2458.pdf

Media

extracted text
OBSTRUCTED LABOUR

References
7 - 7T

The Wellcome Tropical Institute

REFERENCES PROVIDED WITH THE MODULE

A

Driessen, F. Obstetric problems - a manual (1985)
Chapters: 5, 20, 42, 44, 45, 46, 48, 53, 54, 55.

B

Philpott, R. H. Obstructed labour.
Clinics in Obstetrics and Gynaecology (1982) 9, 625-640.

C

Gebbie, D. A M. Symphysiotomy.
Tropical Doctor (1974) 4, 69-75.

D

Heij, H.A.; Te Velde, E.R.; Cairns, J.M. Management of
rupture of the gravid uterus.
Tropical Doctor (1985) 15, 127-131.

E

Lawson, J. Embryotomy for obstructed labour.
Tropical Doctor (1974) 4, 188-191.

F

Mphahlele, M; Van Der Meulen, A. J. Obstructed labour at
the University Teaching Hospital, Lusaka, Zambia,
April 1972-Dec 1973.
South African Medical Journal (1975) 49, 1204-1206.

G

Dutta, D. C. Destructive operations in obstructed labour.
Journal of the Indian Medical Association (1979) 72, 204-206.

H

Bullough, C. Postnatal care. [Influencing the outcome of
future pregnancies].
Tropical Doctor (1988) 18, 79-80.

I

Lawson, J.B.; Steward, D.B. Obstetrics and gynaecology in the
tropics.
pp. 189-192; 203-210.

J

Harrison, K.A.; Rossiter, C.E. Child-bearing, health and
social priorities: a survey of 22,774 consecutive hospital births
in Zaria, northern Nigeria. 12. Maternal mortality.
British Journal of Obstetrics and Gynaecology (1985) 92,
Supplement 5,100-115.

K

Thornton, J.G. Should vesicovaginal fistula be treated only by
specialists?
Tropical Doctor (1986) 16, 78-79.

L

Oronsaye, A.U.; Asuen, M.I. Obstructed labour - a four-year
survey at the University of Benin Teaching Hospital, Benin
City, Nigeria.
Tropical Doctor (1980) 10, 113-116.

M

Chiphangwi, J. Antenatal care in a district hospital.
Tropical Doctor (1987) 17,124-127.

N

Chabot, H.TJ.; Eggens, K.H. Antenatal card for illiterate
traditional birth attendants.
Tropical Doctor (1986) 16, 75-78.

O

Burgess, H.A. Use of the labor graph in Malawi.
Journal of Nurse-Midwifery (1986) 31, 46-52.

P

Philpott, R.H.; Castle, W.M. Cervicographs in the
management of labour in primigravidae. I. The alert line for
detecting abnormal labour.
Journal of Obstetrics and Gynaecology (1972) 79, 592-598.

Q

Philpott, R.H. The recognition of cephalopelvic
disproportion.
Clinics in Obstetrics and Gynaecology (1982) 9, 609-624.

R

Lennox, C.E. The cervicograph in labour management in the
Highland of Papua New Guinea.
Papua New Guinea Medical Journal (1981) 24, 286-293.

S

Naeye, R.L.; Dozor, A.; Tafari, N.; Ross, S.M.
Epidemiological features of perinatal death due to obstructed
labour in Addis Ababa.
British Journal of Obstetrics and Gynaecology (1977) 84, 747750.

T

Adetoro, O.O. Maternal mortality - a twelve-year survey at
the University of Ilorin Teaching Hospital (U.I.T.H.) Ilorin,
Nigeria.
International Journal of Gynaecology and Obstetrics (1987) 25,
93-98.

U

Maternal mortality: helping women off the road to death.
WHO Chronicle (1986) 40, 175-183.

327
Tm HrALTR CEU
32b. V Mam, I Block
Korambng^lg
eangaiore-56j034

Chapter 5

PREGNANCY FOLLOWING CAESAREAN
SECTION (S)
RISKS
- Rupture of the uterus:
- usually during labour.
- sometimes before labour (with a classical scar, inverted
T incision).
- Repeat cephalopelvic disproportion and obstructed labour.

MANAGEMENT
IN EARLY AND MID PREGNANCY
• Encourage the patient to start antenatal clinic at ± four
months.
• Check the haemoglobin and correct anaemia.
• Assess and record the fundal height accurately : this may help
later to establish when the pregnancy is term.
• Arrange for review by gynaecologist, doctor or clinical officer
at about 36 weeks.

AT 36 WEEKS
Patients with two or more previous caesarean sections:
• Estimate the gestational age.
• Admit into hospital unless the patient lives nearby and has
transport easily available, in which case admission can be
postponed until 37 - 38 weeks.
• Arrange for caesarean section to be done either electively or in
very early labour [see below]
Patients with only one previous caesarean section:
• Make an assessment:
• find out why the caesarean section was done
• estimate the gestational age
• check the presentation and size of the baby
• assess the pelvis clinically and by x-ray if necessary
[see Chapter 1]
• Make a decision:
Trial of scar is only allowed with:*
- a cephalic presentation
- a normal size baby
- an apparently normal pelvis
- a normal lower segment scar
* For the conduct of a trial of scar see chapter 31.

12

Caesarean section is indicated for:
- breech presentation or transverse lie
(if persisting to term)
- borderline pelvis
- a large baby (estimated weight 3,500 gm or more)
- a classical scar or inverted T incision
THE PLACE OF ELECTIVE CAESAREAN SECTION
(Elective caesarean section is the operation before the onset of
labour)

Elective caesarean section is a must if the patient is known to have a
classical scar or an inverted T incision. It is best done between 36
and 38 weeks.
Elective caesarean section is desirable in all patients in whom a
repeat operation is indicated, provided the gestational age is
reasonably certain. It is then best done around 38 weeks.

However, if you have reason to believe that the patient does not
want the operation, it may be better not to book her for an elective
caesarean section. Such patients often give themselves a “trial of
scar” at home and only come to hospital when labour has become
obstructed!
It is better, although not ideal, to admit such a patient at 36 weeks,
say that she will be allowed a short trial of scar and then do a caesa­
rean section in early labour. Careful observation in the ward will
still be necessary because these patients tend to retire to a quiet
spot as soon as labour starts to make sure that they do not miss their
“trial”.

THE PLACE OF TUBAL LIGATION
Some patients will want to be sterilised at the time of a repeat
caesarean section. Discuss this during the antenatal visits.
Tubal ligation can be done provided:
- both she and her husband agree to the operation and give
written permission.
- they both understand that ligated tubes cannot be reopened
- they are both aware of the existence of other forms of
contraception

The chance that tubal ligation will be necessary on medical grounds
(extensive adhesions, weak scar) increases with the number of
previous operations.
Tubal ligation is usually done with the fourth operation.

13

Chapter 20

THE LABOURGRAPH
Read this chapter with a real labourgraph in front ofyou.
BACKGROUND
During the first stage of labour the cervix effaces and dilates. If one
plots the cervical dilatation in a graph, the progress of labour can be
seen at a glance. Figure 20.1 shows the progress of an ideal labour. In
the first part of the first stage the cervix dilates only slowly, this is
called the latent phase. The latent phase ends when the cervix has
become fully effaced and is 3 cm dilated. Between 3 cm and full
dilatation the cervix dilates faster; this is called the active phase of
labour.
Both the latent and active phases of labour can be prolonged. The
labourgraph helps us to recognise this quickly, before serious trouble
has developed. A prolonged active phase in particular is often the first
warning of cephalopelvic disproportion.
The Malawi labourgraph spans 22 hours. The first eight hours on the
chart are reserved- for recordings during the latent phase. How
recordings are made when the latent phase lasts longer than eight
hours is discussed later in this chapter. The remaining part of the
chart is for recordings made during the active phase. In it four lines
have been drawn:
- A is the alert line. If the cervical dilation follows this line, the
cervix dilates 1 cm per hour; this is the slowest rate of dilatation
in the active phase which is still normal.
- M and P are the action lines for multiparae and primigravidae
respectively. Action to correct a delay in the active phase should
have been taken before these lines have been reached [see
Chapter 22]
- The broken vertical line indicates the expected time of delivery
[see Chapter 22]
RECORDING ON THE LABOURGRAPH
CERVICAL DILATATION
On admission
If the cervix is still less than 3 cm dilated, the dilatation is recorded
at time 0 with an x. Write date and clocktime underneath [see Figure
20.2]. If the cervix is 3 cm or more dilated, the dilatation is recorded
on the alert line in the active phase [see Figures 20.3 and 20.4].

56

Later recordings
1. In the latent phase when the patient is examined again
four hours after admission, there are two possibilities:
- the cervical dilatation is still less than 3 cm. The dilatation
is then recorded at time 4 hours [see Figure 20.2).
- the cervix is 3 or more cm dilated. This dilatation is also
recorded at time 4 hours, but the curve is then
“transferred to the alert line” [see Figures 20.5 and 20.6].
If the latent phase is prolonged, the cervical dilatation will
still be less than 3 cm after eight hours. It can still be
recorded at the appropriate times on the chart, but when
the active phase is reached eventually, a new chart must be
started [see Figure 20.6].

2. In the active phase. The cervical dilatation is plotted at
the appropriate times [See Figure 20.3].
If the cervix dilates normally the curve is on or to the left of
the alert line. If the curve is on the right of the alert line,
the active phase is becoming prolonged.
How often should vaginal examinations be done?

- Once labour is established vaginal examination is done four
hourly in a nullipara and three hourly in a multipara
— It should also be done at expected time of delivery
It should be done earlier if:
- the membranes rupture
- the patient wants to push
- the cervix was 7 cm or more dilated at the last
examination
- signs of fetal distress develop.
DESCENT OF THE HEAD
The baby’s head is, on examination, divided into five horizontal,
equal parts and as labour progresses the number of parts (fifths)
remaining above the brim is assessed and recorded on the labour­
graph with a dot ‘ ” or an “o” [see Figures 20.2 - 20.6]. The number
of fifths still above the brim should be determined by abdominal pal­
pation or better still by bimanual palpation at the time of a vaginal
examination. The number of fingers that can be placed between the
anterior shoulder and the symphysis indicates the number of fifths
above the brim. Vaginal examination alone can be very misleading:
due to moulding and caput formation the fetal scalp may be seen at
the outlet while the largest part of the head is still above the brim.
57

CONTRACTIONS
The duration of the contractions is recorded as follows:
more than
20-40
p"n le8S than
40 seconds
seconds
Liil 20 seconds
I

□

The number of contractions per ten minutes is recorded as follows:

__ 1/10 minutes

^■3/10 minutes

5/10 minutes

FETAL HEART RATE
The fetal heart rate is counted between the contractions, preferably
with the patient lying on her side. If she lies on her back the supine
hypotensive syndrome can cause slowing of the fetal heart.
The normal heart rate between contractions is 120 - 160 beats per
minute. Changes during the contractions are best ignored. They are
difficult to assess with the fetoscope and their significance is open to
question.

MEMBRANES AND LIQUOR
The following symbols are used:
I = membranes intact
C = clear liquor draining
M = meconium stained liquor draining
R = membranes ruptured but no liquor draining at present
If the membranes rupture during labour, this should be recorded at
the appropriate time [see Figure 20.5].
The following abbreviations are used:
SRM = spontaneous rupture of membranes
ARM = artificial rupture of membranes

MOULDING
This is graded as follows:
0 = bones normally separated
+ = bones touching-but not overlapping
++ = bones overlapping but easily separated
+++ = bones overlapping and cannot be separated
MATERNAL CONDITION
Pulse rate, blood pressure and temperature are recorded in their
appropriate columns.

58

I

ADDITIONAL NOTES
Short comments can be written on the labourgraph. Longer
notes should be written on a separate sheet of paper.

cervical
dilatation
descent.

time

utnin

Figure 20.1
The curves of cerival dilatation and
descent during a normal labour (not
plotted on a labourgraph)

Kim

FHAM

PH AH

Z

/

i ^3=

iLz

US
Wo-If 41

IF
jh

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

I

Figure 20.2
Labourgraph of a patient admitted in the latent phase at 02.00 on
7.10.84. Four hours later she was still in the latent phase.

59

I

LATENT

PHASE

t;

I HASE

activx

✓
C

■
i

r

D

a

5
0fe'/O

|

TIME
MMMM

I

|

1

I

I

I

I

I

1

Figure 20.3
Labourgraph of a patient admitted at 06.15 on 10.10.84 with the
cervix 100% effaced and 3 cm dilated. Good progress after four
hours.

I-A TENT

ACTIVE

PHASE

IH VSE

z
I
c

5
t—)

r

D
S

5
TIME

I

ih

a

ih

ill

i

I

M'”!

ai

it

JL

I

l

I

I

I

I

Figure 20.4
Labourgraph of a patient admitted at 09.30 on 3.7.84 at 7 cm
dilatation.

60

I

I

term

LATF.XT

FHAtt

“7

✓

$

1 r:
I

■

I

c

LI:
nm

i

1» in
illf

it'#

I

I

I

I

I

I

4*

j*

I

I

I

I

I

11

I

I

I

Figure 20.5
Labourgraph of a patient admitted at 14.15 on 9.9.84 with 2 cm
dilatation. Four hours later the cervix was 5 cm dilated,

rturr cmawt
UTtirT

T I
i_ L.

SECOWO CMAITT
*CTTV«

! ”T ! I

\7

k

fl*

Figure 20.6
Labourgraph of a patient admitted at 03.45 on 2.9.84 who
developed a prolonged latent phase. After 18 hours she was at last
in the active phase.

61

I

I

Chapter 42

SYMPHYSIOTOMY
DEFINITION
Symphysiotomy is the division of the symphysis pubis in order to
enlarge the pelvic diameters during delivery.

INDICATIONS
Symphysiotomy is indicated for moderate cephalopelvic dispro­
portion which has resulted in one of the following problems:

1.

OBSTRUCTED LABOUR WITH A LIVE BABY
If the baby’s head is deeply jammed into the pelvis and the
caput is visible in the vulva, symphysiotomy is life-saving for
the mother and therefore a “must”! The alternative,
caesarean section, will be disastrous because of tears in the
lower segment, bleeding and sepsis.

2.

FAILED TRIAL OF VACUUM EXTRACTION
Symphysiotomy can be done when a trial of vacuum
extraction has failed by a small margin.
It will not work, of course, if the cephalopelvic disproportion
is gross and the indication for vacuum extraction was wrong
in the first place.

3.

DIFFICULT VACUUM EXTRACTION
Sometimes vacuum extraction succeeds but only with
difficulty after prolonged traction and at the expense of
brain damage to the baby.
Symphysiotomy will make the delivery much easier and
helps to avoid injury to the baby.

4.

PROLONGED SECOND STAGE
If the criteria for symphysiotomy are met [see below] and
vacuum extraction alone is unlikely to succeed, immediate
symphysiotomy is better than trying a vacuum extraction
first.

CONTRAINDICATIONS
- Severe cephalopelvic disproportion
- Malpresentation: breech, brow, face, transverse lie
- Dead fetus (craniotomy is preferable)

116

r

- Previous caesarean section
- Maternal deformity (spine, leg)

CRITERIA
- Live baby
- Cervix 8 cm or more dilated
- No overlap (the head should not bulge above the symphysis)
- The head is not more than 2/5 and not less than 1/5 palpable
above the symphysis
If the head is too high, symphysiotomy will fail because of severe
disproportion, if it is too low, a simple vacuum extraction will be
sufficient.

PREPARATION
Ideally you need three helpers; two to hold the legs and one to keep
the i.v. drip going, catch the baby, resuscitate it, etc.
• Explain the procedure to the patient
• Give antibiotics as for caesarean section
• Check that:
- the vacuum extractor is in working order
- the delivery trolley is ready
- the resuscitaire for the baby has been prepared.
• If the contractions are weak and the patient is a nullipara,
start a pitocin drip
• Put the patient in lithotomy position.
Two helpers, who do not do anything else, support the legs; the
thighs should be abducted at an angle of 80°.
This position should be carefully maintained or the strain on
the sacroiliac joints becomes too great; lithotomy poles are
unsuitable.
• Disinfect the skin over the lower abdomen, symphysis and
vulva; use iodine over the symphysis
• Infiltrate the skin and subcutis over the symphysis and the
fibrocartilage with 1 or 2% lignocain; also infiltrate the
perineum for the episiotomy
• Insert a plain catheter and leave this in
• Apply the cup of the vacuum extractor in the usual manner
• Pull once gently on the vacuum extractor (sometimes delivery
is easier than expected and symphysiotomy is not necessary)

117

I

I

THE ACTUAL PROCEDURE

• Insert one finger between the fetal head and the symphysis,
this automatically pushes the catheter and urethra out of the
midline (if you use two fingers, the catheter will tend to stay
in the midline)
• Using a scalpel (ideally with a fixed handle) make a vertical
stab incision over the symphysis midway between its upper
and lower border
• Keeping exactly to the midline, push the knife through the
fibrocartilage [see Figure 42.1] until your finger in the vagina
feels its point. If this is difficult, you are not in the midline;
retreat and try again
• Guided by the finger in the vagina, the knife cuts the lower
half of the fibrocartilage and the fibres of the arcuate
ligament; stop there!
• Turn the knife 180° and cut the upper part of the
fibrocartilage; the joint should now separate and one finger
should fit the gap. If this is not the case, the division is
incomplete and you must cut a few more fibres

Figure 42.1 The anatomy of the
symphysis pubis
1. arcuate ligament
2. fibrocartilage of symphysis pubis
3. urethra
4. vagina

DELIVERY
• Make a large episiotomy
• Pull on the vacuum extractor during a contraction; when the
head descends, you will often feel the last fibres of the
symphysis snap

118

r

• Deliver the head and the rest of the baby “away” from the
symphysis to avoid more strain on the urethra
• After delivery of the placenta, check the cervix and vagina for
tears
• Repair the skin incision over the symphysis, the episiotomy
and whatever tears there are (usually none, fortunately!)
• Replace the plain catheter by a Foley’s with an open drainage
system
• Put the legs flat

AFTERCARE
• Keep the patient in bed for about three days, there is no need
to strap the legs together
• After three days allow the patient to mobilise at the pace she
finds comfortable; almost all patients walk well within ten
days
• Leave the catheter in:
- for two or three days if the operation was uncomplicated
and the urine is clear
- for one to three weeks if the urine is blood stained or if
pressure necrosis due to obstructed labour is suspected
• Prescribe antibiotics as you would for caesarean section
• Give a simple analgesic like aspirin for a few days
PROBLEMS AND COMPLICATIONS

FAILURE
If you stick strictly to the criteria, failure should be rare. If it
happens, caesarean section is the only alternative
BLEEDING
This stops after delivery of the baby. Before delivery it can be
controlled by firm pressure

IJ.

OSTEITIS PUBIS
This infection of the symphysis pubis is serious and may cripple the
patient for life. However, it should not occur if reasonable aseptic
precautions are taken

INJURY TO BLADDER OR URETHRA
Fistulae can result from faulty technique but are more commonly
due to the obstructed labour which necessitated the symphysiotomy
in the first place.

Incontinence of urine without a fistula is due to avulsion of the

119

urethra from the pubic bones. It is rare but serious as it is
apparently difficult to treat.
PELVIC INSTABILITY
This is remarkably uncommon. Most patients walk well soon after
the operation and have little or no pain. It is, of course, important to
avoid too much abduction of the thighs during the procedure. If this
does occur, a situation comparable to a fractured pelvis results with
extensive soft tissue damage as well. So beware!!

120

Chapter 44

CAESAREAN SECTION
I

i

FOR CLINICAL OFFICERS AND DOCTORS

INTRODUCTION
This chapter is for those of you who have done some caesarean
sections already and who are, therefore, familiar with the technique
of the uncomplicated operation.

L

PREPARATION
ANTIBIOTICS
Caesarean section during labour
Clean cases
Before operation:
- x-penicillin 5 mega i.v. stat. + streptomycin 1 gm i.m. stat.
After operation:
- x-penicillin 2 mega i.v. or i.m., 6 and 12 hours after
operation

!

Infected cases (temperature 38C° or more, offensive liquor or
ruptured membranes of more than 12 hours)
Before operation
- x-penicillin 5 mega i.v. stat. + streptomycin 1 gm i.m. stat.
or
- chloramphenicol 1 gm i.v. stat
After operation:
- x-penicillin 2 mega i.v. or i.m. 6 hourly + streptomycin 1 gm
i.m. daily for seven days
or
- chloramphenicol 0.5 gm i.v. or p.o. six hourly for seven days.

Elective caesarean section
No antibiotics!
I

BLOOD
• Check the haemoglobin
• x-match one pint of blood if possible
CATHETERISATION AND VAGINAL EXAMINATION
The doctor or clinical officer who will do the operation should
catheterise the bladder in theatre.

123

If the patient is in labour do a vaginal examination at the same time
in order not to miss unexpected progress and the possibility of a
vaginal delivery.

TILT THE PATIENT TO THE LEFT
Tilt the patient approximately 15° to the left by placing a pillow or
rolled up towel under the right buttock
ANAESTHESIA

GENERAL ANAESTHESIA
This is only safe if given by a trained anaesthetist who intubates the
patient.
Never assume that the stomach will be empty because the patient
has not taken food for a long time. Emptying of the stomach slows
down during labour.
Aspiration of stomach contents will cause a most violent pneumonia
which will probably kill your patient.

SPINAL ANAESTHESIA
This is a good technique provided:
- you know the technique and its complications in detail
- you put up an i.v. drip and run in 1 - 2 litres of fluid fast
- you tilt the patient to the left I see above]
- you observe the contraindications
- severe antepartum haemorrhage
- severe anaemia
- severe hypertension or preeclampsia
- serious heart disease

LOCAL INFILTRATION ANAESTHESIA
Safe for mother and baby.

Disadvantages are:
- uncomfortable for the mother
- unsuitable if extensive dissection is required (old scars)
- unsuitable for infected cases because packs cannot be placed
by the side of the uterus and thorough cleaning of the
abdominal cavity is difficult
Technique
• give diazepam 10 - 20 mg i.v. stat.
• prepare a ’/2% solution of lignocaine
• inject 10 - 20 ml under the skin in the line of the incision

124

I

I

i

•
•
•
•
•

after a few minutes incise the skin and subcutis
inject another 10 - 20 ml under the fascia
after a few minutes incise fascia, muscles and peritoneum
put a few ml of anaesthetic under the peritoneum of the uterus
deflect the bladder peritoneum, incise the uterus and deliver the
baby
• as soon as the baby is out, inject 50 mg of pethidine i.v., this may
have to be repeated once or twice.
• complete the operation
INCISION OF THE UTERUS

I.

There are four acceptable ways of making an incision in the uterus:
- a transverse incision in the lower segment
- a vertical incision in the lower segment
- a transverse incision in the upper segment
- a vertical incision in the upper segment

The techniques of making and closing these incisions are described
here, the indications follow later in this chapter.
I

A fifth method of incising the uterus, the inverted T incision, is not
described, as, with forethought, it can always be avoided. It heals
poorly and should not be used.
TRANSVERSE INCISION IN THE LOWER SEGMENT
1. Opening of the uterus
• Cut the peritoneum over the lower segment transversely
and mobilise the peritoneum and bladder well down
• Identify the midline of the uterus (it may have rotated to the
right or the left)
• Incise the lower segment transversely with the scalpel over a
short distance
• Enlarge the opening to the right and the left with scissors
(this is safer than stretching the incision with your fingers
which does not allow you good control over where the lower
segment will tear)

2.

Repair of the incision
• Identify the lower edge of the incision by placing GreenArmytage forceps
• Start at a point where the lower edge is easy to see
• When you have placed the first clamp, lift it and a further
stretch of the edge will come into view
• Place another clamp and continue like this until you have
identified the lower edge over its whole length.

125

A common mistake is to pick up a fold of the posterior wall of
the uterus. If you use the above method, that mistake is
almost impossible to make.
• Now repair the uterus with two layers of continuous chromic
catgut No. 1 or No. 2

3.

Difficulties
Heavy bleeding
Bleeding from the edges can be controlled by Green-Armytage
clamps until the incision has been sutured.
A (partly) torn branch of a uterine vessel in one of the corners is
sometimes easier clamped with an ordinary artery forceps. It
may also need to be ligated separately.
Continued bleeding after the incision has been sutured in two
layers needs extra figure-of-eight stitches. Do not pull these too
tight or they will cut through and more bleeding will result.
Continued oozing is best controlled with a hot pack.

Vertical tears
These are often located in the corners and run downward behind
the bladder. Very often there is heavy bleeding as well.
In order to repair these you must be able to see.
• If you are alone with a scrubnurse, ask for an extra assistant
• Identify the edges of the incision and the tear as described
above
• Mobilise the bladder further downward if necessary
• Divide the round ligament if necessary : this often makes the
lower end of the tear much easier to see
• In the area of the tear use interrupted sutures : these are
easier to unpick if you catch the bladder or ureter by mistake
• After repairing the tear check that the ureter has not been
caught in a stitch
• Remember that all bleeding can always, at least, temporarily,
be controlled by pressure with a hot pack

THE VERTICAL INCISION IN THE LOWER SEGMENT
(DE LEE INCISION)

1.

Opening of the uterus
• Incise the peritoneum transversely high on the lower segment
• Mobilise the peritoneum and bladder well down
• Identify the midline of the uterus

126

• Incise the lower segment vertically with the scalpel
• Enlarge the incision with the scalpel or scissors : often the
upper segment has to be entered

2.

Repair of the incision
• Suture the incision with two layers of continuous chromic
catgut No. 1 or No. 2
• Make sure that you include the uterine fascia in the second
layer or it will continue to bleed
• Repair the peritoneum and pull it high up so that the top end
of the incision is covered
• If the upper segment was incised over a long distance, do
tubal ligation

TRANSVERSE INCISION OF THE UPPER SEGMENT
1.

Opening of the uterus
• Check that the uterus is wide enough for an adequate incision
• Incise the peritoneum over the lower part of the upper
segment transversely with the scalpel
• Mobilise the peritoneum away from the incision with scissors
• Incise the uterus transversely in the midline
• Enlarge the incision to the right and left by stretching it with
your fingers (it is usually too thick to be cut with scissors)
• Deliver the baby by breech extraction

2.

Repair of the incision
• Suture the incision in two layers with continuous chromic
catgut No. 1 or No. 2
• Do not catch the full thickness of the uterine wall in the first
layer : it is often too thick
• Repair the peritoneum over the incision, preferably with a
locking stitch

THE VERTICAL
SEGMENT

(CLASSICAL)

INCISION IN THE UPPER

This is a poor incision and should only be used as a last resort.
It has two dangers:
- it may rupture before labour in a future pregnancy
- infection can spread directly from the uterine into the
peritoneal cavity
127

Opening of the uterus
• Identify the midline of the uterus
• Incise the uterus vertically with the scalpel
• Enlarge the incision by stretching it with the fingers
• Deliver the baby by breech extraction
2. Repair of the incision
• Repair the uterus in three layers with a continuous chromic
catgut No. 1 or No. 2 suture; each layer should catch about
one third of the thickness of the uterine wall; the last layer
should be a locking stitch.
• always tie the tubes
1.

CAESAREAN SECTION FOR SPECIFIC INDICATIONS

REPEAT CAESAREAN SECTION
Entry into the abdomen
This can be difficult but the following method is safe:
• Excise the skin scar
• Lift the fascia with two dissecting forceps as close to the
umbilicus as possible and incise fascia, muscles and perito­
neum to open the abdomen
• If entry into the abdomen at this site proves difficult, do not
muddle on
• Extend the skin incision higher up the abdomen, round the
umbilicus if necessary, and try again; by going higher it is
always possible to find an area free of adhesions
• After the abdomen has been entered, lift the abdominal wall
on two fingers and incise the layers one by one: fascia, muscle,
peritoneum; in this way the bladder is easily recognised when
it appears at the lower end of the incision
Separation of the adhesions
After a previous operation there can be dense adhesions between the
uterus and abdomen wall
• Open the parietal peritoneum as far as possible
• Lift the parietal peritoneum with two dissecting or artery
forceps to stretch the adhesions
• Cut the adhesions close to the uterus with curved scissors,
keep the points of the scissors directed at the uterus
• If a plane of loose connective tissue is found, do the further
separation with a finger or a swab; fibrous bands must be cut
• Stay close to the uterus to avoid the bladder
• If dissection of the adhesions proves difficult, give up and
make an upper segment incision

128

I

I

Mobilisation of the bladder
After a previous operation the bladder is stuck to the lower segment
and cannot be mobilised with the finger or a swab
• Incise the peritoneum on the uterus transversely about 2 cm
above the bladder
• Lift the lower edge with two dissecting or artery forceps to
stretch the adhesions between the bladder and the uterus
• Cut the adhesions close to the uterus with curved scissors
keeping the points of the scissors directed at the uterus
• If this proves difficult, give up and make an incision higher in
the uterus
The uterine incision
Usually a transverse incision in the lower segment is possible and
with an elective caesarean section this should provide no further
difficulty. If the section is done during labour, you may find the head
tightly stuck under the old scar in the uterus. If you then make your
incision just above the scar, the old scar will probably rupture
during delivery of the head of the baby. It is safer to make a
V-shaped transverse incision with the point of the V lying across the
middle of the scar. This is thus divided and the tension removed. If
any tearing occurs, it will be near the midline where it is relatively
easy to see and repair
OBSTRUCTED LABOUR WITH A CEPHALIC PRESENTATION
• Enter the abdomen just under the umbilicus in order to avoid
the bladder
• If catheterisation before the operation was impossible, empty
the bladder now with a needle and syringe; a lot of its swelling
will be oedema and this, of course, does not go away
• Mobilise the bladder of the lower segment in the usual manner
• If someone is going to push up the head from below through
the vagina, let them start now before the uterus is opened; if
you postpone this until after the uterus has been opened, the
shoulder will prolapse into the incision making delivery more
difficult
• Make a transverse incision in the lower segment.
Choose the level of the incision carefully:
- if it is too high, delivery of the baby will be difficult
- if it is too low you may enter the vagina
• Take time to deliver the baby.
• between the contractions work two to four fingers
between the uterine wall and the head until they are
under it

129

• lift the head from the pelvis by flexing your fingers
• try to avoid “levering” the head out with your whole
hand because this often causes vertical, downward tears
in the lower segment
• After delivery of the baby and placenta, repair the uterus
• Clean the abdomen very carefully, if the liquor was purulent or
offensive smelling, wash the pelvis with warm saline

BREECH PRESENTATION
There should be no difficulty. Make sure, however, that you deliver
the baby through the incision with exactly the same manoeuvres
that you would use for a vaginal delivery. Failure to do so may cause
extensive tearing of the lower segment.

r

TRANSVERSE LIE
The choice of the incision in the uterus is what matters here.
You will meet the following situations
- Labour obstructed
Most of the baby is in the overdistended lower segment. Simple
delivery through a transverse lower segment incision will
cause extensive tears. Therefore:
- if the baby is dead
• make a transverse incision in the lower segment
• decapitate or eviscerate the baby
• now deliver it whichever way is convenient
- if the baby is alive
• make a vertical incision in the lower segment
• extend the incision into the upper segment until you
have enough room to deliver the baby safely

- The patient is in early labour but the lower segment is
poorly developed. Most of the baby is in the upper
segment
• make a transverse incision in the upper segment
• deliver the baby by breech extraction

- The patient is in early labour, the lower segment is well
developed and the membranes are still intact
• make a transverse incision in the lower segment
• deliver the baby by breech extraction
PLACENTA PRAEVIA
Usually the normal transverse incision in the lower segment is
possible. If the lower segment is too small for an adequate incision
or if it is very vascular, make a transverse incision in the lower part

130

/

I

of the upper segment. If this is also very vascular, make a classical
incision.

CONSTRICTION RING
Sometimes the cause of obstruction is a constriction ring either in
the lower segment or between the lower and upper segment.
If the baby is entirely above it, make a transverse incision above the
constriction. If the constriction is round the baby’s neck, make a
vertical incision across the constriction.

OTHER PROCEDURES DURING CAESAREAN SECTION
FIBROIDS
Leave these alone. Removal will cause heavy bleeding. If necessary,
they can be removed three months later by a second operation.

OVARIAN CYSTS/TUMOURS
These should be removed. Ovarian cystectomy is possible but if
bleeding is a problem, salpingo-oophorectomy may be quicker and
safer.
ADHESIONS
Adhesions should be separated sufficiently to gain good access to the
uterus. However, there is no point in removing adhesions round the
adnexa. This often causes troublesome oozing and the adhesions will
invariably form again.

BLADDER INJURY
See Chapter 47
TUBAL LIGATION
See Chapter 52

POSTOPERATIVE ORDERS
Routine are:
- intravenous fluids: one to three litres over the first 24 hours,
depending on the patient’s condition
- pain relief : pethidine 50 - 100 mg six hourly for two days;
after that a simple analgesic like aspirin
- antibiotics : as outlined under PREPARATION
- catheter
: open drainage only when indicated (blood
stained urine, bladder repair, obstructed labour)
- observations: vital signs, fundus, vaginal bleeding
131

Chapter 45

REPAIR OF A RUPTURED UTERUS
INTRODUCTION : REPAIR OR HYSTERECTOMY?
In the course of an operation for ruptured uterus you will have to
decide what to do; repair the rupture or do hysterectomy. The
following hints are meant to help you make the right decision.
If you have very little or no experience of hysterectomy, repair of the
uterus is nearly always best. Hysterectomy will take longer and can
cause more bleeding. Only do a hysterectomy when extensive
tearing of the uterus makes a repair impossible.
If you do have some hysterectomy experience, your decision can
depend on the situation.
Factors in favour of a repair are:
- rupture not too large
- edges clean and easy to see
- little or no infection present.
Factors in favour of hysterectomy are:
- extensive or multiple tears of the uterus
- edges which are necrotic or not easy to reach for suturing
(some posterior ruptures, ruptures extending down into the
vagina)
- gross infection of the uterus
This chapter describes repair of the uterus, hysterectomy is dealt
with in the next chapter.
PREPARATION
Rescuscitate the patient
• Put up i.v. drip with large bore needle (No. 18) or cannula
• Give 1-2 litres of saline or sodium lactate before starting the
operation
• x-match 2 pints of blood
Antibiotics
• Give, for example:
- x-penicillin 5 mega i.v. stat. + streptomycin 1 gm i.m.
stat
or
- chloramphenicol 1 gm i.v. stat.
• Catheterise the bladder

132

r

I

ANAESTHESIA
If the patient’s condition is poor, local infiltration anaesthesia [see
Chapter 44] is safest.
If general anaesthesia is used, endotracheal intubation is essential.
Do not: use spinal anaesthesia for patients with a ruptured uterus.
-------

STAFF AND INSTRUMENTS
Needed are:
- you, the surgeon
- a scrubnurse
- if at all possible, a second scrubbed assistant
- a “runner”
- an anaesthetist
The set of instruments used for caesarean section will do, provided
you add some large curved clamps or artery forceps.
TECHNIQUE

• Open the abdomen through a midline incision
• Remove fetus and placenta:
- this is easy if they are free in the abdomen
- if the fetus is in the broad ligament, open the broad
ligament: this is often most easily done by dividing the
round ligament over it
- if the fetus is still inside the uterus (with a posterior
rupture for example) you may have to make the
transverse incision in the lower segment, as for caesa­
rean section, in order to deliver the baby
• Suck away most of the blood and liquor
• Lift the uterus from the abdomen and assess the damage
• Identify the edges of the tear along its whole length
• Make sure to separate the bladder well away from the edge*
• Divide the round ligament if this makes the tear easier to see
• Trim obviously dead tissue away. Do not trim too much as this
makes the repair more difficult and causes bleeding
• Repair the tear in one layer with a continuous chromic 1 or 2
suture. A vertical tear going down to the cervix can be
repaired from below upwards but sometimes the other way
round is easier: traction on the suture helps to bring the lower
end of the tear into view. Carefully identify the edges before
*For repair of bladder injury See Chapter 47.

133

putting in the stitches in order not to include the ureter
• If there is oozing from the broad ligament, put in a drain. This
can be brought out either through the tear into the vagina or preferably extraperitoneally - through the abdominal wall
• Repair the peritoneum over the uterus
• Tie the Fallopian tubes
• Clean the abdomen and wash it with warm saline
• Close the abdominal wall in layers; use deep tension sutures if
you expect peritonitis
POSTOPERATIVE ORDERS
• I.v. fluids and blood depending on the patient’s condition
• continue antibiotics, for example:
- x-penicillin 2 mega i.v. six hourly + streptomycin 1 gm daily
for seven or 10 days
or
- chloramphenicol 0.5 gm i.v. six hourly for seven or 10 days
• Pethidine 50 - 100 mg i.m. six hourly for two days
• Remove the drain after one or two days
• Open bladder drainage for 10 - 14 days if the bladder was
damaged
• Nasogastric tube if the bowels are distended or peritonitis is
expected.

134

r

Chapter 46

SUBTOTAL HYSTERECTOMY
INDICATIONS
There are two situations in which even the inexperienced surgeon
must attempt hysterectomy if somebody more senior is not
available:
- a ruptured uterus with tearing so extensive that repair is
impossible
- postpartum haemorrhage not responding to treatment [see
chapters 34, 49, and 56).
PREPARATION

Resuscitate the patient
• Put up an i.v. drip with a large bore (No. 18) needle or cannula
• Give 1-2 litres of saline or sodium lactate before starting the
operation
• x-match 2 pints of blood
Antibiotics
• Give, for example:
- x-penicillin 5 mega i.v. stat. + streptomycin 1 gm i.m.
stat.
or
- chloramphenicol 1 gm i.v. stat.

Catheterise the bladder

ANAESTHESIA
If the patient s condition is poor, local infiltration anaesthesia [see
chapter 44).
I

I

If general anaesthesia is used, endotracheal intubation is essential.
Do not use spinal anaesthesia.

Needed are:

STAFF AND INSTRUMENTS
- you, the surgeon
- a scrubnurse
- a scrubbed second assistant (essential!)
- a “runner”
- an anaesthetist

135

The set of instruments used for caesarean section will do, but you
must add large curved artery forceps or kochers, at least 12! A
self-retaining abdominal retractor is a great help.
GENERAL POINTS OF TECHNIQUE
Subtotal hysterectomy, which leaves the cervix and perhaps part
of the lower segment in place, is easier to perform than total
hysterectomy. It causes less bleeding and there is almost no danger
to the ureters. Subtotal hysterectomy is still possible when a uterine
rupture extends down into the cervix and vagina. In that case the
tear in the cervix and vagina is repaired after the body of the uterus
has been removed.

Removing the adnexa (adnex = tube + ovary) is often easier than
leaving them in place. The reason is that the pedicle of the
infundibulopelvic ligament is usually smaller and easier to handle
than the pedicle of the cut Fallopian tube and ovarian ligament.

Traction on the uterus throughout the procedure is the key to
success in hysterectomy. Traction makes it easier to identify the
structures that have to be divided and it helps to keep bladder and
ureters out of the way.
The anatomy. Make sure to identify the important structures and
landmarks before you start removing the uterus. With a ruptured
uterus the anatomy can be difficult to recognise and this does not
become any easier if you start cutting without knowing exactly
what you are doing.

Control of bleeding. For the control of bleeding concentrate on:
- the ovarian vessels either in the pedicle of the infundibulo­
pelvic ligament or in the stump of the cut tube and ovarian
ligaments.
- the uterine vessels
- the stump of the cervix or lower segment.
There can be some bleeding from other small vessels but this
does not usually cause problems.
TECHNIQUE IN DETAIL

(AS FOR RUPTURED UTERUS)
• Open the abdomen
• Remove fetus and placenta [see chapter 45]
• Clean away most of the blood and liquor
136

• Insert a self-retaining abdominal retractor if this is available
• Lift the uterus from the abdomen
• Maintain traction on the uterus with one hand or put in a
traction suture
• Identify the following structures:
- the corpus uteri
- the round ligaments
- tube and ovary on both sides
- the infundibulopelvic ligaments on both sides
- the avascular area in each broad ligament
- the lower segment
- the bladder
- the rectum
[see figure 46.1]
..

4

5

Wk
v*Figure 46.1 The uterus seen from behind prior to hysterectomy
1. uterine corpus
6. uterine vessels
2. Fallopian tube
7. rectum
3. ovary
8. sacro-uterine ligament
4. round ligament
9. avascular area in broad ligament
5. ovarian vessels
Note that the uterus is being pulled over to the left and that the right adnexa are
being lifted in order to show the anatomy clearly

137

• Identify the rupture and clamp obvious bleeding points
• Pull the uterus to the left and divide the right round ligament
between clamps about 2 cm from the uterus; this step opens
the anterior peritoneal leaf of the broad ligament
• Enlarge the opening in the anterior leaf of the broad ligament
with scissors in a downward direction towards the bladder
• Lift the right adnexa with one hand and push a finger of the
other hand from behind through avascular area in the broad
ligament; this step helps to define the infundibulopelvic
ligament
• Clamp the infundibulopelvic ligament with two artery forceps
and cut it
Alternatively, if you wish to leave the adnexa in place, clamp
and divide the tube and ovarian ligament near the uterus. If
the tube and ovarian ligament are very thick and vascular, the
clamping and cutting may have to be done in two steps
• Suture ligate the pedicles of the round ligament and
infundibulopelvic ligament (or the cut tube and ovarian
ligament)
• Repeat the same procedure on the left side
• Now pull the uterus well up in the midline and cut the
peritoneum between the uterus and bladder; extend the
incision laterally to meet the incisions in the anterior leaves of
the broad ligaments
• Push the bladder off the lower segment with your finger or a
swab on a holder; two or three centimeters is enough, pushing
it down further can cause bleeding;
if the rupture is in the anterior lower segment, you have to put
its edge on stretch with Green-Armytage forceps before you
can separate off the bladder
• Now expose the posterior lower segment by pulling the uterus
forward over the symphysis pubis
• Divide the peritoneum over the posterior lower segment at
about the same level as this was done anteriorly
• Extend the incision laterally to join the openings in the broad
ligaments
• Push the lower flap of the peritoneum off the lower segment
with a swab on a holder or if this is difficult, cut it loose with
scissors
• Now review the situation:
on either side of the uterus you should see a bundle of loose

138

r

connective tissue which holds the uterine vessels; you may
have to strip down the peritoneum of the broad ligaments a
little further to see them more clearly
• Pull the uterus to the left and clamp the uterine vessels on the
right with strong curved kocher or artery forceps just above
the level where the bladder is still attached to the lower
segment; make sure the points of the clamp are very close to
the uterus (there is no harm in including a little uterine wall!)
• Place a second clamp inside the first and cut the uterine
vessels in between
• Sutureligate the pedicle
• Repeat this procedure on the left
• Now amputate the uterus through the lower segment just
above the level of the cut uterine vessels; have artery forceps
ready to pick up the cut edge of the lower segment before it
disappears in the depth of the pelvis
• Clamp obvious bleeders
• If there is a downward tear in the cervix, repair this now after
making sure that bladder and ureters are well out of the way
• Now suture the anterior wall of the lower segment to the
posterior wall with figure-of-8 stitches; make sure you
include the angles on the left and right as these tend to bleed;
leave the centre open for drainage
At this stage the pelvis should be more or less dry.

• Look for remaining bleeding points and ligate these
• If there is a lot of oozing from one of the broad ligaments, place
a rubber drain in that area and bring it out either through the
cervix into the vagina or (preferably extra-peritoneally)
through the abdominal wall
• Close the pelvic peritoneum with a continuous suture; start on
the left at the pedicle of the infundibulopelvic ligament and
suture the anterior edge of the peritoneum to the posterior
edge and place all vascular pedicles under the peritoneum
• Wash the abdomen with warm saline and close it.
POSTOPERATIVE ORDERS

• i.v. fluids and blood depending on the patient’s condition
• Continue antibiotics, for example:
- x-penicillin 2 mega i.v. six hourly + streptomycin 1 gm
daily for seven or 10 days

139

or
- chloramphenicol 0.5 gm. i.v. six hourly for seven or 10
days
• Pethidine 50 - 100 mg i.m. six hourly for two days
• Remove the drain after one or two days
• Open bladder drainage for 10 - 14 days if the bladder was
damaged
• Nasogastric tube if the bowels are distended or peritonitis is
expected.

I

140

I

Chapter 48

DESTRUCTIVE OPERATIONS
GENERAL
You should try not to do a caesarean section for obstructed labour, if
the baby is already dead. Often a destructive operation is easier and
safer, because it carries less risk of bleeding and infection and also
because it leaves no uterine scar. However, for a destructive
operation to be safe, three points should always be observed:
- the indication must be correct
- you should follow the correct operative technique
- you should be able to do an immediate laparotomy when you
discover a rupture of the uterus during the operation.

The following sections explain how destructive operations should be
done.

CRANIOTOMY
(The destructive operation for cephalic presentations)

INDICATION
You should do craniotomy if:
- the fetus is dead
- the head presents
- 2/5 or less of the head is above the brim (if the head is higher,
caesarean section is usually safer!)
- the cervix is fully dilated
- the uterus is not ruptured
PREPARATION
In the labour ward

• put up an i.v. drip with dextrose 5% in water
• give high doses of antibiotics, e.g.
x-penicillin 5 mega i.v. stat, with streptomycin 1 gm i.m. stat
or
chloramphenicol 1 gm i.v. stat
• give painrelief, e.g.
pethidine 50 mg slowly i.v.

• take bloodsamples for haemoglobin and x-matching
• shave for a vaginal operation and laparotomy

142

In theatre (Do not do a craniotomy anywhere else!)
• check the preparations done in the labour ward
• check the i.v. drip
• check the instruments on the trolley
• check that everything is ready for laparotomy
Staff
Needed are:
- you, the surgeon
- one anaesthetist
- one scrubnurse
- one “runner”

Anaesthesia
General anaesthesia with intubation is best. Do not use general
anaesthesia without sia without intubation. If general anaesthesia is
not possible, use spinal anaesthesia or local infiltration anaesthesia
of the perineum and vulva.

)

If local infiltration anaesthesia is used, sedate the patient with:
- pethidine 25 - 50 mg i.v. slowly (N.B. Check what she was
given in the labour ward)
- diazepam 10 - 20 mg i.v. slowly.
Technique
• Put the patient in lithotomy position
• Clean and drape the vulva and perineum
• If local anaesthesia is used, infiltrate the perinuem with ’A or
1% lignocaine
• Catheterise the bladder with a rubber or plastic catheter [see
Chapter 38]
• Put one or two Sims specula into the vagina so that you can see
the head well. Ask an assistant to hold the specula
• With the knife make a cross-shaped incision through the skin
of the head right down to the bone
• With a finger feel for a gap (a suture line or a fontanel)
between the bones
• Push a closed pair of scissors between the bones
• Now open and close the scissors a few times while turning
them round (the brain should now be coming out from the
hole)
• Put a finger through the hole in the skull and check that all
brain compartments have been broken up
• Put 3 or 4 strong volsellum forceps or kochers (even better are
Willet’s forceps) on the skin or the skin and the bone
• Pull on these forceps and try to turn the posterior fontanel
under the symphysis

143

• If sharp edges of bone come sticking out, protect the vagina
with your fingers or remove the offending bone
• Make a large episiotomy
• Deliver the head

After delivery of the head, sometimes delivery of the shoulders is
still difficult. In that case:
• Put a hand behind the baby in the vagina and turn the
shoulders through 90° or even 180°. Try delivering the
shoulders again.
If the shoulders cannot be delivered by turning them, you
must bring down the arms one by one.
• Put a hand behind the fetus in the vagina and feel for the
posterior shoulder and arm. Pull the arm down gently (the arm
can break, but you should not damage the cervix or vagina).
• After delivering the first arm turn the fetus 180° and deliver
the second arm in the same way. Further delivery should now
be easy.
After delivery of the fetus:
• remove the placenta manually
• Give ergometrine 0.5 mg i.v. stat.
• After removal of the placenta feel immediately for tears of the
uterus or lower segment
• Inspect the cervix, vagina and vulva carefully for tears
• Repair the episiotomy and tears
• insert an indwelling catheter for open bladder drainage
• If the uterus is not well contracted, put pitocin 10 - 40
units per 1,000 ml in the i.v. drip.

POSTOPERATIVE ORDERS
• continue the i.v. drip slowly for about 24 hours
• continue with antibiotics in high doses, for example:
- x-penicillin 2 mega six hourly and streptomycin 1 gm i.m.
daily for seven days
or
- chloramphenicol 500 mg six hourly for seven days
• open bladder drainage for 14 days

DESTRUCTIVE OPERATIONS FOR TRANSVERSE LIE

GENERAL

Obstructed labour with a transverse lie and a dead baby is a difficult
problem. Destructive operations for transverse lie through the
144

vagina are more difficult than craniotomy. However caesarean
section also has its dangers and should - where possible - be avoided.
The operations that can be done are the following:

!

I

Decapitation : The fetus* neck is divided and the body and head
are then delivered separately
Evisceration : The fetus’ chest and/or abdomen are opened and
all internal organs are removed. The trunk
collapses and delivery by either internal version or
decapitation becomes much easier.

Caesarean section in combination with destruction : A
transverse incision is made in the lower uterine segment.
Decapitation or evisceration is done through the uterine incision.
IMPORTANT DONTS
Never attempt internal version without doing evisceration first. The
risk of rupturing the uterus is enormous. Cutting off the prolapsed
arm does not make version any safer!
Do not attempt decapitation or evisceration through the vagina if
the fetus is still high in the birthcanal. It is dangerous because you
will not be able to protect the vaginal wall and cervix adequately
during the operation.
I

I

At caesarean section do not attempt to deliver the fetus intact
because this will cause severe tearing of the lower segment.

Do not make a classical or inverted T incision in the uterus for a
dead fetus.

CRITERIA FOR DECAPITATION OR EVISCERATION BY THE
I

I

VAGINAL ROUTE
- the fetus is dead
- the lie is transverse
- the cervix is 8 cm or more dilated
- the uterus is not ruptured

You can only decide which operation you are going to perform after
you have examined the patient under anaesthesia.
Preparation
The preparation of the patient in the labour ward is exactly as for
patients undergoing craniotomy. In theatre the preparations are
also the same

145

Anaesthesia
See the section on craniotomy. Good anaesthesia is even more
important than with craniotomy because the operation is done
higher in the birthcanal.
Technique
• put the patient in lithotomy position
• clean and drape the vulva
• catheterise the bladder

Now with the patient under anaesthesia examine her again. Put one
hand in the vagina and support the fundus with the other. Note the
following:
- cervical dilatation: If it is less than 8 cm, caesarean section
is probably safer
- the lower uterine segment : explore it as far as you can
without using force. If you find it ruptured,
do a laparotomy now
- the exact position of the fetus : which arm has prolapsed?
where are the head and the neck?
where are the chest, abdomen and back?
Now decide:
- if the neck and the body are both still high in the birthcanal,
do a caesarean section
- if the neck can be reached easily, attempt decapitation
- if the neck is difficult to reach but the body is well down,
attempt evisceration

Decapitation
a) Using the decapitation saw (this is very well described in
Lawson and Stewart’s Obstetrics and Gynaecology in the
Tropics)
• hook the end of the saw in the thimble
• put the thimble on the best finger of your best hand and try
to bring it round the neck
This is often difficult because there is little room between
the neck, the head and the chest. Sometimes it is easier to
put the saw over the neck and under the arm
• when the saw is in position, protect the vagina with specula
• apply firm traction and saw through the neck
• pull on the arm : usually this delivers the body
• put a hand in the vagina and turn the head so that the neck
points downwards
• put one or two volsellum forceps on the neck and deliver the
head like the aftercoming head of a breech

146

I
1

I

• if the head was delivered first, deliver the body by pulling
on the other arm. Don’t do a version, the cut neck might
damage the uterus.
b) Using scissors
• hook one or two fingers round the neck and pull it down
• protect the vaginal wall with a speculum held by an
assistant
• carefully cut the neck with a pair of strong scissors
• further delivery will be as described above.
Evisceration
• your assistant should pull on the prolapsed arm
• protect the vaginal wall with one or two specula
• with a knife or a pair of strong scissors make a large
opening in the abdomen and/or chest
• put one or two fingers into the opening and remove all
internal organs. Make sure you remove the liver, the heart
and the lungs. Sometimes the diaphragm has to be
perforated with scissors.
• Now reassess the situation:
• sometimes the breech can be brought down easily by
hooking a few fingers behind the fetus pelvis; further
delivery is then no problem
• sometimes a foot or a leg can be felt easily and this can be
brought down. The operation is then completed by a very
gentle version and breech extraction
• if the breech cannot be delivered easily, the neck can be
brought down for decapitation by pulling on the arm
• in the unlikely event that all this fails, don’t hesitate to do
a caesarean section.
After delivery of the fetus see the section or craniotomy

I

POSTOPERATIVE ORDERS
See the section on craniotomy. Obstructed labour with a transverse
lie usually does not cause pressure necrosis of the vagina and
bladder. Open bladder drainage for a few days is sufficient.

HYDROCEPHALUS
i

I

DIAGNOSIS
On abdominal palpation
The head is large in proportion to the fetal body

I

147

On vaginal examination
With a cephalic presentation the wide sutures and fontanels are felt.
If the patient presents with a stuck breech, bimanual palpation will
reveal the large size of the fetal head.
x-ray
An abdominal x-ray may confirm the diagnosis but beware with
breech presentations: on the x-ray the head may seem much bigger
than it actually is (this is due to the way it is projected onto the film)
RISK
Obstructed labour with rupture of the uterus

MANAGEMENT
Cephalic presentation
If progress in the first stage is good, prepare for vaginal delivery.
• When the cervix is about 7 cm dilated, insert a good size needle
(lumbar puncture needle, for example) or plastic cannula into
the head through a suture and drain the cerebrospinal fluid.
The head will collapse and delivery becomes easy
• Do not wait for full dilatation (the large head will stay high
and full dilatation may never come)
• If, during the first stage of labour, progress is poor, do
caesarean section. At caesarean section drain the cere­
brospinal fluid from the head with a needle before you deliver
it. If you do do not do that, there can be nasty tears of the
lower segment.

!

I

Breech presentation
If progress is good, wait until the trunk has been delivered up to the
neck. Insert a needle through and then under the skin of the neck
into the head and drain the cerebrospinal fluid. Delivery then
becomes easy.

If progress in the first or second stage is poor, do a caesarean section
but drain the cerebrospinal fluid from the head before delivering it
through the uterine incision.

I
IMPORTANT DONATS
- Do not try to force a too large head through the pelvis
- Do not use pitocin

i

i

I
THE STUCK BREECH

DIAGNOSIS
The usual manoeuvres to deliver the head (Smellie - Veit, supra
pubic pressure) have failed and the baby is dead.

148

CAUSES
- cephalopelvic disproportion
- incompletely dilated cervix
- hydrocephalus
MANAGEMENT
• For management of the hydrocephalic head see previous
section
• In other cases sedate the patient with pethidine 50 mg i.v. + 50
mg i.m.
• Wait for about an hour, preferably with the patient in
lithotomy position and the baby’s body hanging down;
often the head is delivered spontaneously after it has had time
to mould.
If this fails, craniotomy is necessary. This is best done in theatre
under general anaesthesia but it can be done in the labour ward.
• Retract the anterior vaginal wall with a Sims’ speculum and
expose the posterior aspect of the neck
• With scissors cut a small opening in the skin of the neck
• Make a tunnel under the skin and push the scissors into the
head
• Open and close the scissors a few times in different directions
to break up the brain compartments
• Pull gently on the neck and while the brain gradually escapes,
the head is delivered.
IMPORTANT DON'T
• Do not pull the head with great force through an incompletely
dilated cervix: this can cause cervical tears which extend into
the lower segment.

i

i

I

ii

149

Chapter 53

PUERPERAL INFECTIONS
DEFINITIONS
Fever postpartum : A temperature of 37.5 C° (under the arm) or
more on any two days within 14 days after
delivery
Puerperal sepsis : Infection of the genital tract following
delivery.

CAUSES OF FEVER POST PARTUM
PUERPERAL SEPSIS

M icro-organisms
Many different micro-organisms can cause puerperal sepsis. They
can be classified as follows:

- sexually transmitted organisms : gonococci, chlamydia
These organisms are present in the cervix before delivery
- large bowel bacteria
There are many different kinds of large bowel bacteria.
Important examples are : Escherichia coli and Bacteroides
fragilis. These bacteria are carried from the perineum into the
birthcanal for example by a vaginal examination.
- other bacteria
For example : staphylococci, beta-haemolytic streptococci etc.

Factors which encourage puerperal sepsis
These include:
- poor hygiene or poor aseptic technique during delivery
- manipulations high in the birthcanal
- the presence of dead tissue in the birthcanal after delivery, for
example retained products or sloughing vaginal wall following
obstructed labour.
Forms
Depending on how the infection has spread puerperal sepsis can
present as:
- endometritis
- salpingo-oophoritis
- parametritis (= pelvic cellulitis) : the infection has spread
through the wall of the uterus into the broad ligament

160

- generalised peritonitis
- septicaemia: the infection attacked a vein in the pelvis (a
branch of the ovarian or uterine veins) and infected emboli are
discharged into the circulation.
- abcesses:
- in the pelvis as:
- tubo-ovarian abcess
- broad ligament abcess
- abcess in the pouch of Douglas
- in the subphrenic spaces
- in multiple places in the abdomen:
- in the paracolic gutters
- between small bowel loops
CAUSES OF FEVER OTHER THAN PUERPERAL SEPSIS
Urinary tract infection
- Wound infection (episiotomy, abdominal wound)
- Chest infections : pneumonia or lung abcess (due to aspiration
during anaesthesia)
- Mastitis
- Deep vein thrombosis
- Medical conditions: malaria, typhoid, meningitis etc.

ASSESSMENT
Take a history and examine the whole patient for possible causes of
fever. Include a rectovaginal examination if the cause of fever is not
immediately obvious.
After your clinical assessment you will find one of the following
situations:
1. The cause of fever is (reasonably) certain
For example : endometritis or urinary tract infection
• Treat the patient according to your diagnosis

2.

No cause of fever is found
If the patient is not very ill:
• send blood for haemoglobin, white blood count and malarial
parasites
• Prescribe a full chloroquin course
• review the patient daily : look for localising signs or signs of
improvement

161

I

If the patient is very ill:
• send blood for haemoglobin, white blood count and malarial
parasites
• send urine for microscopy and culture if possible
(take specimen by suprapubic aspiration or catheter)
• do, if possible, one or more blood cultures*
• order a chest x-ray
• prescribe a full chloroquin course
• start a broad spectrum antibiotic after the cultures have
been taken
• review the patient daily, look for “hidden” causes of fever
[see below]

“Hidden” causes of fever
1. The “hidden” abcess
• Suspect an abcess when the patient has been ill with fever for
more than one week
• Look for:
- a subphrenic abcess. Useful signs are:
- poor air entry over the lower lobe of the lung on the
affected side
- tenderness on palpation between the lower ribs
- a raised hemidiaphragm on the chest x-ray
- a pelvic abcess. Sometimes it takes a while before an
abcess in the pelvis becomes palpable: repeated recto­
vaginal examinations every 3-4 days are necessary!
- (multiple) abcesses in the abdomen. Abcesses between
the bowel loops are often difficult to find. Continuing
abdominal distension is often the only sign apart from
fever.
2.

Septicaemia
In the beginning (spiking) fever and chills are the only sign.
Signs of endometritis or infection elsewhere in the pelvis are
usually minimal or absent.
After a while signs of “metastatic” infection may appear, for
example: pneumonia or lung abcess; liver involvement
(jaundice!).
Often after days or weeks an abcess develops at the site of the
infected vein in the broad ligament.

* The value of the high vaginal swab, although it is often done, is very limited. It is
difficult to exclude contamination with material from the lower vagina and many of
the organisms causing puerperal sepsis do not grow in our laboratories (for example,
gonococci, chlamydia, B. fragilis)

162

3.

!

Generalised peritonitis and/or multiple abcesses in the
abdomen following caesarean section or ruptured uterus
This diagnosis is often made too late.
Suspect generalised peritonitis if, in the presence of fever:
- three or four days after the operation the
abdomen is still distended
- the abdomen remains tender:
- sometimes all over
- sometimes only in the flanks where
most of the pus collects
Additional signs can be:
- vomiting
- poor bowel sounds
- diarrhoea

The diagnosis is certain and laparotomy urgently needed if these
signs persist for more than a week after the original operation.

1
!

MANAGEMENT OF PUERPERAL SEPSIS
THE USE OF ANTIBIOTICS
General rules
• Use broad spectrum antibiotics (reason, you do not know
which organisms you are dealing with)
• Give high doses (reason: many of the bacteria are not very
sensitive to antibiotics)
• If the infection develops in a patient who had prophylactic
antibiotics with a caesarean section, do not use the same
antibiotics to treat the infection (reason: the infection is
probably due to a bacteria resistent to your prophylactic
antibiotics)
• If infection develops in a patient who is already on antibiotics,
change the antibiotics unless they were started only one or
two days ago.
• Do not in general change antibiotics too soon: they need at
least three or four days to do their work
Choice of antibiotic
Patient not very ill. Useful regimens are:
- ampicillin 500 mg per os or per injection six hourly
- x-penicillin 2 mega i.m. or i.v. six hourly + streptomycin 1 gm
i.m. daily
- bactrim 2 tablets bd
- tetracyclin 500 mg per os six hourly
All these regimens must be given for at least five days,
preferably seven

163

Note that:
- ampicillin or x-penicillin/streptomycin are not effective
against chlamydia
- none of these regimens is effective against B. fragilis
Patients severely ill. Often more than one micro-organism is
involved. B. fragilis is frequently one of the micro-organisms. B.
fragilis only responds to: chloramphenicol, metronidazole or
clindamycin.
Useful regimens are:
- ampicillin 500 mg by any route six hourly + metronidazole
400 mg eight hourly
- x-penicillin 2 mega i.v. six hourly + chloramphenicol 500 mg
i.v. six hourly
These regimens must be given for at least one week.
Follow-up
• Review the patient daily
• If she does not improve, consider:
- is the original diagnosis correct?
- is there an abcess anywhere?
- is she on the right antibiotic?
• Do not change the antibiotics until you have answered the
first two questions.

I

I

MANAGEMENT OF THE VARIOUS FORMS

OF PUERPERAL SEPSIS

Endometritis
After vaginal delivery
• mild : - ergometrine tablets I tds for five days
- simple analgesic
• severe: - same + antibiotics
After caesarean section
I

• antibiotics in all cases
Salpingo-oophoritis or parametritis
• antibiotics

• pain relief: simple analgesic, pethidine 50 - 100 mg six hourly
if necessary
164

Generalised peritonitis
• Treat on suspicion alone
• Antibiotics
• i.v. fluids
• Pethidine 50 - 100 mg i.m. six hourly
• Nasogastric tube
If no marked improvement after 24 hours, laparotomy. This is a
major procedure. Refer to gynaecologist or surgeon if possible. Do
not postpone.

Septicaemia
• Antibiotics
• Consider adding: heparin 5,000 - 10,000 units i.v. six hourly
for 10 days
• Refer to gynaecologist if possible
Abcesses
• Posterior colpotomy for abcesses in the pouch of Douglas*
• Laparotomy for abcesses elsewhere in the abdomen.
• Refer to gynaecologist if possible.

I

♦The effect of a posterior colpotomy on the patient’s feeling of well-being should be
dramatic and manifest within one or two days. If the temperature remains high and
S e is Still ill after two to three days, do a laparotomy to drain the remaining
abdominal abcesses.

I
I

165

I
I
i

Chapter 54

SECONDARY POSTPARTUM HAEMORRHAGE
DEFINITION

Bleeding from the genital tract in excess of normal lochial loss after
the first 24 hours postpartum and until the end of the puerperium.
CAUSES

- Retained products (membranes, placenta)
- Endometritis
- Sloughing of dead tissue (cervix, vagina, bladder, rectum)
following obstructed labour
- Breakdown of the uterine wound after caesarean section or
ruptured uterus
(

RISK

- Repeated attacks of (very) heavy bleeding
- maternal death, particularly if it happens far from hospital

MANAGEMENT
Admit into hospital as an emergency
Check the haemoglobin
x-match a pint of blood
Give ergometrine 0.5 mg i.m. or i.v.
Put up an i.v. drip
Add 10-40 units of pitocin per litre to the drip, if necessary
Start broad spectrum antibiotics in high doses by injection.
For example:
- x-penicillin 2 mega i.v. 6 hourly + streptomycin 1 gm i.m.
daily
or
- chloramphenicol 0.5 gm i.v. 6 hourly
• Do an examination under anaesthesia as soon as possible.
Do not postpone this\
• Treat according to your findings:
•
•
•
•
•
•
•

Retained products or endometritis
• finger curettage
• avoid the use of a curette if possible; if it has to be used, it
should be large and blunt.
166

Sloughing of dead tissues
• remove the slough
• pack the bleeding area tightly for 24 - 48 hours
• if packing does not not stop the bleeding, hysterectomy or
ligation of the internal iliac arteries may be necessary.
Breakdown of the uterine wound after caesarean section or
ruptured uterus
• do laparotomy
• resuture the wound or do hysterectomy

167

Chapter 55

COMPLICATIONS FOLLOWING OBSTRUCTED
LABOUR
The main complications following obstructed labour are:
- puerperal sepsis [see Chapter 53]
- secondary post partum haemorrhage [see Chapter 54]
- urinary or rectal fistulae
- nerve injuries

URINARY FISTULAE
CAUSES
The common cause is pressure necrosis of bladder and/or urethra
during delivery
Less common causes are:
- injury to bladder or urethra during:
- caesarean section
- the operation for ruptured uterus
- symphysiotomy
- injury to the ureter during:
- caesarean section
- the operation for ruptured uterus

DIAGNOSIS
The diagnosis is usually obvious: the patient is continuously wet.
Confusion is possible with:
- severe stress incontinence, in which the urine flows from
the urethra
- retention and overflow, in which the (over) full bladder is
easily demonstrated
MANAGEMENT
• Continue open bladder drainage for about three weeks to
promote healing of the vagina. This is easy if the fistula is
small. With a large fistula it can be difficult to keep the
catheter in the bladder. Try with a finger in the vagina to
guide the catheter into the bladder and inflate the balloon
enough to keep it there. If this does not work, remove the
catheter and allow the urine to drain through the fistula
• Treat infection with antibiotics
• Prescribe sitbaths once or twice a day (vaginal douches are
even better)

168

• Remove sloughs in theatre if necessary
• Discharge the patient home when all dead tissue has sloughed
and the vagina is clean
After four months and not earlier, refer the patient to a
gynaecologist for repair of the fistula.
Before referral check that:
- the patient’s general condition is good
- the haemoglobin is 11 gm% or more
- the vagina and vulva are clean
- urinary tract infection and bilharzia have been treated

RECTAL FISTULAE
CAUSES
The most common cause is pressure necrosis during labour. A less
common cause is the placing of sutures through the rectum by
mistake during the repair of an episiotomy or perineal tear

i

MANAGEMENT
• Treat infection with antibiotics
• Clean the vagina with sitbaths or vaginal douches
• Remove sloughs if necessary
• Refer to gynaecologist for repair about four months after
delivery
NERVE INJURIES
PRESENTATION
Common
:
Less common :

- dropfoot on one or both sides
- weakness of the hip muscles
- weakness of the quadriceps muscle

CAUSES
The most probable cause is pressure during labour

DIAGNOSIS
• Look for signs of nerve injury in every patient who has had
obstructed labour
• Check that she can move her toes, ankles, knees and hips in all
directions

Nerve injuries are very easily missed if the patient is completely
bed-ridden for a few weeks (due to puerperal sepsis for example). If
they are missed, contractures develop very quickly.
I
I

169

I

MANAGEMENT
Principles
• Await spontaneous healing of the nerves. This usually does
occur, but may take a long time (months)
• Prevent contractures from developing during the time the
muscles are paralysed
In Practice
• Keep the weight of bed sheets and blankets off the legs by
using a bed cradle
• Move the affected joints through their full range of movement
a few times each day (encourage a relative to help with this)
• For a dropfoot a backslab during the night can be useful
• Encourage the patient to walk

170

i

8

Obstructed Labour
R. H. PHILPOTT

Definition

Failure of descent of the fetus in the birth canal for mechanical reasons in
spite of good uterine contractions.
Incidence

This will be related to the incidence of cephalopelvic disproportion present
in the community and to the availability and quality of antenatal and
intrapartum care. Unfortunately disproportion is most prevalent in areas
where health services are deficient as poverty is an underlying cause of both
these factors. Where obstetric care is optimal, obstructed labour should
never occur, even in communities where disproportion is prevalent.
Because of the great improvement in obstetric care in developed countries
Very little has been written about obstructed labour in recent years and most
what follows represents the unpublished experience of a team working in
a.n area of transition.
CAUSES

Cephalopelvic disproportion (CPD). This may be due to faults in the pelvis
or to faults in the fetus, or a combination of both.
Faults in the pelvis: small or abnormal in shape.
Faults in the fetus:
a large baby;
a large defect, e.g., hydrocephaly, ascites or tumours of the fetus;
fetal monsters;
locked twins.

I

Malpresentations or malpositions of the fetus.
Breech presentation:
impacted, large breech;
extended arms;
arrested aftercoming head due to either hydrocephaly, undilated cervix,
deflexion of the head or CPD.
< linics in Obstetrics-iind (iynaecology—Vol. 9. No. 3. December 1982
U3U(>-3356/82/09( 13-625 SO3.IIO0 1982 W. B. Saunders Company Ltd

625

<•
R. H. PHILPOTT j

626

Transverse lie
Brow presentation
Mentoposterior position
Occipitoposterior position.

Impacted shoulders following delivery of the head.
Tight perineum, particularly in the primigravida.

Abnormalities of the vagina: transverse or longitudinal congenital septa;
scarring from the use of caustic traditional medications.
Abnormalities of the uterus: fibroids, congenital malformation.
Ovarian tumours.

COMPLICATIONS

These differ in the primigravid and the multigravid patient. The advanced
complications are seen only when the patient has not received proper
obstetric care.
The primigravid patient

Prior to obstruction, there is delay in the rate of dilatation of the cervix in the
active phase of the first stage of labour to a rate less than 1 cm per hour. If
oxytocin augmentation is used and theCPD is minimal, the rate of dilatation
may increase, but if the CPD is gross the rate of dilatation may increase or it
may not increase at all. Obstruction usually occurs before full cervical
dilatation. If this is neglected, the sequence of events is as follows.
There will be excessive pressure on the placenta from prolonged uterine
activity leading to fetal asphyxia, and compression of the head leading to
cerebral birth trauma. The effects of the head compression are compounded
when there is fetal asphyxia as the latter may cause a clotting defect.
Avascular pressure necrosis will develop in a ring in the pelvis at the level of :the obstruction. If this is severe and prolonged, the cervix, and even the J
lower segment, will slough and eventually a dead, compressed and moulded
baby will be delivered. This might take two or three or more days. When the :
ring of pressure necrosis sloughs further this may leave a large vesicovaginal
fistula involving the proximal half of the urethra and the bladder neck up to
the ureteric orifices. About 15 percent of those who develop a vesicovaginal
fistula also develop a rectovaginal fistula at the level of the pressure ring.
Healing of the ring of necrosis in the vagina leads to considerable contrac­
ture and often almost complete stenosis of the vagina.
Other complications include septicaemic shock, peritonitis with abscess
formation and an atonic postpartum haemorrhage. A few days after the
delivery some patients develop foot drop from sciatic nerve involvement.
The mechanism of this is uncertain, though it is thought to be due to

J

1

OBSTRUCTED LABOUR

627

compression of the sciatic nerve root or possible protrusion of a lumbar disc
during prolonged bearing down. In some parts of the world untrained
assistants apply excessive pressure to the abdomen, even to the extent of
causing rupture of the liver.
It is difficult to comprehend the degree of psychological trauma that these
patients suffer as a result of this terrible experience. Most are young teenasers and whether the pregnancy has been planned or not. it must leave an
indelible emotional scar. Even if the resultant fistulae can be repaired, and
i his is only possible in about 80 per cent of cases, many will be permanently
infertile and some will have so much scarring that further coitus is impossible.
This will mean that, for the majority, there will be no chance of marriage,
and in some cultures this will leave her an outcast and a recluse. For the few
whose injuries are not so severe, and another pregnancy is possible., %
incipient tears in the uterine wall may lead to rupture in subsequent labours.

I'he multigravid patient

Prior to obstruction, progress in the rate of dilatation of the cervix may be
normal and the only evidence of the impending obstruction may be the
failure of the presenting part to descend in the birth canal. However, in
many multigravid patients the abnormality in the cervical dilatation pattern
is similar to that seen in the primigravid in that initial progress is good and
then, during the active phase of the first stage, there is secondary arrest of
dilatation. If left untreated this does eventually progress to full cervical
dilatation.
After full cervical dilatation, if the obstruction is not relieved, the normal
pattern of uterine action continues to excess. Retraction and thinning of the
lower segment continues so that the junction ring between lower and upper
segment rises progressively, often to the level of the umbilicus. It is then
called the pathological retraction ring or Bandl's ring. The thin lower
segment becomes ballooned out and the myometrium is oedematous and
bruised. The bladder is also drawn up into the abdomen and with continuing
pressure of the fetus against the pubis, the bladder wall becomes
oedematous. The next event in neglected obstructed labour in the multi5 ravid is rupture of the uterus. The rent starts in the thin lower segment and
often extends laterally on one side down into the vagina and upwards into
the fundus of the uterus. The fetal presenting part may remain jammed in
the pelvis and so preventing severe haemorrhage from the torn uterine
vessels, or the fetus may be expelled into the peritoneal cavity with severe
haemorrhage both intraperitoneally and per vaginam. Sometimes the
bladder will rupture also, particularly if it is adherent to the lower uterine
''Cgment following a previous caesarean section. Rupture of the uterus in
• 'bstructed labour carries with it a maternal mortality rate that varies from 10
♦o 25 per cent depending on the duration of the rupture and the facilities
available for resuscitation and surgery.
From this description it will be seen that rupture of the uterus is extremely
uncommon in obstructed labour in the primigravid patient and fistulae from
pressure necrosis are seldom seen in the multigravid. Because obstructed

f

f

R. H. PHILPOTT

628

labour in the primigravid is virtually always preceded by delay in the rate of
cervical dilatation, this latter feature acts as a reliable early warning signal.
In the multigravid obstruction may occur without a preceding change in the
normal pattern of cervical dilatation, and in those circumstances the problem
is more difficult to detect and is then only recognized by the slow or arrested
rate of descent of the presenting part, and increasing moulding if it is a
cephalic presentation.

CLINICAL PRESENTATION

The primigravid
The patient is usually a young teenager who may even have become pregnant
before her first menstrual period. The majority will have received no ante­
natal care, either because they were hiding an unplanned pregnancy or
because of inadequate provision of obstetric services. There will be a history
of a verv prolonged labour at home, often extending to days rather than
hours.
On general examination the patient will be tired, exhausted and anxious.
She will be dehydrated and acidotic, pyrexial with a tachycardia but not
shocked (unless there is septicaemia). On examination of the abdomen
there is often evidence of distended, atonic bowel due to hypokalaemia.
Uterine contractions will either be hypotonic or they will be strong and
painful with poor relaxation between contractions. In advanced cases of
obstruction the uterus may be distended with gas and be tympanitic to
percussion. The cause of the obstruction may be apparent, for example a
high fetal head overlapping the pelvic brim or a transverse lie. The liquor
will have drained away and the uterus will be moulded around the fetus. The
fetal heart will give evidence of fetal distress with a tachycardia and
decelerations that start early in the contraction and are delayed in their
return to the baseline until well after the contraction is over. In more
advanced cases the fetus will be dead. The bladder is often distended with
retained urine, or thick and oedematous from compression and containing
surprisingly little or no urine. Catheterization may be very difficult and it
may be necessary gently to dislodge the presenting part with two fingers in
the vagina while passing the catheter. Because the bladder is drawn up, the
catheter may have to be passed a long distance to reach the bladder. The
urine is often blood stained.
On vaginal examination, if there has been prolonged and even premature
bearing down, vulva and cervix may be very oedematous. The vagina is dry
and the cervix not yet fully dilated unless it has already sloughed. The liquor
is meconium stained and often foul-smelling from intrauterine infection.
There will be evidence of the obstruction, for example a high head with
excessive moulding and caput.
The multigravid
The clinical picture varies according to whether the patient is seen prior to or
after uterine rupture.

I

629

OBSTRUCTED LABOUR

Prior to uterine rupture. The genera! findings will be the same as for the
primigravid. On examining the abdomen there is often the ‘three tumour
abdomen’ with an oedematous, enlarged bladder, a distended, tender lower
uterine segment, a Bandl’s ring near the level of the umbilicus and a tonically
contracted upper uterine segment above the ring. The round ligaments can
be palpated and feel like attenuated guy ropes on either side of the ballooned
lower segment. The bowel is distended. The fetus may be difficult to palpate
because the liquor has drained away and the uterus is tonically contracted,
but it should be possible to detect the cause and evidence of the obstruction.
On vaginal examination, the cervix is usually fully dilated and the presenting
part high.
After uterine rupture. Here again there are two different clinical pictures. In
the one, though rupture has occurred, the presenting part is so impacted that
it has produced a tamponade effect on the torn uterine vessels and there is
therefore no haemorrhage or hypovolaemic shock. This clinical picture is
Similar to that prior to uterine rupture. The other picture is more dramatic
With all the features of profound shock, often both hypovolaemic and
septicaemic.
The patient with an evident rupture of the uterus will often give a strange
story of feeling that the baby was moving up rather than down with each
contraction, a sensation that they recognize prior to the rupture. After the
rupture, the contractions may no longer be felt but they are replaced by
continuous very severe abdominal pain with the patient hardly able to move.
There may be a history of vaginal bleeding. She will appear extremely
anxious and distressed with dehydration and shock. The pulse and blood
pressure may be absent and the central venous pressure low. Sometimes the
patient is moribund or dead on admission to hospital.
The abdomen will be grossly distended and the uterus difficult to feel.
Though the fetal parts may be easily palpable outside the uterus often the
degree of peritonitis and tenderness is such that palpation is extremely
difficult. It may be possible to elicit shifting dullness but, again, the degree of
pain and tenderness is so great that it is difficult to move the patient.
Abdominal aspiration with a syringe and needle in the flank may demonstrate
blood in the peritoneal cavity, but the diagnosis is usually so obvious that this
is not necessary. The fetus will be dead.
On vaginal examination the presenting part may be jammed in the pelvis,
or it may have receded above the pelvic brim. Catheterization may produce
heavily blood-stained urine, particularly if the bladder is also ruptured.

I

MANAGEMENT

Prevention
Obstructed labour should never occur if patients receive optimal antenatal
intrapartum care and the requirements of such care are not sophisticated,
elaborate or expensive. The type of care to be provided will need to be

[
I
I

F
R. H. PHILPOTT

630

tailored to the immediate resources of the community. Where feasible, the '
ideal is hospital care for all, but in the very communities where obstructed
labour is prevalent this is not possible at present. The next best alternative is
a system of well-supervised, midwife-run satellite clinics within easy reach of
all women, and groups of these clinics related to a base hospital. This
relationship must include detailed protocols of management with welldefined indications for transfer, ready communication with the base hospital
and adequate transport for referrals. Even these provisions are outside the
minimum facilities available to millions of women in the developing world,
but they must be strived for. as without them the casualties of obstructed
labour will continue to be high. In many parts of the world the traditional
birth attendant cares for women in their homes, and has the respect of the
whole community. These attendants must also be provided with basic train­
ing in antenatal care and intrapartum management, and be recognized as an
integral part of the maternity services of the country.
Where distances are great and the population is scattered it may not be
possible to ensure that all patients can get to a clinic or a hospital in time for a
supervised labour. In these circumstances the provision of a simple ‘Mothers
Waiting Area (MWA)' near to each clinic and hospital will ensure that
patients receive complete intrapartum supervision from early in labour. The
MWA should consist of simple housing similar to that in the patient's village,
and the patient, who arrives at the MWA in late pregnancy, is cared for by a
relative. At a hospital in eastern Nigeria. Ross (1970. personal communica­
tion), reported a perinatal mortality of 70 and maternal mortality of 7 among
109 emergency admissions for severe CPD whereas in 134 similar patients
admitted to their MWA the figures were 6 and zero respectively for 134 CPD
admissions.
In the antenatal care of the individual patient the past obstetric history of
difficult deliveries in the multigravid must alert the midwife to the high risk category of the present pregnancy. Patients short in stature need to be>
watched carefully, and in peripheral clinics midwives need to be trained to"?,
do pelvic assessments at 36 weeks of pregnancy. During labour, the useof^
simple partograms with guidelines to help in the recognition of slow dilata-^tion of the cervix are of great value in predicting and detecting mechanical^
problems in labour. Each labour needs to be conducted with the sarne^
meticulous care and accuracy of clinical observations as in a trial of labour^
and for that matter each patient should be regarded as having potential CPD
until delivery is complete.
<
Because many of these patients are very young primigravidas, often only
12 to 14 years of age, more attention needs to be given to home and^
community education in an endeavour to avoid these very early teenageg
pregnancies.
g
Resuscitation

If the patient is first seen at a peripheral clinic, immediate transfer needs tobe arranged by the quickest means available. Depending on the facilities
available, and it is recognized that these may be severely limited.
’■CT

‘HSTRUCTED LABOUR

631

patient’s prognosis can be improved by giving first-aid treatment before and
during transfer. If it is possible to communicate with the referral hospital to
warn that the patient is on her way, this helps to prepare for her arrival.
Accurate information, preferably on a labour graph if the patient has been
labouring in the clinic, should accompany the patient. Many clinics are able
to set up an intravenous infusion of an electrolyte solution and give a
parenteral broad-spectrum antibiotic before the patient leaves the clinic.
On arrival in hospital, resuscitation needs to be fairly rapid as operative
delivery is usually very urgent. If an expert anaesthetist is available to help in
the resuscitation this is a great advantage. Again it is recognized that the
quality of resuscitation depends not only on the duration of the obstruction,
but also on the facilities and resources available. Where these are limited it is
probably best to concentrate them in referral hospitals and leave the
peripheral clinics to cater for the normal deliveries. If, as is usual, urgent
operative delivery is indicated, it is a good idea to admit the patient straight
to the operating theatre area and carry out the resuscitative measures in the
tkftatre. This avoids moving the patient again and permits operation just as
Soon as the patient is in optimal condition.
The first step is to set up an intravenous infusion and to correct dehydra­
tion. electrolyte deficit and acidosis. If it is possible to measure the serum
electrolytes and blood gases this is valuable, but usually delivery needs to be
attended to before these results become available. For this reason it is
advisable to give a Plasmolyte solution that contains sodium bicarbonate or
to give sodium bicarbonate separately.
If there is evidence of shock, this may be either hypovolaemic or septicaemicshock or both together. It is then mandatory to set up a simple central
venous pressure (CVP) manometer and regulate further fluid replacement
according to the CVP measurements. It is best to maintain a CVP reading of
between 5 and 8 cm of water. If the CVP is within this range or higher and the
blood pressure is still low, it is presumed that part of the problem is
stpticaemicshock. The blood pressure, pulse, CVP and urinary output must
be recorded frequently throughout management.
If there is hypovolaemic shock this should be corrected by blood trans­
fusion as soon as this is available, but in the meantime electrolyte solutions
be used. Fresh whole blood is ideal but packed red cells plus fresh dried
plasma may be more readily available. If there is a concomitant anaemia this
should be attended to at the same time. If there is septicaemic shock, use
steroids and broad-spectrum antibiotics in the first instance. Different
hospitals favour different antibiotic regimens dependent on the type and
Sensitivity of organisms prevalent locally. We use parenteral Ampicillin.
Konamycin and Metronidazole. If this does not correct the hypovolaemic
shock and the blood pressure is still low with poor urinary output and
constricted peripheral vessels, the best treatment is a titrated infusion of
<Aopaniinc. This will result in peripheral vasodilatation and a drop in the
CVP which must be corrected immediately by more intravenous fluids. In
of obstructed labour, infection must be assumed and the possibility
of septicaemic shock predicted and therefore the same regimen of anti­
biotics as outlined above should be commenced prophylactically.

R. H. PHILPOTT

632

With the urgency of resuscitation and operative treatment we often fail to
communicate with the very distressed patient. Recognizing the magnitude |
of suffering they have already experienced (and their probable fear of
institutional treatment), it is important to give some time to allaying their
fears and giving them reassurance.

Anaesthesia
Where an anaesthetist who is expert in handling this type of clinical problem <
is available, the decision on the best method of anaesthesia can be left in his
hands. Unfortunately, as with the skills of resuscitation, it is in those parts of
the world where obstructed labours are most common that expertise m
anaesthesia is least available. The expert anaesthetist will probably choose a
general anaesthetic with endotracheal intubation as the method of choice for
abdominal operations. If that expertise is not available, then the safest
technique for the single-handed surgeon is local infiltration with a local
anaesthetic. If there is no shock present or expected, the single-handed
surgeon will find it easier to operate with (and this is more comfortable for
the patient) a spinal ‘saddle-block' or an epidural block. These can be
supplemented with local infiltration to the upper part of the abdominal
incision. Each of these techniques is best learnt by observing an expert.
.
If the delivery is to be by the vaginal route then a pudendal block,
saddle-block or epidural block will be most suitable.
Of very great importance is the avoidance of inhalation of acid gastric
contents before, during or after the anaesthetic. If this occurs, as it well
might if strict care is not taken, it can prove lethal. A stomach tube should be
passed to the empty stomach. The counsel of perfection is to give a non­
particulate oral antacid within an hour of surgery, followed by Cimetidine,
Metaclopramide and Glycopyrrolate. This combination of drugs reduces the
acidity of gastric juice, empties the stomach downwards and reduces theproduction of gastric acid. These drugs are expensive and may not be readily
available, though the high maternal mortality rate from acid gastric
aspiration merits their use. If they are not available Mist. Magnesium^
Trisilicate orally and Atropine intramuscularly should be used.
It is important to continue frequent (every five minute) checks of pulse,i
blood pressure, CVP and urinary output measurements throughout thy
operation and then to provide intensive care observations in thej
postoperative period.
Operative Delivery
The mother’s life will often depend on the obstetrician making the corrects
choice of operative procedure. The various procedures will be discusseafirst. and then the management of the different causes of obstructed labour^
will be considered.
Episiotomy
V
In primigravid patients who have received no intrapartum care, somenme^-

i

OBSTRUCTED LABOUR

633

all that is needed is a wide episiotomy. These patients are often found to
have an occipitoposterior position of the vertex. Putting the patient in the
lithotomy position may further facilitate delivery.

Ventouse and forceps delivery
I n those patients in whom there is definite CPD the use of the forceps or the
ventouse is absolutely contraindicated. This is the case, in vertex presenkirions. when there is marked moulding and more than one-fifth of the head
palpable above the brim of the pelvis. On the other hand, labour may be
obstructed due to an occipitotransverse or occipitoposterior position with­
out CPD with the head no more than one-fifth above the pelvic brim. Then
an assisted vaginal delivery is indicated.
It must be emphasized that there should never be a difficult operative
delivery, and where there is already fetal asphyxia the assisted delivery
should be no more than a ‘lift out’. The accurate assessment of head level by
^abdominal palpation and the degree of moulding as assessed by vaginal
examination are crucial in making these decisions.
If the head is no more than one-fifth above the brim and the fetus is alive,
the decision has to be made whether to use the ventouse or the forceps.
Personal preference and experience with the instrument are determining
factors. When there is not much moulding or caput, either instrument is
Suitable, and possibly the ventouse is preferable as it is easier to use. When
the moulding and especially the caput are more evident, the fit of the head in
the pelvis is that much tighter and more traction will be required. It is then
■H>a.tihe forceps are to be preferred. If the head is not in the occipitoanterior
pettibon. a Kielland's rotation is to be preferred to a manual rotation, as the
latter might displace the head upwards.
If the ventouse is chosen, then the ‘rule of the three pulls’ must be adhered
to. The first pull must dislodge the head from its arrested position, the
Second pul! must bring the head to the pelvic floor and the third pull deliver
or at least crown the head. If any one of these three pulls does not achieve its
purpose then further traction must be desisted and an alternative method of
delivery carried out. If the baby is alive this will be either a symphysiotomy
or a caesarean section. If the baby is dead, a craniotomy. It is always best to
kave predicted the difficult delivery and to have chosen the symphysiotomy
©rcaesarean section in the first instance. If there is fetal asphyxia present this
is doubly imperative.

Symphysiotomy

This is an operation that needs to be reinstated and given its place in the
Qinge of operative delivery procedures. It has fallen into a degree of dis­
repute mainly because there was a time when it was used to overcome gross
disproportion and this led to major complications. It is not used at all in parts
t-He world where CPD is almost non-existent. This is a problem for it is in
these countries that the textbooks are written and the trends are set. In
countries where CPD is prevalent, symphysiotomy is an excellent procedure

T
634

R. H. PHILPOTT

in labours where the degree of CPD is borderline. Caesarean section must
still be employed where the CPD is marked. The skill is to recognize the
difference.
This operation has its best place at a strategic moment in a well-planned
trial of labour, when borderline CPD is evident and before fetal hypoxaemia
occurs. In well-chosen cases it can be shown to be preferable to a caesarean
section and avoids the difficult or traumatic operative vaginal delivery. In
neglected, obstructed labour it also has a place, for it will avoid the compli­
cations of a major abdominal operation in a high-risk patient. However, in
such patients the exact place of symphysiotomy needs to be carefully
defined.
It is only indicated when the baby is alive, in a cephalic presentation with
the head not less than one-fifth and not more than two-fifths above the brim,
with moulding of the fetal head. This represents borderline disproportion.
When none of the head is felt above the brim, the venlouse or forceps can be
used and when more than two-fifths of the head are felt above the brim the
complications of symphysiotomy become unacceptable and caesarean
section is to be preferred. The fetus should not be under 2.5 kg or over 4 kg in
weight.
If the fetus is under 2.5 kg, the degree of pelvic contracture must be so
great that it is almost certain that, with a normal-sized baby in the next
pregnancy, a caesarean section will be required. This outweighs one of the
major benefits of symphysiotomy, that it provides permanent enlargement
of the pelvis and allows for normal vaginal deliveries in subsequent preg­
nancies. If the fetus is over 4 kg in weight the amount of symphyseal
distraction may be so great as to produce an unstable pelvis and a strong
likelihood of urinary stress incontinence.
There must be no evidence of rupture or impending rupture of the uterus
and though the operation is most commonly performed in the primigravid
patient, it can be done in a multigravid. It is usually done at full cervical dilatation but is sometimes done at 8 or 9cm dilatation when fetal distress
makes delivery urgent. Some obstetricians do symphysiotomy much earlier
in labour to reduce the amount of head compression in late labour, but this
practice does lead to a number of unnecessary symphysiotomies.
The operation is also indicated for unexpected CPD with the aftercoming
head in breech presentation with a live baby. In such cases the decision must g
be made very quickly and the operation performed immediately, otherwise
the baby will die of asphyxia or intracranial birth trauma.
g
-L-.

s

Caesarean section
“
The patient in obstructed labour is a very serious operative risk and the
operation should not be embarked upon lightly. The surgeon should be
experienced in this kind of problem as the price of inexperience can be death
for the patient. Though the operation needs to be proceeded with urgently,
adequate time must be spent in thorough resuscitation of the patient as
considerable and require
outlined earlier. The anaesthetic hazards are also
-------experience if these are to be avoided.

)BSTRUCTED LABOUR

635

Dependent on the patient's circumstances and the facilities available,
some surgeons will be liberal with caesarean section even when the baby is
dead; others will try to avoid caesarean section whenever possible. If it is not
possible to ensure that the patient will have good antenatal and intrapartum
care in a subsequent pregnancy, and if the facilities for a safe caesarean
section in this admission are not good then a section will be avoided if there is
any alternative. The decision is a critical one (Lawson, 1972).
While preparing for the caesarean section continuing uterine activity can
be reduced by giving tocolytic /^-stimulant drugs by intravenous infusion.
The anaesthetist needs to be aware that these can lower maternal blood
pressure.
Caesarean section is indicated in obstructed labour when the fetus is alive
and symphysiotomy is not suitable; if the cervix is not yet fully dilated and
delivery is urgent for maternal reasons, whether the baby is alive or dead; if
there is evidence of imminent uterine rupture, even if the fetus is dead. In
the latter circumstance it is feasible to open the abdomen and if the lower
segment is found to be intact, to proceed to a destructive operation from
Y>eJow. This saves the extreme hazard of a destructive operation with a
uterus that might rupture, and at the same time avoids a caesarean section
for a dead baby in the presence of intrauterine infection.

The alternative types of caesarean section for obstructed labour

IzMtTajieritoHeal caesarean section. This is indicated when there is established
or potential intrauterine infection in cephalic presentations, and has been
means of reducing the incidence of severe postoperative peritonitis in
such patients. The technique used is that described by Crichton (1973). The
cMominal wall can be opened longitudinally or transversely, but the
parietal peritoneum is opened transversely. When the visceral peritoneum
has been opened and the bladder reflected downwards, the upper flap of the
visceral peritoneum is sutured to the upper flap of the parietal peritoneum
and the lower flap of the visceral to the lower flap of the parietal peritoneum.
This artificially excludes the lower segment from the peritoneal cavity by
rendering it extraperitoneal. After delivering the baby and placenta and
suturing the lower segment, the peritoneum over the uterine wound is left as
wos and a drain placed over the wound.

De Lee incision. This is a low longitudinal uterine incision. It is indicated
when the lower segment is thin and distended and there is danger of lateral
tears when extracting the fetus. It is done in cephalic presentations, but very
particularly in transverse lie with a prolapsed arm and a live baby. It is also
tha incision of choice when there is an uncorrectable (with, for example,
amyl nitrite) constriction ring. This incision is made by opening the visceral
peritoneum transversely and reflecting the upper flap as high as possible and
Hne lower flap as low as necessary. To gain adequate access, though twoH-'irds of the incision will be in the lower segment, one-third will extend into
the upper uterine segment.
Lew er segment, transverse incision. Normally when performing a caesarean

636

R. H. PHILPOTT

I

section this is the incision of choice, and it will be so in most cases of
obstructed labour. However, it is dangerous and should not be done if there
is intrauterine infection (then an extra peritoneal incision is indicated) or
lateral extension tears are likely to result (then a de Lee incision is indi­
cated). The dangers of lateral tears cannot be overemphasized. Apart from
compromising the blood supply to bruised and infected myometrium, the
attempts to secure haemostasis can endanger the ureter if great care is not
taken. In obstructed labour the distorted presenting part may be difficult to 4
extract through a lower segment incision. Not only will a tear be dangerous,
but the alternative of doing an inverted-T incision to deliver the baby is also
hazardous, for it does not heal well. For all these reasons it is critical to make .
a careful assessment before making the incision in the uterus, and only doing
a standard transverse incision if this is safe. Because the final decision is not
made until the abdomen is opened and the uterus is inspected, it is best in "
these circumstances to open the abdomen through a midline subumbilical
longitudinal incision rather than the more usual transverse incision. The
peritoneal cavity should be entered high up in the longitudinal incision to
avoid injury to the bladder which is often much higher than expected. When
deflecting the bladder downwards, after opening the vesical peritoneum,
great care must be taken as it is often oedematous and fragile.
If it is decided to do a transverse lower segment uterine incision, there are
important steps that must be taken to avoid lateral tears if the head is
jammed in the pelvis. Once the abdomen is open and the visceral perito­
neum incised, an assistant should slowly and carefully disimpact the fetal
head per vaginam. before the uterus is opened. If this is delayed until after
incising the uterus, the fetal shoulder will prolapse into the wound making
further delivery very difficult. The uterus should be opened with scissors to
ensure the exact directions of the opening. The surgeon should then guide
the delivery of the fetal head with a hand outside the lower flap of the lower
segment. Once the head has been disimpacted to the level of the incision it is g
preferable to deliver the head with Wrigley’s forceps with the sagittal suture
in line with the uterine incision and a degree of anterior asynclitism. If, when
the forceps arc applied, it is recognized that the incision will not be long
enough to deliver the head w'ithout a lateral tear, the lateral ends of the^j
incision should be cut with scissors upwards so that the whole incision now j
forms a broad U-flap.
. J
If an extensive lateral tear does occur, and there is much haemorrhage, it
is wise to open the broad ligament by cutting the round ligament and so
making it easy to palpate the ureter before applying any clamps. The extent f
of the tear can be found by applying Green-Armytage forceps to the edges of
the wound and drawing the angle into clearer vision. Bleeding can always be i
controlled in the first instance by direct pressure w ith a dry pack and then the ..
vessels can be located and tied. Sometimes complete control of the haemor­
rhage is Obtained only by tying off the internal iliac artery on that side.
Caesarean hysterectomy
When the lower segment is severely bruised, major uterine vessels have

J

637

)BSTRL'CTF.D I ABOUR

been torn and there is potential or established intrauterine infection, and the
patient is not planning further pregnancies, caesarean hysterectomy is the
procedure of choice.

Destructive operations
These are unpleasant to perform but are the procedures of choice in ob­
structed labour when the fetus is dead and the cervix is fully dilated. They
arc an alternative to an unnecessary and hazardous caesarean section or a
vaginal delivery that would otherwise cause severe maternal trauma. They
are contraindicated where the uterus is ruptured or where rupture is immi­
nent. Thus, if there has been vaginal bleeding or the lower segment is
ballooned and tender, a laparotomy is essential. If this confirms dangerous
thinning of the lower segment, caesarean section is safest, though hazardous
in itself. If there has not been excessive thinning and bruising of the lower
segment, the vaginal destructive operation may be proceeded with and the
obdomen closed without opening the uterus. It is always imperative to
explore the uterine cavity, the cervix and the vagina for rupture or lacerti^ns after performing any destructive operation. If there has been pro­
longed pressure of the presenting part on pelvic structures, there is danger of
fistula formation and a self-retaining catheter should be left in the bladder to
keep the bladder empty and at rest for 14 days. Because the uterus is liable to
be atonic in the third stage of labour, a postpartum haemorrhage must be
avoided by giving ergometrine intravenously with the delivery of the baby
and then immediately to commence an infusion of 11 of Dextrose in saline
containing 40 u of oxytocin.
For a description of the indications and method of performing the various
destructive operations see the chapter ‘Delivery of the Dead or Malformed
Fetus’ by Lawson, in this issue.

Rupture of the Uterus
Rupture of the unscarred uterus is the end result of obstructed labour in the
multigravid patient. The aetiology, diagnosis and principles of preliminary
management are referred to under that heading.
Many authors (Lawson, 1967; Groen, 1974) advise the simplest and
quickest operation possible in these seriously ill patients. They recommend
an abdominal repair of the rupture in most instances. For many years now
we have found that, in most cases, total abdominal hysterectomy gives much
better results (Mokgokong and Marivate. 1976; Skelly. Duthic and Philpott,
1976). Usually, there is severe infection present from the prolonged rupture
of the membranes, the uterme muscle is bruised and oedematous and the
ntphire has torn one or both uterine arteries, leaving the muscle without a
good blood supply. When this type of rupture is repaired, the healing is poor
ohd postoperative infection is common. If a repair is done, it is almost always
Accessary to do a tubal ligation as subsequent pregnancy would be a great
nsk. For these reasons we have preferred total hysterectomy which can be
^one about as quickly as a repair. Only when the tear mimics a transverse

....I
•t:

R.H. PHILPOTT I

638

lower segment caesarean section incision and the patient is anxious for more
children do we consider a repair. If the surgeon is not experienced at *
hysterectomy, repair may for him be the better option.
4
The basic steps of the operation are similar to a standard abdominal *
hysterectomy, but a few points are worthy of emphasis. When the fetus and
placenta have been removed, the first step is to control the haemorrhage.
The direct pressure of a dry pack will stop the bleeding in the first instance.
Green-Armytage forceps are applied to the edges of the rupture until the
lowermost angle is found. The bladder is reflected downwards and once the A’
broad ligament is opened, the ureter is felt all the way to its entrance into the
bladder. Only then can the pack be slowly removed and bleeding vessels \
clamped. With the uterus open it is easy to see the level of the cervix and a
total hysterectomy is nearly always possible. There is often an extensive
haematoma tracking up from the torn uterine vessels on one side, between
the layers of the broad ligaments, and often upwards towards the kidney.
This must be evacuated and torn vessels ligated.
We used to pack the raw pelvic floor extraperitoneally at the end of the
operation, but in recent years have found that this is seldom necessary. We
leave an abdominal drain in that area to give early warning of secondary
haemorrhage. Continuous bladder drainage is necessary if there has been
prolonged pressure on the bladder or if there is haematuria at the end of the
operation.
~

MANAGEMENT OF INDIVIDUAL CAUSES OF OBSTRUCTED
LABOUR

This is presented in abbreviated note form to clarify the best method of
management in the various conditions causing obstructed labour.
Cephalic presentations
Vertex. If the cervix is not fully dilated, caesarean section should be performed. If the cervix is fully dilated and rupture imminent, caesarean section J
should be performed, but if rupture is not imminent then management^
depends on the level of the head, the degree of moulding and evidence of^
fetal distress:
0/5 above and minimal moulding — episiotomy and ventouse or forceps, ®
1/5 above and marked moulding - forceps;
1/5 above with marked moulding and fetal distress - symphysiotomy (org
caesarean section);
1/5, 2/5 or 3/5 above with dead baby - craniotomy:
i
4/5 above with dead baby - may need caesarean section;
i?
2/5 above with moulding and'baby alive - symphysiotomy (or caesarean ;
section);
3/5 above with moulding and live baby - caesarean section.

4

Brow. Caesarean section should be carried out.

4

i

OBSTRUCTED LABOUR

639

Mentoposterior. If the cervix is fully dilated, the baby alive, no CPD, no
imminent rupture - Kielland's forceps rotation and delivery. If CPD caesarean section. If baby is dead and cervix fully dilated - craniotomy.

Aftercoming head of the breech
I f the fetus is dead, craniotomy should be performed by perforating the head
through the occiput. If the fetus is alive, management must be immediate
and depends on the cause. If it is a hydrocephalic head, perforate and drain
the. CSF. If it is an undilated cervix trapping the head, lift up the body by the
feet and cut the cervix right-posteriorly (at eight o'clock) and left-posteriorly
(at four o'clock), and then deliver the head with Wrigley's forceps. If the
head is deflexed deliver by the Mauriceau-Smellie-Veit manoeuvre. If this
fails, there is CPD and if the fetus is still alive the only management is an
immediate symphysiotomy.

Transverse lie
If the baby is alive, caesarean section should be performed. If the mem­
branes are intact or recently ruptured this should be a lower segment
caesarean section, but if the arm is prolapsed it should be a de Lee incision.
If the baby is dead and the cervix fully dilated with no imminent rupture,
do a decapitation. If rupture is imminent or the cervix is not fully dilated, do
a.caesarean section. This should be a lower segment, transverse incision. If
necessary the baby can be dismembered to remove it through the transverse
tndsion.
There is no place for internal version and breech extraction in obstructed
labour with a transverse lie. Rupture of the uterus is almost inevitable it this
rule is broken.

Congenital septum of the vagina
If this is the cause of an obstructed labour it is seldom feasible to incise the
Septum and deliver the baby vaginally. The septum is usually too thick and a
caesarean section is necessary. The septum can be excised in the non­
pregnant state.

Ovarian tumour or uterine fibroid

If a tumour is obstructing labour a caesarean sectior. will be necessary. If this
i$ an ovarian tumour, it can be removed at the completion of the caesarean
Section. If it is a uterine fibroid it is best to leave it alone ar.d deal with it as an
uxferval procedure.
Follow-up advice
U there is the possibility of a future pregnancy, it is important to ensure that.

r

3F

R. H. PHILPOTT

64()

*

if this obstructed labour has been due to neglected obstetric care, the „
opportunity is taken to plan for thorough care in the next pregnancy. This
entails contraceptive advice to help plan the timing of the next pregnancy
and good antenatal and intrapartum care in that pregnancy. It is probable
that an elective caesarean section will be needed next time.
REFERENCES
Crichton. D. (1973) A simple technique of extraperitoneal lower segment caesarean section.
South African Medical Journal, 47.2011-2012.
Groen. G. P. (1974) Uterine rupture in Nigeria. Obstetrics and Gynecology. 44.682-687.
Lawson. J. B. (1967) Rupture of the uterus. In Obstetrics and Gynaecology in the Tropics and
Developing Countries (Ed.) Lawson. J. B. & Stewart. D. B. p. 189. London. E. Arnold.
Lawson. J. B. (1972) The place of caesarean section in developing countries. Tropical Doctor,
2. 30-32.
Lawson. J. B. (1974) Embryotomy for obstructed labour. Tropical Doctor. 4. 188-191.
Lister. U. (I960) Obstructed labour. Journal of Obstetrics and Gynaecology of the British
Commonwealth, 67. 188-198.
Mokgokone. E. T. & Marivate. M. (1976) Treatment of the ruptured uterus. South African
Medical Journal. 50. 1621-1624.
Seedat. E. K. & Crichton. D. (1962) Symphysiotomy: technique, indications and limitations.
Lancet, i. 544-558.
Skelly. H. R.. Duthie. A. M. & Philpott. R. H. (1976) Rupture of the uterus. South African
Medical Journal, 50.505-509.

¥

i
!

-‘i

f

■■

Tropical Doctor3 April 1^4

SYMPHYSIOTOMY | 69

Obstetric Care
Symphysiotomy
Donald A. M. Gebbie, MD, FRCOG
Professor of Obstetrics and Gynaecology3
University of Nairobi, Kenya
tropical DOCTOR, 1974, 2, 69-75

“The main barriers against symphysiotomy remain
in the minds of obstetricians,” says A. D. H. Browne
(i968)^and this operation has had a “most chequered
history” (Moir 1964). Its performance dates back to
1777 when, on the night of September 30, Sigault
undertook it to ensure the safe delivery of Madame
Souchet, a rachitic Parisienne. There had been four
previous stillbirths delivered by destructive operations. That it is in fact older than this is suggested by
both Bowesman (i960) and Pereira (1964) who have
j
1
1
! J .
.
produced evidence of its being undertaken for many
years by tribal practitioners in unsophisticated areas
in Africa.
The purpose of symphysiotomy is to divide one
side of a disproportionately small and irregular
cylinder, the maternal pelvis, to permit the easier
passage of a global object, the fetal head, thereby
securing spontaneous or gently assisted delivery. If
this is undertaken skilfully and with proper indication.
there should be no harm to the fetus but a permanent
change will occur in the maternal skeletal system,
This alteration does not necessarily cause disablement
but may prove beneficial in future pregnancies. Done
badly, or with improper indication, the trauma which
can be inflicted upon the maternal hard and soft
tissues may be great and severe morbidity may follow,
It is the purpose of this article to describe the
indications for symphysiotomy, the conditions which
must be present before it is done, and the dangers and
difficulties which may be encountered with particular
reference to how they may be avoided.
Most symphysiotomies are undertaken for cephalopelvic disproportion with the vertex presenting and a
live fetus. Furthermore they are performed either
late in the first stage of labour or in the second stage.
As such it is an alternative to either Caesarean section
or to a difficult and often traumatic operative vaginal
delivery. To begin with, therefore, the operation will
be described in this context and reference will be
made later to less common circumstances in which
symphysiotomy may be useful.

INDICATIONS

When a contracted pelvis is diagnosed during the
antenatal period, the obstetric practitioner is fore­
warned of possible cephalopelvic disproportion. Pro­
vided there are no contraindications such as a previous
bad obstetric history, a previous Caesarean section, or
significant maternal disease, a trial of labour is the
rule. The progress of the subsequent labour will
determine whether the initial surmise was correct or
not. At any time during labour, the trial may have w
to
be terminated by artificial means for one or more of
several reasons. These will include fetal distress,
distress.
failure to progress, and prolonged labour. If there has
been no antenatal
supervision the woman may be
;
admitted as an emergency at any time during labour
and the assessment will of necessity have to be made
more rapidly. Along with such decision, the most
appropriate method of operation should be chosen,
“
The choice lies among Caesarean section, operative
vaginal delivery by forceps or vacuum extraction, and
symphysiotomy. Whichever method is decided upon
will depend upon three related factors, the station or
descent of the fetal head, the degree of overlap, and
the dilatation of the cervix.

Station or descent of fetal head
It is current practice to describe fetal head descent or
station as the relationship of the level of the presenting
part
part to
to the
the maternal
maternal ischial
ischial spines.
spines. In
In cephalopelvic
cephalopelvic
disproportion this is dangerous as caput formation
and fetal head moulding falsify the extent of fetal
head descent. It is better to use the classification
originally suggested by Crichton and recently illustrated by Simons and Philpott (1973) of dividing the
fetal head into fifths and describing descent as the
number of fifths of the fetal head which have passed
into the brim of the maternal pelvis. Better still for
present purposes is to divide the fetal head into three
equal portions or thirds and relate these to the brim
of the maternal pelvis.
Thus, less than one third, between one third and
two thirds, or over tw’o thirds of the fetal head may
be described as having entered the pelvic brim.
No symphysiotomy should be undertaken unless
more than one third of the fetal head has entered the
pelvic brim. If this has been the maximal descent of
the fetal head which has been possible either labour
has not progressed far enough or the cephalopelvic
disproportion is severe.

I

?

I

70 | SYMPHYSIOTOMY

Symphysiotomy may be indicated when one third or
more of the fetal head has entered the pelvic brim,
provided other circumstances are favourable.

Overlap
During descent of the fetal head the anterior parietal
eminence passes underneath the pubic symphysis. If
disproportion is present, fetal head moulding will be
necessary before this can happen.
I
~
Progress
with
regard to this is easily determined when, with thei
bladder empty, the flat of a hand is placed along the
anterior surface of the symphysis pubis and the lower
abdominal wall. Either the parietal eminence will not
be felt (no overlap), or it will be flush with the
anterior aspect of the pubic symphysis (first degree
overlap), or it will be felt jutting out in front of the
pubic symphysis (second degree overlap).
It is unwise to attempt symphysiotomy if there is
second degree overlap.
Symphysiotomy may be indicated when more than
one third of the fetal head has entered the pelvic brim
and there is less than second degree overlap, provided
other circumstances are favourable.

Tropical Doctor, April 1974

high, the second stage not yet reached, and when a
clinical diagnosis of disproportion is made, it is best
to undertake symphysiotomy before wasted attempts
at operative vaginal delivery are made.
If the situation has been well chosen, division of the
symphysis pubis will be followed by rapid progress
in labour and spontaneous delivery. If delivery is
delayed, the vacuum extractor should be applied to
’
assist.
Forceps
should not be used as the soft tissues
1 *
under
the symphysis are by now completely un­
protected,
THE OPERATION

The description of the operation which follows is
essentially the one first suggested by Seedat and
Crichton (1962). A closed technique is preferable to
a
an open
surgical procedure as it may be done rapidly
and without general anaesthesia. Good results have
also been obtained by following the principles of
Zarate (1955) but this particular technique appears
to be more complicated.

(a) Position of the patient
Throughout the operation the patient’s thighs should
Dilatation of the cervix
be held at an angle of no greater than 90 degrees
After division of the pubic symphysis, it is essential (Fig. 1). If assistants are present, one should hold
to deliver the baby wdthin a reasonable period of time, each leg. With no assistance, the lithotomy position
Accordingly, labour should have progressed to either should be used and a bandage tied between the
late first stage or into the second stage. Very often patient’s knees to prevent over-abduction of the
when there is cephalopelvicdisproportion, labour
thighs. This position, which should be maintained
will initially progress satisfactorily until the dilatating throughout the operation, is necessary to prevent
cervix becomes compressed between the fetal
head
---------J over-separation of the symphysial joint. If the joint is
and the pelvic brim. Stasis will occur 1within
‘ ‘ ‘ the widely separated there will be an associated strain or
cervix and further dilatation will not take place. even subluxation of the sacroiliac joints.
Instead, the cervix will become oedematous. Relief of
the disproportion is invariably followed by rapid (b) Catheterization
cervical dilatation. It is safe to consider symphysio- A catheter should be passed and the bladder emptied
tomy when the cervix is more than 5 cm dilated (Fig. 1). To do this, the head may need to be dis­
in a multiparous patient or 7 cm dilated in a primi- placed upwards if the disproportion is significant. A
gravida.
firm catheter should be used as there may be difficulty
Symphysiotomy may be indicated when more than in inserting a soft rubber one. Haematuria, if present,
one third of the fetal head has entered the pelvic brim, should be noted.
when there is less than second degree overlap, and when
the ccrt'ix is more than 5 cm dilated in a multipara or (c) Local anaesthesia
more than 7 cm dilated in a primigravida.
The skin and tissues in front of, above, and below the
pubic symphysis should be liberally infiltrated with
Failed vacuum extraction
local anaesthesia. Ten to 15 ml of a 1% solution of
There remains one further indication for symphysio­ procaine or lignocaine are usually sufficient. Particular
tomy. When vacuum extraction has been attempted in attention should be made to the area just below the
seemingly favourable circumstances and delivery has pubic symphysis where there is a large concentration
failed after 15 minutes* skilful traction, a sym- of sensory nerve endings (Fig. 1). The perineum
physiotomy may be performed. This is the origin of should also be anaesthetized in preparation for an
“trial of vacuum extraction”. Too many such trials episiotomy. A full pudendal block is not necessary as
are not recommended. When the head is relatively forceps delivery is not contemplated.

Tropical Doctor, April 1^74

SYMPHYSIOTOMY

JI

inserting this finger, the catheter and with it the
urethra are displaced to one side of the midline.
The catheter and the finger will remain in position
until the operation has been completed.
(e) The incision and division of the symphysis
A solid-bladed scalpel is taken in the right hand and
inserted through a stab incision into the centre of the
joint. The blade should be pointing downwards
(Fig. 3). If the joint has been entered correctly there
should be little resistance and very little force is
required. Pressure should continue until the tip of
the scalpel is felt by the index finger within the
vagina. If the knife is pressed too far into the joint
little damage to the mother will be inflicted but the
operator will cut his own finger. The knife is now

Fig. 1. The patient's legs are held at an angle of
abduction of less than 90°; a catheter is inserted;
local anaesthesia is inserted over the pubic
symphysis and into the perineum.

V

1

I

I

(d) Displacement of urethra
The index finger of the left hand is inserted into the
vagina between the fetal head and the posterior aspect
of the pubic symphysis (Fig. 2). To do this the fetal
head may again need to be displaced upwards. While

Fig. 3. Division of the lower joint fibres.

Fig. 2. The index finger of the left hand,
inserted behind the symphysis pubis, displaces the
ureter to one side.

rotated upwards, the junction of the upper part of
blade and the anterior surface of the symphysis pubis
being the fulcrum for rotation. This action will divide
the greater part of the lower half of the joint and the
attendant ligaments (Fig. 3).
The scalpel is then removed, rotated through 180
degrees and reinserted through the same stab incision
with the cutting surface upwards. The upper part of
the joint is divided by rotating the scalpel downwards
(Fig- 4)By now most of the joint and its ligaments will have
been divided and the joint will begin to separate.
Some of the ligaments will, however, still be intact.
These can readily be felt by the vaginal finger and it

72 I SYMPHYSIOTOMY

Tropical Doctor, April 1974

may be used to guide the scalpel blade to them when
division may be completed. When the vaginal finger
can be inserted into the gap created by the joint
division, the operation is finished. The scalpel and
vaginal finger are withdrawn. If the operation has been
correctly undertaken, the head will now descend and
full cervical dilatation will occur if it has not already
done so.

Episiotomy
To avoid injury to the vestibule, an <episiotomy
/ '
is
obligatory unless the perineum is markedly deficient.

Delivery
Whether spontaneous or assisted by vacuum extrac­
tion, the baby should be delivered over the perineum
and not, as is usual, upwards over the mother’s
abdomen. This manoeuvre will avoid trauma to the
soft tissues under the symphysis pubis.

Repair
The stab wound should be closed. One catgut suture
is usually sufficient. This suture should be inserted
fairly deeply as the edges of the skin incision usually
bleed freely. The episiotomy should also be repaired,
aftercare

If there is no haematuria at any stage during the
operation, bladder drainage is not required. If haematuria is present, continuous bladder drainage should
be undertaken until the urine has been clear for at
least three days. It has been found that strapping is
not required. The patient should be sedated for about
eight hours. From then on ambulation should be
encouraged as soon as the patient wishes to get up.
Care should be taken to ensure that someone is
present the first few times that she gets out of bed
to prevent falling or other injury. As far as possible
it is the patient herself who should decide limitation
of movement. At first, walking should be aided by her
supporting herself round the bed and from bed to
bed. She should use short shuffling steps or 1walk
"
initially with a broad base. A physiotherapist, when
available, may be of considerable assistance. The
patient may be discharged when her walking is
confident and when she is free of pain. The immediate
convalescence may vary from five to 14 days. On
discharge she should be warned against undue
exercise for a period of three months.
COMPLICATIONS

Immediate complications from symphysiotomy are
not common. Haemorrhage from the vascular area
over the pubic symphysis is sometimes alarming. The

Fi^. 4. Division of the upper joint fibres.

bleeding almost always stops as soon as the baby is
bom and may be controlled before delivery by firm
pressure. Occasionally large haematomata occur which
track into the vulva or to the perivesical or parametrial
areas. Such bleeding is not dangerous if recognized
and dealt with promptly. Infection at the operation
site may also occur. If left unchecked it may affect the
joint and bones of the pubis. Severe infection is
readily prevented by the judicious use of antibiotics.
Urinary infection is as frequent as it is in any patient
who has been catheterized in labour. Soft tissue
injury is rare if the operation is reserved for cases in
which disproportion is not gross, although urethral
incontinence, due to the avulsion of the urethra from
its normal anatomical situation, is encountered. Treat­
ment of such a complication is not easy. Fortunately
its occurrence is rare and it is doubtful if it is more
t
frequent than ureteric injury
during Caesarean section.
Of greater importance is permanent pelvic injury with
subsequent ambulatory difficulty. It is not certain
how frequently women who have had symphysiotomy
develop such a complication. Follow-up is not easy in
the areas where symphysiotomy is frequently under­
taken, but any evidence available suggests that
ambulatory difficulty is not as common as might be
expected. The stability of the pelvis is dependent
mainly on the integrity of the sacro-iliac joints. Apart
from the intrinsic joint structure, there are powerful
anterior ligaments to prevent subluxation. When the
pubic symphysis separates spontaneously in labour.

Tropical Doctor, April 1^4

there is almost always associated damage to the sacro­
iliac joints with accompanying pain and often severe
ambulatory problems. During surgical symphysio­
tomy, the sacro-iliac joints are protected by controlled
femoral abduction which will prevent dangerous
separation in the sacro-iliac joints. In surgical sym­
physiotomy, too, the strain upon the sacro-iliac joints
is present for a short space of time, whereas when the
symphysis separates spontaneously during labour a
considerable time may elapse before delivery. Healing
within the symphyseal joint is by fibrous union, which
may be as strong or even stronger than it was before.
There can be little maternal mortality from sym­
physiotomy provided the operator is careful. Haemor­
rhage is a possibility as it is in any operation. Sepsis
is less likely to cause death as general spread from
such a limited area is not easy. The main complica­
tions are therefore intractable urethral incontinence,
injury to the urethra and bladder, and pelvic in­
stability. The first is seldom encountered, the second
is due to failure to undertake the operation properly,
and the last is probably less likely in practice than in
theory. All major complications will be due to bad
selection of cases, in particular the attempt to under­
take the operation in the presence of gross dispropor­
tion. From the fetal point of view there can be no
doubt regarding the safety of the operation. From the
time of decision to undertake symphysiotomy until
the actual delivery most operations take less than
15 minutes and the baby is born in much the same
condition as it was at the time the operation was
considered necessary.
SYMPHYSIOTOMY IN TRIAL OF LABOUR

Although symphysiotomy has been accepted by many
doctors who work in developing countries as a useful
operation in an emergency situation, its place in the
conduct of a trial of labour in hospital is in dispute.
Such patients may have attended an antenatal clinic
regularly and it is likely that they will return for
subsequent delivery. It is argued, therefore, that
Caesarean section, followed thereafter by a series of
such operations is preferable to symphysiotomy. This
is not necessarily the case. Many patients, having once
had a Caesarean section, may wish to prove to them­
selves that normal vaginal delivery is possible and
ignore the medical advice. They may not come to
hospital during the next pregnancy and be seen only
when the uterus has already ruptured. The only
justifiable argument against symphysiotomy is that
the complication rate can be high, not only imme­
diately post-operative but also at a later date. If the
complication rate is likely to be small then the
operation is justifiable in any circumstance.

SYMPHYSIOTOMY | 73

SYMPHYSIOTOMY BEFORE LABOUR

At one time the operation was advocated as an
elective procedure either before pregnancy or at the
time of emergency Caesarean section for proven
disproportion in the hope that the consequent pelvic
enlargement would ensure subsequent normal de­
livery. Such practice has little to commend it.
Disproportion cannot easily be diagnosed prospec­
tively unless the pelvis is extremely small or deformed,
and these very small pelves cannot be enlarged
sufficiently by symphysiotomy to guarantee future
adequacy without gross damage to surrounding soft
tissues.
SYMPHYSIOTOMY IN EARLY LABOUR

It is equally unwise to undertake symphysiotomy
early in labour. The operation may easily be under­
taken unnecessarily. Alternatively, and of greater
importance, prolonged strain on the sacro-iliac joints
which accompanies division of the symphysis without
rapid delivery may lead to ambulatory difficulties after
delivery in the same way that this follows spontaneous
separation of the symphysis during labour. It is also
notable that haemorrhage from the operation site
usually does not cease until the baby is delivered, and
when rapid delivery does not follow symphysiotomy,
this haemorrhage may be impossible to control.
SYMPHYSIOTOMY IN ABNORMAL PRESENTATION

Transverse lie
There is no place for symphysiotomy in transverse
lie, no matter the circumstances.

Brow presentation
When the pelvis is capacious, and the obstruction is
due to brow presentation, symphysiotomy may be
undertaken provided the rules regarding fetal head
descent overlap and cervical dilatation are followed.
Such circumstances are rare and only found in
emergency situations when the presentation is ob­
scured by a large caput succedaneum. The diagnosis
is then made with surprise at delivery. There is,
however, no justification in permitting labour to
continue when the diagnosis is made in early labour
in the hope that a symphysiotomy will eventually be
possible.
Face presentation
Symphysiotomy may be undertaken in face presen­
tation. Subsequent vacuum extraction is, however,
impossible and it is important to ensure that further
descent of the fetal head and greater dilatation of the
cervix has occurred before attempting the operation.
It is therefore limited to the occasional case when the

k
I

74 | SYMPHYSIOTOMY

second stage has been reached, the chin anterior and
the head engaged.

Tropical Doctor, April 1974
for cephalopelvic disproportion, an elective Caesarean
section should be undertaken. There is no place for
permitting labour in the hope that eventually a stage
will be reached when the abdominal and vaginal
findings will allow symphysiotomy. If labour is per­
mitted and operative delivery becomes necessary
because of unsuspected disproportion. Caesarean
section rather than symphysiotomy should be under­
taken. In an emergency situation when the patient is
admitted in advanced labour and conditions are
favourable for symphysiotomy, the operation may
occasionally be justifiable in the fetal interest. After
delivery uterine exploration should be undertaken to
ensure that the uterus is intact. Wherever possible,
however, symphysiotomy is to be avoided where there
is a uterine scar.

Breech presentation
The place of symphysiotomy in breech presentation
is quite different from when the head presents. If
the pelvis is small and the fetus of fair size, the best
results for fetal survival will be obtained when
Caesarean section is undertaken as an elective pro­
cedure. However, unless all breech presentations are
to be delivered in this way and until emergency
situations have been eliminated from clinical practice,
symphysiotomy is often useful to reduce some of the
very high perinatal mortality in such deliveries. If the
fetal head is disproportionately large for the pelvis and
this has not been anticipated, great difficulty in
delivery is encountered with much mortality and
morbidity to the fetus. In this dangerous situation a FAILED SYMPHYSIOTOMY
speedy symphysiotomy undertaken for delivery of the In the past eight years I have undertaken and assisted
aftercoming head may eliminate the difficulty, but the at over 500 symphysiotomies. There have been three
operation cannot be expected to give good results. failures. The reason for such failures is worth
However quick it may be, the time factor may be too reporting. In the first case, the patient had outlet
great for fetal survival and there is much greater disproportion which is unusual in Africa. Symphysiochance of the inexperienced causing severe damage to tomy, when recommended at all in Western practice,
the maternal soft tissues. It is possible to reduce fetal is considered to be suitable only for disproportion at
mortality in such a situation but impossible to elimi- the outlet. In this case, although the symphysis was
well divided, there was no appreciable separation at
nate it.
The operation may, however, be undertaken earlier the outlet. The second case is equally interesting as it
in the second stage in breech presentation. When demonstrates the importance of obeying the rules. A
there is slow descent of the buttocks in spite of good primigravida was admitted in obstructed labour. The
uterine and maternal effort, difficulty in delivering fetal head was present and on vaginal examination the
the aftercoming head may be anticipated. Un- head could be seen on separating the labia. There
doubtedly symphysiotomy at this stage, provided the was, however, second degree overlap present which
buttocks have descended to mid-cavity, may be was at first thought to be a shoulder. Division of the
followed by rapid progress and easy delivery. In the symphysis was not followed by descent of the head
15 cases personally undertaken in this situation, all and at the subsequent Caesarean section a hydro­
babies have survived.
cephalic fetus was delivered. The third failure had
had a previous symphysiotomy undertaken in another
Symphysiotomy in the presence of a dead fetus hospital. She arrived late in labour and because the
If the baby is dead and the delivery imperative, vaginal and abdominal findings suggested dis­
destructive operations are best provided there is no proportion of first degree a repeat operation was
danger of uterine rupture. Occasionally, when attempted. This proved impossible as the fibrous
destruction is not easy at a high level and the mother’s union between the joint surfaces was dense and
condition is poor, it may be better to undertake impossible to divide, thus demonstrating the sound­
ness of the joint.
symphysiotomy rather than Caesarean section.
Symphysiotomy in the presence of a previous
Caesarean section scar
During the antenatal assessment of a patient with a
previous Caesarean section scar a decision must be
taken as to whether to permit a test of labour (or trial
of scar) or to undertake an elective Caesarean section.
If the patient has had more than one Caesarean
section or if the previous Caesarean section has been

SYMPHYSIOTOMY IN THE FUTURE

There is no evidence to suggest that the frequency of
cephalopelvic disproportion is decreasing in develop­
ing countries. Kwashiorkor and allied diseases are as
prevalent as ever and as long as these diseases exist in
infancy there will be disproportion for a period of
40 years thereafter. Developing countries tend to
increase their population at alarming rates to such an

I

Tropical Doctort April 1974

extent that the medical services are unable to keep
pace. More and more doctors are being trained for
service in these countries. This has as one of its
results a tendency to increase the number of
Caesarean sections undertaken as that operation is
often an easy way out of an obstetric difficulty. It is as
easy to teach a doctor the indications for and the
technique of symphysiotomy as it is the indications
for and the technique of Caesarean section.
To train young doctors in the use of the vacuum
extractor or forceps is not difficult either. In practice,
however, where cephalopelvic disproportion exists,
delivery is not often easily accomplished without
some injury to the fetal head. As long ago as 1935,
Sir Albert Cook, pioneer of medicine in East Africa,
stated, “When I left England in 1896, it was with the
firm belief in the efficiency of extraction by forceps in
difficult labour. Greater experience has taught me
the folly and danger of this course.”
Symphysiotomy is able to reduce the number of
Caesarean sections which are undertaken. Moreover,
it can eliminate all cases of difficult vaginal delivery.
That it is not used more frequently is entirely due to
the seeds of prejudice sown in the minds of doctors in

SYMPHYSIOTOMY | 75

the nineteenth and early twentieth centuries from
ill-chosen and illogical evidence. The results which
were produced at the time by those who undertook
symphysiotomy in a rational fashion were good. The
operation has stood the test of time and will continue
to be practised in spite of this prejudice.

REFERENCES

Bowesman, C. (1960). Surgery and Clinical Pathology in the
Tropics, p. 665. Edinburgh: E. & S. Livingstone.
Browne, A. D. H. (1968). Emergency symphysiotomy.
Practitioner, 200, 537.
Cook, A. (1933). The influence of obstetric conditions on
vital statistics in Uganda. E. Afr. med. J., 9, 316.
Moir, J. C. (1964). In Munro Kerr’s Operative Obstetrics,
6th edition, p. 621. London: Bailliere, Tindall and Cox.
Pereira, J. C. (1964). Symphysiotomy in modern obstetrics.
An. Inst. Med. trap. (Lisboa'), 21, 153.
Seedat, E. K., and Crichton, D. (1962). Symphysiotomy;
technique, indication and limitations. Lancet, 1, 554.
Simon, E. J., and Philpott, R. H. (1973). The vacuum
extractor, Tropical Doctor, 3, 34.
Zarate, E. (1955). Subcutaneous partial symphysiotomy.
English edition, T.I.C.A. Buenos Aires.

D
Tropical Doctor, July 1985

Management of rupture of
the gravid uterus
H A Heij md1
E R te Velde md
J M Cairns mb frcs
St Francis Hospital, Katete. Zambia
TROPICAL DOCTOR, 1985, 15, 127-131

INTRODUCTION

OBSTETRICS AND GYNAECOLOGY | 127

be anywhere in body and lower segment, either
anterior or posterior, and the direction of the rent
may be transverse, vertical, or a combination (Tor L-shaped tears). Site and type of the rent can be
related to the cause: for instance, an anterior, trans­
verse tear may be due to dehiscence of the scar of
a previous lower segment caesarean section: or a
posterior tear may be due to the “boot scraper
effect of the fetal head in a forceful breech
extraction.
Uterine rupture usually occurs spontaneously, i.e.
without (iatrogenic) trauma, and is caused by
cephalopelvic disproportion, malposition and malpresentation. Traumatic rupture may occur during
obstetric intervention, either instrumental, manipu­
latory or pharmacological.
Rupture of a caesarean section scar may occur
either spontaneously or as a result of trauma and
offers a somewhat different picture. In particular,
the classical caesarean section scar in the body of
the uterus is at risk of rupturing in subsequent
deliveries.
Uterine rupture is rare in the primigravid, but
“the grand multigravid uterus is prone to rupture”
(Smith 1982).

Rupture of the gravid uterus is an obstetrical cata­
strophe associated with a high maternal and fetal
mortality rate. The incidence is particularly high
in developing countries: e.g. 1 in 112 hospital deliv­
eries in Nigeria with a maternal mortality rate of
7.6% (Groen 1974), against an incidence of 1 in
6673 deliveries in the USA with a maternal mor­
tality rate of almost 0 (Spaulding & Gallup 1979).
The reason for the high frequency in developing
countries is that cephalo-pelvic disproportion is
common because of growth stunting of future
mothers in childhood and adolescence, through mal­
nutrition and recurrent untreated infections. Also,
in the absence of prenatal care, disproportion and
malpresentations are undetected until the resulting CLINICAL SIGNS AND SYMPTOMS
obstructed labour brings the patients to hospital Rupture of the uterus is in many cases the final
when rupture of the uterus is imminent or has result of obstructed labour. For recognition and
already occurred. Unfortunately, unskilled manipu­ management of obstructed labour, the reader is
lations, either by untrained attendants or inexperi­ referred to specific texts (Lawson & Stewart 1967,
enced doctors, may also contribute to the increased Philpott 1982). When obstructed labour is not
incidence of uterine rupture in developing countries. relieved without delay, impending rupture of the
will develop.
The extension of antenatal screening to detect uterus
- • , The
, junction
m ring between body
patients at high risk would certainly reduce the and lower segment (the ring of Bandl) rises and the
incidence. Furthermore, the proper management of lower segment becomes stretched and painful to
disproportion and obstructed labour is important in touch, even between contractions, which are
increasing in strength and duration. The patient
the prevention of uterine rupture (Philpot 1982).
The high fetal and maternal mortality rate of becomes anxious and restless, with a rapid
- pulse
uterine rupture could be reduced by early diagnosis and irregular respiration. The bladder may be disand active ore-,
pre-, intra- and post-operative managemanage­ tended and catheterization may yield blood-stained
ment of these patients. This paper will concentrate urine. At this stage immediate intervention is
required to prevent rupture. However, injudicious
on these curative issues.
manipulations may further stretch the already
thinned lower segment and produce rupture. When
DEFINITIONS AND AETIOLOGY
ui
mt
utviuo
hivoho
w.
«
the
fetus
Rupture of the uterus means the development of a is dead, caesarean section is to be avoided
tear of theu terine wall.TewiliTimit ourselves here by the performance of embryotomy (LawsonJ974y
to intrapartum rupture. The site of the rupture may When the fetus is still alive, symphysiotomy and
ventouse extraction should be seriously considered.
Caesarean section may be inevitable to deliver a
’Correspondence to: Department of General Surgery, Dijkzigt
live baby which is in transverse lie or compound
University Hospital, Dr Molewaterpiein 40, 3015 GD Rotter­
presentation.
dam, Netherlands.

I

F

i

128 | OBSTETRICS AND GYNAECOLOGY

Table 1. Diagnostic features of uterine rupture
Impending rupture
Anxiety, restlessness
Painful lower uterine segment
Rising contraction ring (of Bandl)
Distension of bladder
Blood-stained urine
Rapid pulse rate, irregular respiration

Rupture of the uterus
Sudden cessation of contractions (history and examination),
followed by:
Little or no pain (often), or severe continuous pain (seldom)
Cessation of fetal heart sounds
Haemorrhage per vaginam (often only a little)
Fetal parts felt in abdominal cavity
Abdominal distension
Hypovolaemic shock (varying from mild to severe)
Vaginal examination

Tropical Doctor, July 1985

short period of intensive resuscitation, preferably
in the operating theatre, is justified. Preoperative
management is summarized in Table 2.
At this point, it should be emphasised that any
attempt to extract the fetus in this period will not
only remove the tamponading effect but also
increase shock and extend the rupture; it should
therefore be discouraged strongly.
First, at least one large-bore drip should be set
up. If necessary, the internal jugular or subclavian
route can be used, and are preferable to a cutdown.
Normal saline is still the best resuscitation fluid in
shock, and at least 1 litre should be given before
anaesthesia is started. Sodium bicarbonate, 100 ml
of 8.4% solution, will help to correct metabolic
acidosis. It is of course desirable to establish an
adequate circulation, with warm extremities and a
recordable blood pressure, but if haemorrhage is
continuing, this aim may not be achieved without
surgical measures to stop the bleeding. Transfusions
have to be started as soon as blood becomes available
- Hpreferably
i^viauiy ncsu,
ucuaubc ui
fresh, because
of viuiung
clotting uisiur
disturbances
like disseminated intravascular coagulation,
Vasopressor agents are of little use in hypovol­
aemic shock, but dopamine may be useful in relaLively low dosages (up to 10 /xg/kg per minute), to
improve renal blood flow and protect against acute
tubular necrosis. Higher infusion rates of dopamine
result in an alpha-adrenergic vasoconstrictor effect
that is useful in septic shock. Monitoring of the

The diagnostic features of uterine rupture are
listed in Table 1. When rupture actually develops,
the clinical picture may be either dominated by
severe haemorrhage and hypovolaemic shock, or it
may be more insiduous if the presenting part exerts
a tamponading effect on the tear. In both instances,
contractions come to a sudden end, sometimes to
be followed by a continuous severe pain; but more
often pains stop completely or ralmost
1---- * completely,
The patient is sometimes rather talkative,r which
may mislead attendants into thinking she is less ill
than she is. External vaginal haemorrhage may be
surprisingly little, even in shock. Often fetal parts
canlbe felt
' *• easily
" through
*
the abdominal wall (usu- Table 2. Preoperative management of uterine
ally the fetus lies in the peritoneal cavity). There; rupture
is usually some distension of the bowel: infrequently
this is gross - when the patient presents late. Vaginal Correction of hypovolaemia:
Normal saline, 1-2 litres
Sodium bicarbonate
examination may demonstrate disengagement of
8.4%, 100ml
the presenting part: often the examining hand passes
Blood transfusion
easily through the rupture into the abdominal
Antibiotic treatment:
Chloramphenicol 500 mg qds
cavity. Sometimes the placenta is born before the
i.v.
fetus. Infrequently, loops of intestine prolapse
Metronidazole 500 mg tds
through the rupture into the vagina.
p.r. (much cheaper than i.v.
route)
In some instances, uterine rupture will only be
detected at vaginal examination after an operative Prevention of aspiration:
Nasogastric tube
Aspiration of gastric contents
vaginal delivery, although this does not necessarily
through tube
mean that the obstetric interference caused the
Antacids e g. Mist. Magn.
rupture.

Trisil.

Circulatory support:

i

PREOPERATIVE MANAGEMENT

Once uterine rupture has been diagnosed, prepara­
tions should be made for emergency laparotomy.
There is no room for conservative treatment, but a

Dopamine, 10 Mg/kg per
minute
Digitalis preparations

DO NOT EXTRACT FETUS WITHOUT LAPAROTOMY

OBSTETRICS AND GYNAECOLOGY | 129

Tropical Doctor, July 1985

central venous pressure is time consuming and usu­
ally unnecessary, unless there are signs of congestive
heart failure, in which case digitalis preparations
are also indicated.
Antibiotics should be started in the preoperative
phase as well. As faecal flora are the main contaminent, a combination of chloramphenicol and
metronidazole is rational and, in our experience,
very effective.
ANAESTHESIA

Always pass a nasogastric tube before induction of
anaesthesia, empty the stomach carefully and instil
20 ml of Mist. Mag. Trisil. to counteract acid
contents and prevent regurgitation into the airway.
The choice of anaesthetic methods will be as limited
as the number of expert anaesthetists in most rural
hospitals. In fact, many doctors will have to deal
with these patients single-handed. General anaes­
thesia with tracheal intubation is the method of
choice, when available. Cricoid pressure should be
used during intubation to close the oesophagus and
prevent reflux of stomach contents. Ketamine drip
appears a safe alternative, as this agent does not
lower the blood pressure, but it gives no relaxation
of the abdominal muscles. Local anaesthesia may
be resorted to in some instances when no other
method is available. Spinal or epidural anaesthesia
are contra-indicated because of their hypotensive
effect.
THE OPERATION

There are two ways to deal with the ruptured uterus:
to repair it, or to remove it. Each policy has its
advocates, but all studies published are biased by
the authors’ preference, and no controlled trials
have been performed on the results of each method.
The advocates of repair of the ruptured uterus
argue that this operation is usually easier and speed­
ier to perform, with less trauma and blood loss than
in hysterectomy (Aguero & Kizer 1968, Sheth
1969, Groen 1974, Nasah & Drouin 1978). Others
prefer to extirpate the poorly circulated and con­
taminated uterus to prevent haemorrhagic and sep­
tic complications (Paydar & Hassanzadeh 1978,
Rahman & Fothergill 1979). However, it seems
wise to tailor the operation to the individual situa­
tion, thereby aiming at the smallest possible pro­
cedure that gets the patient off the table as soon as
possible, and in the best possible condition
(Mokgokong & Marivate 1976, Golan et al. 1980).

An example of an algorithm to help in decision­
making is presented in Figure 1.
The abdomen is opened by a generous midline
incision. First the fetus and placenta are extracted.
Large bleeders in the uterine tear are grasped,
taking care to avoid bladder and ureters. Diffuse
oozing is controlled by dry gauze packs or towels.
CHARACTERISTICS OF UTERINE TEAR

anterior

posterior
^—vertical

<-

transverse
body^
^ower
ragged

J*

extending
into
paracolpos
no

'V

segment
c^eanedged

repair

yes

J,

subtotal
__
total
hysterectomy
Figure 1. Algorithm for surgical management of
uterine rupture

After the bleeding has been temporarily arrested in
this way, and while fluid resuscitation is continued
with all efforts to re-establish the circulation, it is
time to assess the damage and to decide on the
further management.
When the tear appears suitable for repair, the
bladder is deflected first, as in lower segment
caesarean section. The broad ligament is now open
and both ureters can be identified by palpation. The
angles of the rent are located next and two corner
stitches are inserted to aid in exposure. Then the
rent can be sutured by a haemostatic running stitch,
taking large bites of uterine tissue. Subsequently,
additional bleeding points are ligated until the whole
field is dry. Finally, bilateral tubal ligation is per­
formed, because the scar that results from repair
of a uterine rupture is not sound enough to withstand
a subsequent labour. The only exception to this rule
is when a ruptured lower segment caesarean scar
has been repaired in a patient who is likely to return
for an elective repeat section before the onset of
labour next time. In all other patients bilateral tubal

F

130 | OBSTETRICS AND GYNAECOLOGY

I

I

l!

ligation is necessary to prevent another calamity in
the future.
If the patient’s condition is still not stable at this
point, it is probably wise to finish the operation and
close the abdomen. In most cases, however, the
combined resuscitative and operative efforts will
have resulted in a stable patient, and a search can
be made for additional injuries, particularly of the
urinary tract. Inspection of the bladder may reveal
a tear, but to be certain about its integrity, Bonney’s
blue should be instilled via the catheter. Full­
thickness tears should be closed in two layers and
a catheter be left in the bladder for ten days postoperatively. Lesions of the ureters are much less
common and more difficult to detect. Intraveneous
injection of methylene blue or indigo carmine may
help to discover leaks of a ureter. If there is serious
suspicion that one of the ureters has been caught
in a ligature, retrograde cannulation after opening
the bladder is indicated. If severe damage is con­
firmed, reanastomosis over a splint, or reimplanta­
tion in the bladder is necessary to preserve the
kidney. Other organs, especially the bowel, should
be inspected carefully for injuries, and then the
abdomen is meticulously cleansed. Lavage with at
least two litres of warm saline reduces bacterial
contamination by dilution. A drain should be left
in the pouch of Douglas as an early warning should
secondary haemorrhage occur. It must be removed
after 24 hours.
Hysterectomy is the operation of choice in all
complicated tears of the uterus. It is sometimes
surprisingly easy to perform when the tear is exten­
sive and the uterus almost completely detached. In
general, attempts to remove the cervix routinely are
not only unnecessary but also harmful, for the
urinary tract may be damaged during the procedure
and precious time and blood may be lost. Subtotal
hysterectomy should not be spurned in these cir­
cumstances when that is sufficient to obtain
haemostasis. Also, the cervix may be very difficult
to identify when the pelvic tissues are oedemateous
and bruised.
For hysterectomy, the bladder should be deflected
and the ureters identified over their whole length
in the operative field. The round ligaments and
adnexal attachments are clamped and divided and
the broad ligament opened. The uterine vessels are
clamped, after checking the ureters and the uterus
can be amputated at the supravaginal level unless
the tear extends into the cervix. In that case,
removal of the cervix is the only method to obtain
haemostasis. Finally, the cervical remnant or

Tropical Doctor, July 1985

vaginal cuff can either be closed with a running
stitch or sutured circumferentially with a haemo­
static stitch, leaving a communication with the
vagina for drainage of the haematoma. As in repair,
the peritoneal cavity is carefully cleansed and rinsed
with several litres of normal saline, a drain is left
in the pouch of Douglas and the abdomen is closed.
POSTOPERATIVE MANAGEMENT

Even when the patient is stable towards the end of
the operation, intensive postoperative treatment will
be required, as many complications may occur
(Table 3).

Table 3. Postoperative management
Circulation :
Respiration:
Nutrition:
Sepsis:

Monitoring of blood pressure, pulse rate,
hourly urine output, peripheral temperature
Fluid administration (supportive drugs)
Physiotherapy
Antibiotics if necessary
Oral intake as soon as possible
Parental feeding in peritonitis
«
Antibiotics
Surgical drainage

Psychological
support

Circulation
Close observation after the operation is necessary
to keep the fluid balance in order, and to detect
secondary haemorrhage as early as possible. The
best parameter for the circulation is the urine out­
put, which should be at least 1 ml/kg per hour.
Fluids should be administered in order to attain this
output. Blood pressure, pulse rate and peripheral
temperature (e.g. of the hallux) are also helpful,
but interpretation may be difficult when peripheral
vasodilatation occurs in the course of septicaemia.
If urine output decreases in spite of large amounts
of fluid, the patient may have acute tubular necrosis
as a result of (prolonged) shock. In that case, blood
pressure is usually normal again and fluid admin­
istration should be restricted to prevent severe
overloading of the circulation. Enough fluids should
be given to compensate for the previous day’s losses.
In most cases of acute tubular necrosis, the anuria
is reversible and is followed by an episode of polyuria
as a sign of recovery of the kidneys (Lawson &
Stewart 1967). As mentioned above, low doses of
dopamine in the acute phase may exert a protective
effect on the kidneys.

Tropical Doctor, July 1985

]

I

OBSTETRICS AND GYNAECOLOGY | 131

Respiration
Respiratory problems are common in the post­
operative period. The risk of aspiration of gastric
contents has been mentioned and remains actual in
the postoperative period. Therefore, the naso-gastric
tube should be left in place for at least 48 hours.
Atelectasis due to elevation of the diaphragm and
to pain should be anticipated early and combatted
by physiotherapy, mobilization and adequate anal­
gesia. Antibiotic treatment is necessary when pul­
monary infiltrates develop.

Nutrition is another important aspect of the
severely ill patients. When there are no signs of
peritonitis, oral feeding should be started a few days
after the operation. High calorie and protein feeds
can be made by adding sugar and electrolytes to
milk and administering these by nasogastric tube
if necessary. Total parenteral nutrition via a cen­
trally placed catheter is ideal in general peritonitis,
but is beyond the scope of most rural hospitals.
Alternatively, calories can be given via a peripheral
drip as Intralipid.

Septic complications
Peritonitis is likely to occur after laparotomy for a
heavily contaminated uterine rupture. Therefore,
antibiotic treatment should be continued for at least
48-72 hours after the operation until one is certain
that there are no signs of intraabdominal sepsis.
General peritonitis may develop, with a continous
fever, abdominal distension and ileus, and massive
fluid requirements because of third space losses and
respiratory problems. Alternatively, localized intra­
peritoneal abscesses may develop in the second
postoperative week, for instance in the pouch of
Douglas or in the subphrenic space. A spiking
temperature and clinical deterioration should initi­
ate a search for these abscesses by physical exam­
ination and chest X-ray. Drainage is indicated,
preferably by the extraperitoneal route, but re­
laparotomy may be inevitable when circulatory col­
lapse is impending.

Finally, the psychological aspects should not be
forgotten in these women, who have often not only
lost their baby but also the possibility of having
another.

General remarks
Continuous bladder drainage via an in-dwelling
urethral catheter should be maintained for at least
one week in all instances where labour has been
obstructed by disproportion, in the hope of avoiding
or limiting full-thickness pressure necrosis of the
bladder and subsequent fistula formation.

REFERENCES

AgUero O & Kizer S (1968) Obstetrics and Gynecology 31,
806-809
Golan A, Sandbank O & Rubin A (1980) Obstetrics and Gynec­
ology 56, 549-554
Groen P (1974) Obstetrics and Gynecology 44, 682-687
Lawson J B (1974) Tropical Doctor 4, 188-191
Lawson J B & Stewart B D (1967) Obstetrics and Gynaecology
in the Tropics. Edward Arnold, London; pp 203-210
Mokgokong ETA Marivate M (1976) South African Medical
Journal 50, 1621-1624
Nasah B T A Drouin P (1978) Tropical Doctor 8, 127
Paydar M A Hassanzadeh A (1978) International Journal of
Gynaecology and Obstetrics 15, 405-409
Philpott R H (1982) Clinics in Obstetrics and Gynecology 9,
625-640
Rahman MSA Fothergill R J (1979) Journal of the Royal
Society of Medicine 72, 415-420
Sheth S S (1969) American Journal of Obstetrics and Gyne­
cology 105, 440-443
Smith A M (1982) Journal of Obstetrics and Gynaecology 2,
245-248
Spaulding L B A Gallup D G (1979) Obstetrics and Gynecology
54, 437-444

l

I
F

I

E
188 | EMBRYOTOMY FOR OBSTRUCTED LABOUR

(m)

Tropical Doctor, October 1974

Obstetric Care
I

Embryotomy for
obstructed labour
John Lawson, FRCOG
Formerly Professor of Obstetrics and Gynaecology,
University of Ibadan, Nigeria9

TROPICAL DOCTOR, 1974, 4, 188-191

This article is thefourth in a series on obstetric operations
comwonZy performed in the tropics to relieve mechanical
difficulties in labour. The first article, on Caesarean
Section, appeared in the J
January I9f2 issue. The?
second, on the
T "
Vacuum Extractor, by Simons and
Philpott in January 1973, and’ the
' third,
’
on Symphysiotomy, by Gebbie in April 1974.
Destructive operations on the fetus are 1hardly ever
indicated nowadays in Europe, North America, andI
other developed regions. Disproportion and malpresentations are usually detected during pregnancy:
even when they are not, the resulting mechanical
difficulties are dealt with early in labour before the
stage of obstruction is reached.
In developing countries, however, skilled antenatal
care is by no means universally available. Contracted
pelvis is much more common, due to stunting of the
growth of future mothers by malnutrition and disease
in childhood and adolescence. Lack of confidence in
modem obstetric care may cause delay in seeking
treatment, and if the nearest hospital can only be
reached after an arduous journey, disaster may be
imminent by the time patients in prolonged labour
arrivc.------

The case against Caesarean section
At first sight, Caesarean section might appear to be
the complete answer to the relief of obstructed labour.
However, this operation carries immediate and
remote risks which cannnot be ignored.
Caesarean section is considerably less safe in small
rural hospitals, where most obstructed labours have
to be dealt with, than in well-equipped obstetric units
staffed by specialists. The operation itself may carry
considerable risks when performed by the “occasional
surgeon”, particularly from haemorrhage when
stored blood is not available. The difficulties of
•Present address: Department of Obstetrics & Gynaecology,
University of Newcastle upon Tyne, England.

anaesthesia for Caesarean section when the patient is
in poor condition after a prolonged labour may defeat
the anaesthetic skills available in these surroundings.
Furthermore, Caesarean section late in labour when
infection is already established carries grave danger of
general peritonitis which even massive doses of
broad-spectrum antibiotics may not control.
The later risk of rupture of the uterine scar in a
subsequent labour is even more important, although
less obvious. A woman who does not come to
hospital until she has been in labour for several day’s
does so only as a last resort. Next time she is pregnant,
she may deliberately avoid hospital care because she
associates a frightening operation and an uncomfortable puerperium with hospital delivery. If a section
does not result in a live baby, she may attribute the
loss of her child to the operation, and thus be even less
likely to return. In rural areas, many women live so
far from hospital that they cannot get there in labour
even when they wish to do so, particularly in nomadic
communities.
Another neglected obstructed labour after a
Caesarean section will probably be disastrous, as the
uterine scar will almost certainly rupture. Operative
vaginal deliveries should therefore be preferred.
particularly if the fetus is dead, if the patient is
unlikely to return to hospital for the next delivery
What kind of operative vaginal delivery?

Labour obstructed: fetus alive
If the fetus is still alive and labour is obstructed by
only a moderate degree of cephalo-pelvic disproportion, symphysiotomy should be seriously considered. If there is gross disproportion, however, this
will not secure delivery and Caesarean section may
have to be performed in spite of the risks.
When labour is obstructed by transverse lie or
compound presentation and the fetus is still alive,
Caesarean section is nearly always indicated. Internal
version and breech extraction is not a safe alternative:
when the uterus is firmly contracted round the fetus,
an attempt at version is almost certain to rupture the
lower segment.
Labour obstructed: fetus dead
When the fetus is dead, obstruction due to cephalopelvic disproportion can be relieved by reducing the
size of the head by craniotomy, and obstruction due
to transverse lie can be similarly relieved by decapi­
tating the fetus. These procedures are not difficult

Tropical Doctor, October

EMBRYOTOMY FOR OBSTRUCTED LABOUR : 1S9

with practice, and are safe provided they are gently
performed. Certain rules mu>t be observed, however,
if an embryotomy is to be successful.

Pre-operative management
Before embarking on any operation to relieve
obstruction, correction of the effects of the preceding
prolonged labour should be started. Dehydration and
ketoacidosis indicate the rapid intravenous infusion of
at least one litre of 5'?o dextrose. To combat the
inevitable intra-uterine infection, the circulation
should be flooded with a broad-spectrum antibiotic,
preferably intravenous ampicillin or tetracycline. If
stored blood is available, at least two units should be
cross-matched. These measures should be initiated
while preparations are being made for operative
delivery, but they should not be allowed to cause
undue delay.
Since a rupture of the uterus may be discovered
during or immediately after delivery, all destructive
operations must be performed under general anaes­
thesia in the operating theatre, with a trolley of
laparotomy instrumems ready for immediate use if a
rupture is found.
Immediately before embarking on a destructive
operation, the abdomen should be re-examined for
signs of uterine rupture. It should be established by
vaginal examination that the true conjugate of the
pelvic brim is not less than 8 cm.: the cervix should
be at least three quarters dilated.

■t'

I

i

■ V

CRANIOTOMY

When labour becomes obstructed in a cephalic pre­
sentation, the head is usually driven down firmly
into the pelvic brim where it becomes impacted:
craniotomy is then fairly easy. However, if the head
is high and mobile above the pelvic brim, craniotomy
is difficult and dangerous and subsequent extraction
may even be impossible. In these rare circumstances,
delivery of the dead fetus by Caesarean section, with
all its risks, will be safer.
Technique
The skull is perforated, preferably with a Simpson’s
perforator. The instrument is passed into the skull
up to the shoulders of the blades and opened widely:
it is then closed and turned through 90° and opened
again, to produce a cruciate opening in the vault. The
closed perforator is then inserted deep into the skull,
opened and rotated briskly to break up the septa and
the brain substance. The obstruction has now been
relieved by reducing the size of the head, so extraction
of the fetus can proceed.
Crushing instruments such as the cephalotribe and
cranioclast are completely obsolete: the three-bladed

Fig. 1. Extraction, using Morris's forceps, after
craniotomy.

combined instrament is particularly clumsy and
should be relegated to the museum. Extraction is
most easily achieved by pulling on several pairs of
forceps anached to the edges of the hole in the skull.
(See Fig. 1). Morris’s craniotomy forceps, which are
strong 10 inch clamps with curved toothed blades,
are most suitable for this: if they are not available,
strong volsella or even heavy* hysterectomy clampscan
be used instead.
The bones of the vault are firmly grasped by up to
four pairs of forceps, care being taken that folds of

i

I

I
i

vaginal wall or cervix are not accidentally included.

Firm traction on the bundle of forceps (and also
usually rotation) will cause the collapsed head to
descend. Sometimes a piece of the cranium to which
one of the forceps is attached will pull off: this does
not matter as the instrument can be reappb’ed taking a
deeper bbe, bringing the point nearer to the base of

I

f

190 ; EMBRYOTOMY FOR OBSTRUCTED LABOUR

Tropical Doctor, October 1^4

a

r

I

Fig. 2. Decapitation using Blond-Heidler saw. (a) Passing the thimble zeith sazu attached up to the neck.
{F) Drawing the thimble down over 1the
’ other side of the neck. {Posterior viezv.') (c) Severing the neck with the
saw. {d) Vclsellum attached to the stump of the neck to deliver the after-coming head.
the skull. With this method, the collapsed head is
elongated by traction on its apex. Crushing is there­
fore not required and jagged pieces of bone cannot
lacerate the vagina.
If the fetus is ven large, reduction in the size of the
shoulder girdle may be required after delivery of,the
head. This is achieved by cleidotomy, in which one or
both clavicles are divided with stout embryotomy
scissors.

DECAPITATION’

When labour is obstructed by a transverse lie, a
preliminary reconnaissance per vaginam determines
the exact position of the fetal neck. If possible, an arm
is brought down and firmly pulled on by an assistant.
This prevents upward displacement of the fetus
during manipulations and thus limits further stretching
of the distended lower segment, and it also brings the
neck lower to make it more accessible.

I
E

Tropical Doctor, October 1974

{

V

r

EMBRYOTOMY FOR OBSTRUCTED LABOUR | 191

Technique
A
A self-retaining
self-retaining urethral catheter should
should then
then be
be
If the fetus is small or macerated, the neck can usually inserted to drain the bladder continuously for at least
be severed^ v-ithout difficulty^ with embryotomy 48 hours. If the fetal head has been impacted in the
scissors. (The
hook is clumsy and difficult
for time before delivery, pressure
/tt.-decapitation
j
J
*vagina
• a long
tto use, particularly
................................................................
as the blade is usually blunt.)- If the necrosis of the bladder wall is likely. In these cases,
fetus is large or the neck is relatively inaccessible, the therefore, bladder drainage should be continued for
Blond-Heidler decapitation saw is best. (See Fig. 2).
at least 10 days in the hopes of preventing full­
The thimble of the Blond-Heidler set to which one thickness sloughing and thus the formation of a
end of the wire saw blade is attached is passed up vesico-vaginal fistula.
behind the fetal neck. The thimble is then drawn over
During the post-operative period, the correction of
the neck and brought down in front of it. Next, the dehydration and keto-acidosis begun before delivery
handles are attached to the ends of the saw and, will necessitate further intravenous therapy for at least
keeping them close together so that the vagina
is not 24 hours. Similarly, antibiotic therapy should be
«
injured, a few firm strokes soon sever the neck. The continued intravenously until the oral route can be
trunk is then delivered by traction on the arm: the substituted.
operator’s hand protects the vagina from laceration by
jagged spicules of bone protruding from the thoracic LOOK TO THE FUTURE
inlet.
When the patient has recovered, the cause of her
To deliver the after-coming head, the operator traumatic experience must be fully explained both
rotates i*
----uterus
*---- until the stump of the
• neck
...
- it in the
is her and- to her relatives.
That it could have been
pointing down the birth canal. The* stump is thenl 1prevented' \
by adequate obstetric care should be
grasped with a heavy volsellum and, with a finger in firmly emphasized. *
the mouth to flex it, the head is delivered by traction
T*
1 patient
„J
If the
is satisfied with the number of
on the volsellum as if it were the after-coming head of surviving children she has, sterilization1 or
or ccontraa breech. This is a simple procedure which prevents ceptive advice should be offered. If further’ PreS’
the stump of the neck from injuring the birth canal.
nancies are desired, however. Caesarean section will
be required next time. (It may sometimes be unwise
AFTER-TREATMENT
to tell the patient this before discharge, if it would
Extraction of the fetus after a destructive operation deter her from returning to hospital for her next
should be followed immediately by manual explora­ delivery.) In any case, it is essential to convince the
tion of the uterus to detect rupture and remove the patient before she leaves hospital of the importance of
placenta. Delay in diagnosing rupture of the uterus is skilled obstetric: care in all subsequent pregnancies
disastrous, as laparotomy must be quickly performed and deliveries.
if the patient is to survive. Indeed, if a rupture is
APPENDIX
discovered during an embryotomy before the fetus is The following instruments are recommended for the labour
delivered, vaginal manipulations must be abandoned ward embryotomy set:
4—Morris’s craniotomy forceps.
immediately and delivery effected per abdomen
1—Simpson’s perforator.
without delay.
1—Queen Charlotte’s pattern embryotomy scissors.
Having excluded uterine rupture the cervix and
1—Blond-Heidler decapitation saw with spare blades.
vagina are examined and any lacerations are repaired.
1—Combined breech hook and crotchet.

F
0 & G 46

16 July

SA Medical Journal

1204

(Supplement—South African Journal of Obstetrics and Gynaecology)

|
j

Obstructed Labour at the University Teaching
Hospital, Lusaka, Zambia—April
1972-December 1973
M. MPHAHLELE,

A. J. VAN DER MEULEN

SUMMARY
Sixty-three cases of obstructed labour encountered at the
University Teaching Hospital, Lusaka, Zambia, are analysed

for the period April 1972 - December 1973. For the same
period there were 27 348 deliveries and 1 432 Caesarean
sections. The management of choice was Caesarean
section, because of a lack of experienced medical staff
and the poor results obtained, together with the serious
complications which follow destructive operations before
vaginal deliveries. Eighty-five per cent of the babies were
delivered alive. There was no maternal death. Twenty-six
mothers remained in hospital for longer than 10 days.

MATERIAL
Obstructed labour was diagnosed in 100 patients during
the period April 1972-December 1973. Only 71 case
records were available and, of these, 8 were rejected as
not suitable for study, leaving 63 to form the basis of
this study. For the same period there were 27 348 delive­
ries and 1 432 Caesarean sections.

PRESENTATION AND CLINICAL
SUMMARIES

There was 1 case of a burst abdomen.

Head Presentation

5. Afr. Med. J.. 49, 1204 (1975).

There were 19 cases where the head presented. Nine
were brow presentations, 1 a face presentation, 2 occipitoposterior, and in 8 the exact position was not determined
owing to marked caput formation. Seventeen patients had
audible fetal heart sounds on admission, and 12 patients
had 8 cm or full dilatation of the cervix. In 1 patient the .
degree of dilatation could not be assessed and in 7 cases
the cervical dilatation was less than 8 cm. Four patients
were febrile on admission. In the group of 20 patients .
with the head presenting, only 1 was delivered vaginally, ;
the fetus presenting by the face, and the remaining 19
patients were delivered by Caesarean section, which
resulted in 15 live babies and 5 stillborn babies (2 macera­
ted). Sixteen babies weighed over 2,5 kg, and only 3 weighed
less than 2,5 kg. One baby’s weight was not recorded |
There was only 1 baby with an Apgar score of 4/10, and|
in 14 babies the Apgar score ranged from 6 to 10. Post­
opera lively, 8 patients became pyrexial. Ten patients j
were discharged within 10 days and 9 remained in hospita
for longer than* 10 days; 1 patient absconded after 8 daysThere was 1 case of a burst abdomen in this group.

Antenatal clinics are the backbone of well-run obstetric
services in developed countries. It is at these clinics that the
mothers ‘at risk’ are separated in time for appropriate
management to achieve a successful outcome of the preg­
nancy. As a result, complications of uncorrected abnorma­
lities and malpositions are rarely encountered in the labour
wards.
It is equally true to say that the standard textbooks of
obstetrics in developed countries such as the USA and
the UK mention in passing rather than discuss in detail,
the subject of obstructed labour. In the developing
countries obstructed labour is a major complication which
must be diagnosed early or, better still, must be anticipated,
if disasters are to be averted.
When it arises, it also requires doctors skilled and
experienced in the practice of obstetrics in the tropics.
In this Unit as far back as 1971, it was generally accep­
ted by the senior members of the medical staff that Cae­
sarean section should be the treatment of choice in cases
of obstructed labour, destructive operations and vaginal
deliveries assisted by instruments, since fistulae (urinary
and faecal), severe perineal and pelvic injuries and infec­
tions were well-known serious life-threatening complica­
tions which may follow such difficult vaginal deliveries
when labour is obstructed.
University Teaching Hospital, Lusaka, Zambia
M. MPHAHLELE, b.sc., m.b. b.ch., d.p.h., m.r.c.o.g., f.r.c.s.,
Conmltant Gynaecologist
A. J. VAN DER MEULEN, m.p., Registrar_____________
Date received: 28 January 1975.

Compound Presentation

There were 7 patients in this group. On admission
patients had a cervical dilatation of less than 8 cm and
had fetal heart sounds recorded; 2 patients were pyrexiaL _
Caesarean section was performed to deliver them
?
resulting in 7 live births; all but 1 weighed more th
2,5 kg. Two mothers had puerperal pyrexia; 5 werC '
for discharge within 10 days and 2 remained more tha
10 days. There was no fetal loss in this group.

" If '

I!'

16 Julie 1975

Mediese

SA

Tydskrif

1205

(Byvoegsel-Suid-Afrikaanse Tydskrif vir Obstetric en Ginekologie)
Transverse Lie without Prolapse of a Limb

There were 7 patients in this group. In all, the mem­
branes had ruptured before admission. One patient had
•;?.dergone 2 previous Caesarean sections. Two patients
h:.d reached 8 cm or full dilatation of the cervix. The
remaining 4 had a cervical dilatation of less than 8 cm.
Caesarean section was the mode of delivery in all, resul­
ting in 6 live infants; all weighed more than 2,5 kg. Two
mothers remained in hospital for more than 10 days.
Five had puerperal pyrexia.

There were 27 patients, 6 of whom had twins.
Three patients were pyrexial on admission; 8 patients
had 8 cm or full dilatation of the cervix on admission,
and 21 had reached less than 8 cm. Fetal heart sounds
were recorded in 18 mothers and in 5 they were not
heard. One patient had internal version for breech,
followed by extraction resulting in the stillbirth of an
infant weighing 1,9 kg (no fetal heart sounds on admis­
sion). The remaining 28 mothers were delivered by Caesa­
rean section. Six healthy sets of twins and 19 singletons
were bom. while 4 babies were stillborn. The firstborn of
each of 4 sets of twins presented with prolapse of the arm
The
fifth
patient
had
a
normal
vaginal
c^’ivery of the first twin at home, but on admission she
hau a Bandls’ ring, and a Caesarean section was perfor­
med. The sixth patient also had home delivery of the first
twin and on admission to the hospital the second twin’s
arm had prolapsed into the vagina. A Caesarean section
was performed, and a live baby was delivered. Seven­
teen patients in this group had pyrexia on admission.
There were 15 mothers who remained in hospital for less
than 10 days and 12 stayed more than 10 days. Thirteen
babies, including 9 of the twins, weighed less than 2,5 kg,
and 18, including 2 of the twins, weighed more than 2,5 kg.
Of the 5 babies who were stillborn^ 1 was thought to
have had feeble heart sounds; at 8 cm cervical dilatation a
Caesarean section was performed and a stillborn baby
weighing 2,5 kg was delivered. Two patients had no fetal
heart sounds on admission, but the cervical dilatation
was less than 5 cm. Caesarean sections were performed,
delivering 2 stillborn babies, each weighing 2,8 kg. The
fourth stillborn baby, weighing 1,9 kg, had had internal
version and breech extraction.

There were 3 mothers, paras 6, 9 and 4, who had pre­
vious Caesarean scars — the first 2 for the fourth and
third pregnancies respectively, and the third mother for
the second and fourth pregnancies. All 3 had cervical
dilatations of less than 8 cm. In 1 no fetal heart sound
was present on admission. Two of the babies survived,
each weighing 2,5 kg.
The parities are shown in Table I.

COMMENT

TABLE I. PARITY INCIDENCE
Parity

Presentation

Brow and face

Compound presentation
Transverse lie
Shoulder presentation with arm
or hand prolapse (singletons)
Shoulder presentation with arm
or hand prolapse (twins)
Total

3

1
1

2
0
1
1

1

2

6

2

1

14

7

1

0
8

2

3
0

47

Obstructed labour is a challenge to the midwife and obste­
trician; if unrelieved, death of the fetus is the outcome.
The mother runs the risk of dying undelivered, from a
rupture of the uterus. The mother may deliver, but nr
may
succumb to infection from general peritonitis or after
fistula formation and the concomitant pelvic infection
with or without renal failure.
Success in management of obstructed labour depends
on (a) early diagnosis; (6) early presentation of the mother
in the labour ward; (c) expert active management to
relieve the obstruction.
In our circumstances the Caesarean section offers the
best mode of treatment, because the majority of our staff,
both medical and nursing, have been trained in countries
where obstructed labour has virtually disappeared from the
labour wards. It is our policy that Caesarean section be
the treatment or management of choice, and vaginal deli­
very, with or without a destructive operation, is left to
experienced and senior medical staff members, even in
cases where the fetus is dead, for unless the cervix is
fully dilated there will be minimal intra-uterine manipu­
lation under good anaesthesia, and the uterus will contain
a fair amount of liquor amnii. Unless these conditions are
present, we strongly condemn vaginal delivery in cases
of obstructed labour. Symphysiotomy may be offered as
an alternative.
There is still a small place for destructive operations
in our circumstances: the lack of experienced personnel,
the type of pelvis, the not so common severe injuries to
the perineum and the complications that follow the socalled easy vaginal deliveries after destructive procedures,
make these procedures undesirable. Perhaps the time is
ripe for extraperitoneal Caesarean section as practised
by Mokgokong and Crichton1 in Durban to be encouraged.
This, we believe, would result in the birth of more healthy
babies.

Shoulder Presentation with Arm or Hand Prolapse

I

O & G

4
4
0

8
0
0
0

9
0
0
0

Total

1

7
0
0
0

4

1

0

0

0

21

0

0

0

0

0

1

6

8

11

2

0

0

1

59

6
0
0

1

5
4
1
2

4

3

1

1

9

8

1
1
1

19

7
6

I

I

1206
O&G 48

SA

Medical

Journal

16 July 1975

(Supplement—South African Journal of Obstetrics and Gynaecology)

In our circumstances we have often witnessed destruc­ equipment. Sixty-three mothers had obstructed labour,
57 of which were single pregnancies, and 6 twin pregnan­
tive operations performed on so-called well-chosen pa­
cies.
There was no maternal death, and 69 babies were
tients, with disastrous results. We agree with Philpott
et al.2 that ‘decapitation and extraction is a procedure delivered. Sixty babies (85,7%) were delivered alive. Of
which has caused rupture of too many uteri ... if there these, 43 weighed more than 2,5 kg, and 14 were of low
birthweight. Three babies’ weights were not recorded.
is evidence of severe thinning of the lower segment any
There were 9 stillbirths, 2 being macerated. The corrected
vaginal procedure will rupture the uterus. Caesarean
section is the best of the bad alternatives.’ Lawson and fetal loss was 11,4%.'
The over-all picture is certainly commendable — I
Stewart3 agree in principle with this, although they still think
that craniotomy is a good operation. This, in our expe­ burst abdomen and 26 patients remaining more than 10
rience, is not so, because in the majority of these cases days in hospital because of pyrexia, wound infection and
of obstructed labour the head is high, and even in neglec­ clinical intraperitoneal infection, was the price paid. The
ted shoulder presentation cases with the arm prolapsed, latter complication, in our opinion, speaks in favour of
extraperitoneal Caesarean section.
the neck has not been stretched sufficiently to allow easy
access for decapitation.
Rupture of the uterus is still fairly common in our
We wish to thank the Permanent Secretary for Health for
labour wards,4 and any procedure or management which permission to publish. We also wish to thank the medical and
prevents its occurrence, is welcome. Crichton and Boulle
nursing staff of the Units of the Department of Obstetrics
have enlightened us on this disaster in unscarred uteri and for their co-operation and vigilance.
have sounded a clear warning that in the neglected trans­
REFERENCES
verse lie, manipulation with a view to vaginal delivery
should be done by the experienced. Ampofo,4 writing in
1. Mokgokong, E. T. and Crichton. D. (1974): S. Afr. Med.
^88.
1969, stated that he found obstructive labour to be the 2. Philpott. R. H.. Sapire. K? E. and Axion J H. M Obstetrics.
Familv Planning and Paediatrics. .4 Manual of Practical Management
main cause of 59 maternal deaths over a 5-year period
for Doctors and Nurses, p. 31. Family Plannmg Association of
Rhodesia.
_
.
in a West African maternity hospital.
3 Lawson J B. and Stewart. D. B. (1967): Obstetrics and Gynaecology
in the Tropics, pp. 172 - 218. London: Edward Arnold
In our series, 27 348 deliveries and 1 432 Caesarean
Mphahlele. M. (1971): S. Afr. J Obstet. Gynaec.. 9. 45.
sections were performed in 21 months
a heavy work­ 4.
5. Crichton. D. and Boulle. P. (1964): Lancet. 1. 360.
6.
Ampofo. A. D. (1969): West Afr. Med. J., 18. 77.
load which taxed our meagre facilities of manpower and

ioo
COMMUNITY HEALTH CEU
326. V Main, I Block
KoTamangala" ** “ '
Bingalore-560034
India

G
204

J. INDIAN M. A., VOL.,72, NO. 9, MAY 1, 197Q

2
Destructive Operations in
-Obstructed Labour

Craniotomy was done in 129 (74 r
'-"I
'cases. ATcftes'proTcing11 olSrulJiJn^^iS

D. C. DUTTA*

w"
28
Table 2 shows gross complications after J
tractive operation.
atter..deg
s
Table 2-Showing the Gross Complications'
Destructive Operations
'

{Received for publication August 3, 1978]
* .
Obstructed labour is still prevalent amongst
rural mothers in the developing countries. An
incidence of 2.3 per cent and a maternal death
rate of 11.4 per cent are enough to indicate its
prevalence and the magnitude of the risk involv­
ed (Dutta and Pal, 1978).. Specialists either
• trained in urban based institutions of the country
or abroad are confused at time, in tackling such
cases in odd environments. With .this objective,
: a critical evaluation has been attempted about
the use of destructive operations in tackling
such cases who are rushed to the referral hospi­
tals in poor condition.

•

MATERIAL AND METHOD

The materials were collected from the perso­
nal series, while the author was attached to the
District Hospitals of Jalpaiguri, Suri and
Chinsurah, West Bengal, covering the period
from 1965 to 1973. During this period, 173 des­
tructive operations were performed amongst 324
obstructed labour cases, showing a frequency of
z 53.2 per cent. For comparative analysis, the hos­
pital statistics of Suri (1967) inxj of Nilratan
Sircar Medical College and .Hospital,’ Calcutta
(1976), were also taken into account.
OBSERVATIONS

Duration of labour—Labour was prolonged
for more than 48 hours in 24.9 per cent, between
25-48 hours in 54.3 per cent and less than 24
hours in 20.8 per cent cases.
Causes of obstructed labour—The causes c.
of
obstructed labour, which necessitated destruc­
tive operations, is shown in the Table 1.
Table 1— Showing the Causes of • Obstructed Labour

Causes

No. of cases

Contracted pelvis and disproportion
Transverse lie •
Occipitoposterior presentation
Face and brow presentation
Hydrocephalus
Foetal ascites
Conjoined twin
Total

122 (64.7%)
41 (23.7%)
8 (4.6%)
7 (4.0%)
2 (1-2%)
2 (1-2%)
I
(0.6%)
173 (100%)

Nilratan Sircar Medical College and Hospital, Calcutta
•M.B.B.S., M.O. (Cal.), Assistant professor (P. P.
Unit)

Uterine rupture
Craniotomy (129)
Evisceration (41)
Vesicovaginal fistula
Complete perineal tear
Maternal deaths

*7

No. of casesj]
(4;0%) |

2 (1.5%)
5 (12.2%)

5 (3.0%) |
2 (1.5%) 3
10 (5.8%) |
‘ • c-

----------------DISCUSSION

th

’

—

Complication

Iab°Ur’ Which is a^most a Sy

ckorJd;{!on and became a forgotten chapter-ill
e obstetrics of the developed countries is Ytilra

th»
fre-aS’ comPrising about 75 per
the popuiahon. Ill-nourished m.thers, wltSI
mia with1136?13
fre^ent co-existingsB&i
referaal hns^'r063 °f SepSiS’ are ^hedt^gej
tab!0®™uni?ation o^o'herwfst'XSMW

admitted ^ith labour,

lasting

more

thanks

caesareo of .?estnictive operation vis-a-vis
in nr, t sechon—Controversy in teaching and
rl».P[aC-ICe prevaiIs between the developed and
developmg countries about the use of destract ye_ operations vts-a-vis caesarean ^ection^in
such types Of cases. The extensive useW
caesarean section in the developed countries
.V.as advocated by Myerscough (1977). EnipHaHqaai 016 safet-v of caesarean section Taylor
1
t 5?en. adv°cated caesarean section- in
dead babies, -especially in transverse lie and
face presentation.
Table 3 shows comparative risks in caesarean
section and destructive operation in obstructed
labour.
Table 3 shows that there was 12.5 per cent of
maternal deaths, out of 120 patients who under­
went caesarean section, in contrast to that of
mV Pf.r cent iri case of destructive operations.
1 he difference in the death rate was found to be
statistically significant (P Z .05),
Although the materials are not identical, the
comparative result of caesarean section and des-

DESTRUCTIVE OPERATIONS IN OBSTRUCTED LABOUR—DUTTA

205

Table 3—Showing the Comparative Risks in Caesarean Section and Destructive Operations in Obstructed
Labour
Operation

Total
No. of
No. of cases maternal
deaths

Caesarean section: .
Personal series (1965-1973)
District Hospital, Suri (1967)
Nilratan Sircar Medical College and Hospi­
tal, Calcutta (1976)
Total
Destructive operations:
Personal series

No. of
stillborn

No. of
neonatal
deaths

Na of
perinatal
deaths

68
27

6 (8.7%)
6 (22.2%)

15

9

10
2

25 (36.8%)
11 (40.7%)

25
120

3 (12%)
15 (12.5%)

2
26

13

I

3 (12%)
39 (32.5%)

173

10 (5.8%)

173

173 (100.0%)

tructive operation as shown in Table . 3 is 1 is completed by either using cranioclast or
enough to indicate that caesarean section is
giant vulsellum (Lyon’s forceps) holding the
neither safe for the mother, nor appreciably
perforated margin. Care should be taken during
improves the foetal salvage. A fatality of 12
delivery of the shoulder to avoid perineal
per cent at Nilratan Sircar Medical College
injury.
and Hospital, Calcutta, substantiates the fact, if
Decapitation or evisceration—The
classic
the quality of the patient is poor, even improved
teaching to relieve obstruction in transverse lie
ancillary aids and good environment cannot
is decapitation if the neck is easily accessible
improve the results significantly.
and evisceration if the neck is high up. Whatever
There remains no room for controversy in
method is employed, the theme is decompression
deploying destructive operation'where the baby
with minimum intra-uterine manipulation. There
is dead. Similar views have been endorsed by
.should not be any dogmatism and the individual
Bhowmick (1974) and Gogoi (1971). Whether it
can employ his conversant skill and judgement
is to be extended selectively, even in a living
best suited for the case and the . environment.
baby with pronounced effect of obstruction on . The author performed evisceration in all cases
the mother with sepsis, is to be thought of, as
even as an alternative to decapitation. Decom­
an alternative to caesarean section. Similar
pression could be achieved .with minimal intra-v
view was also expressed by Jhirad (1954). The
uterine manipulation, a pair of sharp pointed
risk
involved
in <caesarean section,4 apart
x ‘
scissors and . 2 giant vulsellum forceps are the •
from death, has been pinpointed by the author
only instruments required (sharpness of the
in previous communication (Dutta and Pal,
knife attached with decapitation hook is pro­
loc. cit.). Lister (1960) mentioned of the num­
bably lost after 2 operations) and can be
ber of stillbirth to be 10 out of 41 patients, who
employed whether the neck is accessible or not.
underwent caesarean section, having audible
After taking out the. viscerae, the trunk is
foetal heart sound at the beginning of the
delivered using 2 giant vulsellum forceps giving
caesarean section. In the present series, out of ; successive traction until the podalic pole is
65 patients with audible foetal sounds at the
delivered. In case of difficulty spondylotomy v
beginning of the caesarean section, 15 gave
may be done. This is followed by delivery of the
birth to stillborn babies.
rest by breech extraction.
Craniotomy—It is the commonest method of
Exploration following destructive operation—
extraction of the baby in obstructed labour
Routine exploration of uterus following destruc­
(3/4th in the series). One may find, at times - tive operations should be mandatory. It is
difficult to introduce a catheter to ' empty the
indeed difficult at times to detect the rent
bladder which is only possible with decom­
because of soft, thin and flabby lower segment.
pression of the head following perforation lead­
All the 7 cases so detected were promptly dealt
ing to the escape of ^he brain matter. While in
with by hysterectomy without fatality. In 5 out
hydrocephalic head, a‘ pair of sharp pointed
of 7 cases, the margins of the rent were found
scissors is enough to perforate the skull, in
gangrenous with thrombosed blood vessels
others a perforator have to be used. Because of
suggestive of sloughing of the tissues by
too much moulding of the head and formation
pressure necrosis and in fact those were cases
of big caput, the suture line is mostly not
of pre-existing unsuspected uterine rupture and
demarcated but the perforation may be done in
should have been dealt with by primary
the dependent area of the vertex. After decom­

pression of tHe head tEe extraction of the baby

laparotomy.

206

J. INDIAN M. A., VOL. 72, NO. 9, MAY 1, 1979

Routine postoperative bladder drainage.—Con­
tinuous catheter drainage is a must following
destructive operation. Not only it gives adequate
rest to the bladder but by preventing repeated
catheterisation minimises chance of infection.
Cumulative effect of these favour healing of the
bladder necrosis, if any. How long the catheter
is to be put in depends on the degree of
obstruction, colour of urine and regaining of
bladder tone. One may have to keep it. for 48
hours to as long as 7 days or more'. In the
present series out of 5 sloughing vesicovaginal
fistulae, 3 developed soon after operation and 2
' developed on the 3rd day. In all -the cases, the
duration of labour had been more than 48 hours.
Continuous catheter drainage was given in 4
cases up to 2 weeks. In one. it was kept for'3
weeks when there was spontaneous closure of
the fistula.
Prevention of perineal tear— In long-standing
obstructed labour, the perineal and the adjoin­
ing tissues become oedematous and infected and
one is tempted to deliver the baby without
episiotomv especially in primigravida. rightly so
because of inevitable non-union of the wound.
While the compressed head could be delivered
uneventfully without damaging the perineum,
one should take an attitude of caution in deli­
vering the shoulder. Preliminary cleidotomv
either unilateral or bilateral minimises the girth
of the shoulder and thereby prevents injury of
the perineum.
- . . .
Maternal Mortality—There was 10 maternal ‘
deaths. 4 within 24 hours and 6 after varying
interval between 3-7 days. The causes of deaths
were due to combined effect of dehydration,
sepsis and anaemia. Rapid infusion of few
bottles of 5 per cent glucose to correct keto­
acidosis soon after admission, administration of
appropriate antibiotics and blood transfusion are
the essential adjuvant to reduce the death rate.

SUMMARY
Consecutive 173 .destructive operations per­
formed in 325 obstructed labour cases, giving a
frequency of 53.2 per cent were analysed. Com­
parative maternal risk in caesarean section and
destructive operation in obstructed labour
showed
statistically significant increase of
maternal deaths in the former (P
.05). Cranitomy was done in 74.6 per cent and evisceration
in 23.7 per cent cases. Practical guidelines about
the steps of destructive operation -were dis­
cussed. Complications included uterine rupture
in 7. vasicovaginal fistulae in 5. complete perineal
tear in 2 and maternal deaths in 10 (5.8 per
cent).

ACKNOWLEDGMENT

The author is grateful to the D.M.Os, Sadar Hos­
pital, Jalpaiguri, Suri and Chinsura for kindly allow- ■*
ing him to utilise the hospital records.
REFERENCES

BHOWMICK, A. C.—J. Obstet. Cynaec. India, 24 • g
1974.
’
DUTTA, D. C. and PAL, S. K.—Ibid., 28: 55, 1978.
GOGOI, M. P.—J. Obstet. Gynaec. Brit. Comm., 78 •
373, 1971.
JHIRAD, J.—J. Obstet. Gynaec. India, 5 : 231, 1954.
LISTER, U. G.—J. Obstet. Gynaec. Brit. Emp., 67;
188, 1960.
MYERSCOUGH, P. R.—Munro Kerr’s Operative
Obstetrics, 9th ed., 1977, Bailliere Tindall, Lon-,
don, p. 573.
TAYLOR, E. S.—Beck’s Obstetrical Practice, 8th ed.,
1966, Williams and Wilkins, Baltimore, p. 549.

3
Lung Abscess
■
P. B. DAS*
[Received for publication September 29, 1977]
The importance of lung abscess, as a thera­
peutic challenge to the clinicians, has steadily
declined since the advent of antibiotics (Walcott,
1957 ; Ghosh et al., 1966; Saha, 1966; Chakraborty and Naidu^ 1966; Shankar, 1969 ;Jhaef al.,
1972 ; Nigam et al., 1973 ; Chidi and Mendelsohn,
1974). Besides modern anaesthesia, refinements
in the techniques of oropharyngeal surgery and
better
understanding
of
its pathogenesis
have contributed considerably towards this end.
Consequently, many lung abscesses get arrested
in the stage of pneumonitis and mortality and
morbidity rates have also remarkably diminish­
ed. Newer antibiotics have further augumented
the aramentarium (Chidi and Mendelsohn, loc.
cit.). However, this trend is still quite slow in
developing countries where the incidence of
suppurative lung diseases is still high (Misra
et al.. 1969). Naturally, this continues to be a
major health problem. This presentation aims
at documenting the experience in the manage­
ment of 32 consecutive cases of lung abscess
treated
at the
institution
and attempts
to update its treatment relevant to the needs of
the developing world with a brief review of
related literature.
Department of Cardiovascular and Thoracic Surgery,
Safdariang Hospital, and University College of
Medical Sciences. New Delhi
•M.S., F.R.C.S.(C), F.A.C.S., F.I.C.S., F.C.C.P.,
Head of the Department

.

■

•

H

-

Tropical Doctor, April 1988

Postnatal care*
C H W Bullough VID FRCOG
South Shields General Hospital
South Shields, UK
TROPICAL DOCTOR, 1988, 18, 79-83

This is the last in a series aimed at outlining the
essential elements of an obstetric service run by a
nonspecialist doctor working in a district or mission
hospital. Although postnatal care is the last activity
performed in the management of pregnancy, it is
not the least important. It is, however, a relatively
circumscribed subject, ideally suited for delegation
to midwives or other staff, e.g. clinical officers,
provided clear protocols of management are avail­
able. This policy of delegation with supervision has
been stressed in other contributions to this series.
We can begin with an aspect of care which affects
every mother delivered of a living baby, and that
is the contact which occurs between mother and
baby after birth. The ways in which immediate skinto-skin contact promotes bonding, and the benefi­
cial effect early breastfeeding has on subsequent
breastfeeding performance have been widely docu­
mented. This can be seen in an extensive review of
the subject1. Although it is invariably midwives
who implement such a policy, it is the responsibility
of doctors to see that it is put into practice. A rule
that breastfeeding should be begun in the labour
ward makes supervision of the policy more possible.
SEPARATION OF ABNORMAL
FROM NORMAL CASES

After a normal pregnancy and labour, and a
spontaneous delivery, mothers require little except
hygienic surroundings, rest and observation by
midwives. However, not every woman is so fortu­
nate, and many do require very careful and frequent
nursing and medical care. In the crowded conditions
common in maternity units in developing countries
such patients can easily be overlooked among the
mass of healthy mothers. In these conditions the
first essential to ensure efficient postnatal care is
a physical separation of normal from abnormal
cases. If a separate ward cannot be provided, then
at least certain beds must be put aside for the
*Last of a series on The Essentials of an Obstetric Service. See
also: Volume 17 (1987), pages 77, 124, 174; Volume 18 (1988),
page 25

OBSTETRICS AND GYNAECOLOGY

|

79

Table 1. Patients requiring admission
to abnormal postnatal ward
Care required

Category 1: Complications
of pregnancy
Anaemia
Hypertensive disorders of
pregnancy
Bad obstetric history
Antepartum haemorrhage
Almost all conditions which
result in antenatal admission

Investigation and
treatment of persistent
abnormality. Advice
about recurrence

Category 2: Complications of
labour
Caesarean section
A
Ruptured uterus
Destructive operations
Instrumental deliveries (except
Post-operative care.
for low vacuum extractions)
I Advice about recurrence
Symphysiotomy
Retained placenta
Postpartum haemorrhage
Third-degree perineal tears
J

Category 3: Fetal and
neonatal disorders
Stillbirth
Congenital abnormality
Growth retarded fetus
Pre-term infant
Baby at risk for social
reasons, e.g. unmarried or
very young or old mother
Baby in neonatal nursery for
any other reason

Sympathetic understanding
of midwives and doctors.
Advice about causation.
r Advice about recurrence

J

purpose, so that midwives and doctors may identify
those needing special care. Guidelines as shown in
Table 1 will ensure that such a policy is implemented
efficiently.
INFLUENCING THE OUTCOME
OF FUTURE PREGNANCIES

By giving good advice to patients and by recording
our findings and opinion about how the next
pregnancy should be managed, we can influence the
outcome of future pregnancies. The more usual
forms of advice required concern the need for
antenatal clinic attendance, and the need for
hospital delivery because of the likely recurrence of
an abnormality. The need for an elective Caesarean
section in the next pregnancy must also often be
explained; less frequently patients must be advised
that future pregnancies should be avoided, and
sterilization offered. Such advice is often not

initially understood and, if it is, is readily forgotten.
Midwives therefore need encouragement to repeat

80

|

OBSTETRICS A.\D GYNAECOLOGY

it with further explanation when they can, and
routinely on discharge of the patient.
The more certain way of having this advice and
opinion about future management acted on is to
record it on a permanent record kept by the mother.
Optimally this will be a specially designed
“mother’s card” which contains space for details of
a woman’s entire reproductive history. Such a
document can be as essential to the mother as an
under-5 clinic card is to her child, and yet is lacking
in many maternity services. The best alternative is
to record the details on a stiff cardboard outpatient
card. These are often kept for years, in contrast to
notes written on ordinary paper. A letter written
to a referring doctor or midwife, and the hospital
inpatient notes are the documents least likely to be
available when needed.
The value of passing on details of a pregnancy
and delivery can be illustrated by taking an example
from each of the three main reasons for admission
to the abnormal postnatal ward.

I
I

Antenatal problem: A patient had persistent unstable
lie, and congenital uterine abnormality was finally
diagnosed at Caesarean section. Record this and
what type of abnormality you found. This will make
it clear that elective Caesarean section will be needed
in the next pregnancy if the lie is again unstable.
Problems of labour: A stillborn infant was delivered
by vacuum extraction after a prolonged labour at
home. Postnatal pelvic examination showed the
pelvis to be contracted and lateral erect X-ray
pelvimetry confirmed this. Record this and your
opinion that elective Caesarean section will be
necessary in future. If these details are not known,
the woman may be allowed to labour again next
time, with the same fatal result for the infant.

Tropical Doctor, April 1988

Investigation of cause of perinatal death
Before leaving the subject of the information and
advice patients require, we must give further
attention to the subject of perinatal death. Although
in traditional culture such a happening is often not
questioned but attributed to a misdemeanour of one
or other parent, this must not dissuade us from
looking hard for the cause.
The mother’s obstetric history and the clinical
features of the pregnancy and labour may make
diagnosis obvious. Nevertheless, there should be a
standing order in the labour ward that all stillborn
babies and their placenta, and the bodies of all
babies dying early in the neonatal period should be
kept for examination. A scheme for investigation
is shown in Table 2, although it is recognized that
many of these tests will not be possible under the
conditions of rural hospitals. To make a diagnosis
of the many possible congenital syndromes it is
helpful to have a book of illustrative photographs.
If chromosomal analysis of the baby’s tissues is not
possible the parents can be sent for karyotyping at
a later date.
COMMON CLINICAL PROBLEMS

A number of clinical problems are so common in
the puerperium that there should be an agreed plan
of management, preferably detailed in a ward
protocol. Ward staff can then investigate and
Table 2. Investigation of perinatal death
Baby
Placenta

Cord

ii

Fetal or neonatal problems: In a case of intrauterine
death with a fresh stillbirth the baby was light for
dates (below the 10th centile of weight for gesta­
tion), and the placenta was small and infarcted.
There was no fetal abnormality and so intrauterine
growth retardation was suspected. From other
clinical details you may have decided that the vital
factor in antenatal care in the woman’s next
pregnancy will be the early treatment of hyper­
tension or regular treatment with antimalarials, or
nutritional supplements, or rest and observation in
hospital, or all of these. Record these clearly. Initial
care may be undertaken by a midwife in a peripheral
clinic, who might otherwise take no action.

Look for evidence of:
Congenital abnormality, trauma, hydrops,
growth retardation, macrosomia of diabetes
mellitus
Abruption, infarction, small size in relation
to baby
True knots, abnormal vessels

Further investigation on baby
X-ray for skeletal
abnormalities
Bacterial culture from blood,
skin and pharynx
Autopsy
Chromosomal analysis

requiring increasing
11fiand
ind AvnArti
facilities
expertise

(
Maternal investigations
Serological tests for syphilis
in all unexplained
GTT within first week of
stillbirths
puerperium
Viral studies
where appropriate
Sickling test

C

r

I
c

V

OPERATIONS TO RELIEVE OBSTRUCTION

189

have sub-epicranial haematomas containing more than 200 ml. of blood.
It is claimed by enthusiasts that the vacuum extractor can be used to flex
and draw down the head to a level at which forceps can be used successfully.
The writer has no personal experience of this: even if the procedure is
worth employing when labour is delayed solely by ineffective uterine action
(which is doubtful), it is not likely to be helpful when labour is obstructed
by disproportion.
If labour is obstructed by a combination of disproportion and malrotation,
there is no doubt that Kielland’s forceps are more effective, but the efficiency
is achieved at the risk of damage to both the fetus and the mother. Com­
parison between the two instruments merely emphasizes that both the
vacuum extractor and the obstetric forceps are seldom suitable for com­
pleting delivery when labour is obstructed by cephalo-pelvic dispropor­
tion. The beginner, however, may do less harm with the vacuum extractor,
particularly if he is wise enough to desist after three pulls if they have not
produced considerable progress in rotation and descent of the head.
One useful function of the vacuum extractor in obstructed labour,
however, is to complete delivery after symphysiotomy. In these circum­
stances it is extremely valuable, and far less dangerous to the mother than
obstetric forceps.

i

Symphysiotomy
Enlargement of the pelvis by dividing the symphysis pubis is one way of
dealing with a common problem in tropical obstetrics. This is posed by the
patient who arrives in hospital with labour obstructed by cephalo-pelvic
disproportion, having so far received no obstetric care. The fetus is still
alive: a difficult forceps delivery would almost certainly result in a stillbirth
and possibly in severe maternal injury’. Caesarean section is also unattractive,
because of the very real risk of rupture of the scar in a future unsupervised
labour. In such a case, the balance of risks between this and the possible
sequelae of symphysiotomy may justify division of the symphysis, provided
it is correctly performed on carefully selected patients.
The value of symphysiotomy has until recently been obscured by the
uncritical enthusiasm of its advocates, especially those anxious to avoid
repeated Caesarean sections because of the eventual necessity’ for steril­
ization. However, Seedat, Lasbrey and Crichton of Durban have
now exhaustively evaluated symphysiotomy and have perfected a safe
technique.
Indications. The great difficulty is to decide when symphysiotomy is
indicated. Too ready recourse to it will result in unnecessary operations in
cases when spontaneous delivery would occur in time. On the other hand,
if it is performed when disproportion is too gross, either vaginal delivery

I

190

OBSTRUCTED LABOUR

will not follow and Caesarean section will still be necessary (the worst of
both worlds), or delivery will take place at the cost of excessive separation of
the symphysis which may result in serious urinary and locomotor disabihties.
In the opinion of the writer, the operation undoubtedly has a place in
the relief of established obstruction due to contracted pelvis, but it should
not be employed to anticipate obstruction; that is, not in the planned
management of labour complicated by disproportion. If its use is limited in
this way, unnecessary symphysiotomies will be avoided.
It should be confined to patients who have had no previous obstetric care,
or at least to poor attenders at antenatal clinics and those who live far from
medical aid. These are the women who are particularly exposed to the risk
of ruptured scar in subsequent labours if a Caesarean section is performed.
Disproportion must be of moderate degree only: not more than half of
the fetal head should be still palpable above the pelvic brim. Asymmetry
of the pelvis or extreme degrees of contraction (true conjugate less than 8 cm.
or brim area less than 70 sq. cm.) or a very large fetus (one estimated to
weigh more than 9 lb. or 4 kg.) are contra-indications to symphysiotomy.
The cervix must be at least half dilated: the head will descend into the
pelvis after it has been enlarged and fit snugly into the cervix, which
will reflexly stimulate better contractions: full dilatation of the cervix and
expulsion of the fetus should then soon follow.
Symphysiotomy is contra-indicated if labour is obstructed by a breech,
brow or mento-posterior face presentation. In the opinion of the writer,
a “crash” symphysiotomy has no place in the delivery of an arrested
aftercoming head.
Symphysiotomy should not be performed on a woman with a pre-

i

?

Fibrocarti’oge excavated
jto show hyaline cartilage

-Arcuate ligan-tent
-Perineal membrane
-Urethra
-Vagina

Fig. i 1.7.

Anatomy of the symphysis pubis. (From Crichton, D. and
Seedat, E. K. (1963). S’. Afr. med. J., 37, 227).

t

[

I

■

OPERATIONS TO RELIEVE OBSTRUCTION

-

191

XW'-ymring locomotor disturbance (such as hip joint disease), or gross obesity.
Minor degrees of sacro-iliac instability after the operation may be crippling
when superimposed on these conditions.
In a patient who has been sectioned before, symphysiotomy should not be
used to avoid repeating the operation. A scarred uterus should not be
expected to withstand the extra strain required to overcome disproportion.
Lastly, the fetus must of course be alive. If it is dead, craniotomy and
extraction is obviously the proper treatment.
Points of technique. The subcutaneous closed technique is recom­
mended. The operation is performed in the lithotomy position after
infiltration with local anaesthetic. The legs are supported by assistants to
limit abduction of the thighs to an intervening angle of 80 degrees, to prevent
excessive separation of the symphysis.
Before dividing the joint, a catheter is placed in the urethra. The plate of
fibro-cartilage in the middle of the joint (see Fig. 11.7) is identified with a
needle, which is left in position as a guide. A stab incision is then made with
a strong, solid-bladed scalpel through the central plate, 0 • 5 cm. below the
upper border of the symphysis. This incision is extended downwards to
divide the lower half of the symphysis and arcuate ligament, stopping short
of the perineal membrane, and then upwards to complete the transection of
the joint. The fingers of the left hand in the vagina displace the urethra to
one side to prevent it from being injured, and warn when the knife is
penetrating too deeply through the joint.
The incision must be made strictly in the midline. This avoids injury
to the hyaline cartilage covering the articular surfaces of the pubic bones
(see Fig. 11.7), and thus limits post-operative pain in the symphysis
and reduces the risk of osteitis: it also considerably reduces the amount of
bleeding.
After the joint has been divided, the two sides will be felt to separate.
This separation should be limited to 2-5 cm., which corresponds to a gap
which will just take the thumb. This enlarges the capacity of the pelvis by up
to 25 per cent, the main expansion, of course, being in the transverse
diameter. Wider separation than this will strain the sacro-iliac joints and
endanger the perineal membrane: if this is damaged, disturbance of the
control of micturition will result.
The head usually starts to descend as soon as the symphysis separates.
At this point, it is essential to perform a wide episiotomy in all cases. This
prevents excessive stretching of the anterior vaginal wall: without it,
vestibular tears are common, sometimes involving the urethra.
Descent of the head must be controlled, as anterior lacerations will be
produced if it is too rapid. It is wise to adduct the thighs during delivery of
the head to prevent excessive separation of the symphysis. The thighs should

SI

OBSTRUCTED LABOUR

leverage can be exerted which may cause severe damage.
If spontaneous deliver}7 does not occur very soon after the symphysis has
been divided, the vacuum extractor should be applied to the fetal head,
which is then delivered by controlled traction downwards and backwards.
Obstetric forceps should not be used, and rotation with Kielland's forceps is
especially dangerous. The forceps blades increase the distension of the
vagina and thus the risk of anterior lacerations. Rotation with forceps is
particularly liable to damage the unsupported urethra and bladder neck.
After delivery, the divided symphysis is compressed for some minutes
between the fingers in the vagina and the thumb in front of the joint to
expel clot and arrest venous oozing. When the third stage of labour is
complete, the episiotomy and any lacerations are repaired, and the skin
incision over the symphysis is closed with a single stitch. A Foley’s catheter
with a 5 ml. bulb is inserted for continuous drainage of the bladder, the
legs are adducted and lowered and the knees are strapped together tem­
porarily for 24 hours.
Aftercare. The patient is kept in bed for 3 days and nursed on her side: at
the end of this period the catheter is removed and the patient is allowed to sit
out of bed. Ambulation begins on the fifth day, with the aid of rubber-tipped
walking sticks. If the patient is very unstable initially, a broad belt encircling
the great trochanters is helpful, although not necessary as a rule. The patient
is usually walking well with minimal symptoms by the tenth day, but undue
muscular effort and lifting weights should be avoided for at least a further
month.
Complications. Sepsis and haematomas at the site of operation and
injuries of the urethra and bladder neck should be very rare if the operation
is limited to suitable cases and the correct technique is employed.
Stress incontinence is a fairly common sequel to symphysiotomy, but
full control of micturition is usually rapidly regained without specific
treatment.
Locomotor disturbances are particularly important in communities where
the women do strenuous work on the farms or carry heavy weights on their
heads. Laxity of the sacro-iliac joints usually results from excessive
separation of the symphysis during delivery’, and should be avoided by
following the rules laid down above. However, pubic pain and backache
may persist for a long time, and constitute the main objection to the
operation.
Pregnancy and labour after symphysiotomy. The divided joint heals
by fibrous union: along with the pelvic ligaments, this softens during
subsequent pregnancies, which may lead to pelvic instability. As a result,
a woman who has previously had no symptoms may suffer from pubic pain

?
1

I

I

f

s

CHAPTER TWELVE

Sequelae of Obstructed Labour
J. B. Lawson

Rupture of the uterus
In the previous chapter the response of the uterus to obstructed labour was
described. If the obstruction is not relieved, the overstretched lower se<mient
will eventually rupture spontaneously. Rupture may also be induced by
oxytocics or by operations for the relief of obstruction which farther distend
the vulnerable lower segment.
The liability of the uterus to rupture in obstructed labour is influenced b v
fnf17 S°7e extent-.SPontaneous rupture is very rare in a first labour,
although it does occasionally occur after very prolonged obstruction The
grande mulupara, on the other hand, is at particular risk because her uterus
is weakened by an increased proportion of fibrous tissue. However, the peak
incidence of spontaneous rupture is in third and fourth labours: presumably
SjC°nd
the Progressive increase in
birth weight is likely to produce a fetus large enough to cause obstruction
when the pelvis is contracted.
Spontaneous and induced ruptures of the uterus following obstructed
labour all start in the lower segment, although they may extend into the
upper segment It is customary to distinguish between complete and incomp e e ruptures, depending on whether the peritoneal coat is involved This is
unimportant, however, as the treatment is the same.
The anterior wall is most commonly involved, in which case the direction
whJh tea?S UuUaiU r tra"sv"se> often with a vertical extension at one end
ch makes the defect L-shaped: sometimes the neighbouring bladder is
also tom. Longitudinal ruptures of the lateral waU of the lower segment are
quite common and may extend up to the fundus or downwards into the
vagma. These lateral tears open up the broad ligament and may involve the
uterme artery or its mam branches. Posterior ruptures are least common
and are usually transverse. Occasionally multiple ruptures occur, in which
case the uterus may be almost detached.

Diagnosis
Dramatic symptoms are unusual at the moment of spontaneous rupture

204

SEQUELAE OF OBSTRUCTED LABOUR

period of slow deterioration. The appearance of blood at the vulva when
labour is obstructed usually indicates that the uterus is rupturing. Blood in
the uterus may not be able to escape if the presenting part is firmly impacted,
however, so this important sign is not always present. If palpation of the
lower segment produces severe pain, this is very significant, and it is later
followed by generalized tenderness all over the abdomen as blood and
liquor leak into the peritoneal cavity. The contour of the uterus may alter
when part of the fetus begins to extrude through the rent.
In view of the difficulty of diagnosing rupture of the uterus in its early
stages, it is wise when in doubt to anaesthetize the patient in an operating
theatre prepared for both operative vaginal delivery and for laparotomy.
Further examination may then confirm that rupture has occurred, or the
tear may be felt during extraction of the fetus. If so, vaginal manipulations
must be abandoned immediately and the abdomen opened, as otherwise
extension of the rent and removal of the tamponading effect of the fetus will
increase the blood loss.
In patients seen at a later stage, when the fetus has already been extruded
into the peritoneal cavity, clear physical signs usually make the diagnosis
obvious, although the patient often experiences a temporary’ feeling of relief.
The tense uterus no longer fills the centre of the abdomen, which becomes
flattened, with generalized distension filling out the flanks. Tenderness all
over the abdomen becomes more marked, and a fluid thrill and shifting
dullness can often be elicited. The fetal limbs usually become abnormaUy
easy to feel, and the presenting part may move away from the pelvic brim
into which it was previously firmly impacted. The empty uterus can some­
times be felt separate from the fetus either in front of it or to one side,
although it may be concealed behind the fetus.
Catheterization of the bladder is usually easy as the urethra is no longer
compressed by the impacted presenting part. The urine is nearly always
blood-stained: if copious bright red blood escapes through the catheter the
rupture has probably involved the bladder. Vaginal examination may
confirm that the uterus is empty, but more commonly the lower pole
remains in the uterus after the rest of the fetus has escaped through the rent.
In these circumstances, gentle upward pressure on the presenting part
shows it to be surprisingly mobile, and displacement releases a gush of blood.
The placenta usuaUy follows the fetus into the peritoneal cavity, but

si

§
4

I

«

RUPTURE OF THE UTERUS

205

^■■^occasionally it is expelled per vaginam. In an obstructed labour, the appear­
ance of the placenta at the vulva when the fetus is still undelivered
indicates rupture of the uterus.
Sometimes the rupture may not be discovered until after the fetus has
been removed vaginally, particularly if it has been induced by an obstetric
manoeuvre. The importance of digital exploration of the lower segment
immediately after all vaginal operations to relieve obstructed labour has
already been stressed.
The diagnosis is easily missed in cases seen for the first time an hour or
two after deliver}-, as vaginal bleeding and shock may be attributed to
haemorrhage from the placental site. If a patient who is in shock after a
difficult delivery of a stillbirth does not immediately respond to liberal
blood transfusion, rupture of the uterus should be suspected, particularly if
the placenta is still retained.
Occasionally patients are brought to hospital many hours or even days
after a delivery during which the uterus has ruptured. Abdominal distension
develops very quickly due to reflex ileus, and a tender broad ligament
haematoma may be felt beside the uterus: this is commonly accompanied by
retention of urine. In later cases, general peritonitis may be already estab­
lished, and sometimes loops of intestine are found prolapsed into the vagina
through the rent in the uterus, which carries a grave prognosis.
The differential diagnosis of rupture of the uterus is not as a rule difficult.
However, it should not be forgotten that collapse during an obstructed labour
does not necessarily mean that the uterus has ruptured, as circulatory
failure may be the sequel of intrapartum sepsis, dehydration and electrolyte
imbalance. Massive concealed haemorrhage due to abruptio placentae, in
which there is severe shock coupled with a tense, tender uterus, absent
fetal heart sounds and vaginal bleeding, is not preceded by prolonged
labour, but the physical signs may lead to misdiagnosis if the patient is too
ill to give a history. However, vaginal examination is conclusive, as in
abruptio the cervix is usually closed or only slightly open. A full-term
extra-uterine pregnancy can be mistaken for a fetus which has been extruded
through a uterine rupture, although again the history and vaginal findings
are distinctive.
It is much better to diagnose rupture of the uterus too readily than to miss
it, as any delay gravely impairs the patient’s chances of survival.
Management
Attempts to repair uterine ruptures per vaginam are ineffective and
dangerous, except occasionally when a cervical tear has extended an inch or
two into the lower segment. Packing the rent from below usually does more
harm than good; with the whole hand in the vagina, however, massive

I
206

if

-r-:

SEQUELAE OF OBSTRUCTED LABOUR

bleeding after an induced rupture can sometimes be temporarily arrested by
pressing a rolled-up perineal pad through the rent firmly against the pelvic
sidewall while preparations for laparotomy are completed.
Almost all ruptures of the uterus must be dealt with by the abdominal
route, and speed is vitally important: even if the patient’s condition may still
appear to be quite good, it will soon deteriorate. In the need for haste
resuscitation must never be neglected: superimposed on the effects of the
preceding labour, shock from blood loss may otherwise be fatal.
An intravenous drip must be set up immediately in all cases, starting
with 5 per cent glucose. A large vein in the arm should be chosen to
ensure a rapid flow, and if there is any difficulty in inserting a wide-bore
needle, a cannula should be introduced by the cut-down technique without
hesitation. Meanwhile, the patient’s blood group is determined, and if
stored blood is available the first bottles are given very rapidly after
the quickest possible cross-matching has been done {see Appendix I).
Since massive transfusion is almost always required, at least five bottles
should be prepared, particularly as hypofibrinogenaemia occasionally
increases the blood loss. If blood is not available, the circulation may
have to be maintained with plasma but this, of course, is much less
effective.
It is not often possible to complete the resuscitation of the patient before
beginning the operation, as the rise of blood pressure increases the bleeding,
but at least the laparotomy should not be started until the patient’s condition
is improving. However—usually when insufficient blood is available—the
patient occasionally remains so deeply shocked that her blood pressure
cannot be raised to recordable levels. In these circumstances it may be
necessary to operate with the circulatory failure uncorrected, a forlorn
hope which sometimes ends in success. An improvement often follows the
removal of the fetus from the peritoneal cavity, and with the abdomen
open 1 or 2 pints can be transfused under pressure direct into the common
iliac artery, which usually has a dramatic effect.
Resuscitation must be continued energetically during the operation. The
aim should be to maintain the systolic blood pressure above 100 mm.
throughout, so that unsecured bleeding points are revealed and dealt with
before the abdomen is closed, which prevents reactionary’ haemorrhage
later. The importance of starting the treatment of sepsis as soon as
possible with intravenous antibiotics should not be forgotten.
Surgical technique. As soon as the abdomen is open, if the patient is
still undelivered the fetus and placenta are removed. If they are free in the
peritoneal cavity this offers no difficulty, but if only part of the fetus is
extruding through the rent this should be extended towards the midline to
permit removal of the fetus and placenta through it. In posterior ruptures, it

31

p
i
=i

I
I

-t
RUPTURE OF THE UTERUS

207

is usually necessary to make a transverse incision through the anterior wall
of the lower segment to evacuate the uterus.
When the uterus is empty, it should be eventrated and carefully inspected,
including the posterior wall where there may be an unsuspected second
rupture. Any obvious arterial spurters are then tied off, and oozing from the
depths of the broad ligament is temporarily arrested by pressure with a hot
pack. The rent in the uterus must now be dealt with.
The aim should be to secure haemostasis and close off the infected birth
canal from the peritoneal cavity as quickly as possible, producing the min­
imum shock in the process. As a rule, repairing the rent is easier and quicker
than hysterectomy, except when the rupture is very extensive. The decision
to repair the uterus should not be influenced by a desire to conserve child­
bearing function or menstruation, nor should hysterectomy be performed
just because it removes a damaged and infected organ. The correct procedure
in each individual case is the one which is shortest and produces the least shock,
and thus gets the patient off the operating table in the best condition.
When repairing the uterus, no attempt should be made to excise the edges
of the tear or to repair the defect in layers to produce a sound scar. The
irregularity of the rent and the widespread infarction and softening of the
muscle makes this impossible when the rupture has followed obstructed
labour (although it is often practicable when repairing a ruptured Caesarean
section scar). The defect should therefore be rapidly closed with a continuous
haemostatic suture which takes large bites through the full thickness of the
muscle.
When repairing anterior transverse ruptures, particular care must be
taken to avoid penetrating the posterior wall of the bladder and including it
in the sutures. It is therefore wise first to reflect the bladder downwards off
the inferior flap of the lower segment, as in a Caesarean section, so that it is
well out of the way.
Lateral ruptures are more difficult to deal with than anterior ones because
efforts to control associated haemorrhage from the depths of the broad
ligament endanger the ureter. After opening the peritoneal roof of the
broad ligament widely to give access, visible bleeding points should be
tied off individually. Often there is also a general ooze which has to be dealt
with by under-running stitches. These may catch the ureter, particularly if
they include the posterior leaf to which it is related. To prevent damage to
the ureter in these cases, it should be identified on the pelvic sidewall and
traced down, if necessary passing a tape under it with which to draw it out of
the way.
The repair of a lateral longitudinal rupture should start at the apex of the
tear and work downwards: traction on the running suture will then bring the
depths of the tear into view. The repair should stop short of the inferior

208

SEQUELAE OF OBSTRUCTED LABOUR

margin, to leave a defect through which a strip of corrugated rubber is
passed into the vagina to drain the cavity in the broad ligament. If there is
still much oozing here, before closing the visceral peritoneum the cavity can
be packed with gauze bandage which is brought out into the vagina instead
of the rubber drain.
Because the scar of the repair is very likely to rupture in a subsequent
pregnancy, particularly if labour becomes obstructed again, the Fallopian
tubes should be tied before closing the abdomen. This can be done without
appreciably lengthening the operation.
The alternative surgical management, hysterectomy, is simple when it is
confined to cases where the uterus has been almost detached by very ex­
tensive damage to the lower segment. It is also usually indicated in posterior
ruptures (which are difficult to repair), especially if the fetus has had to be
removed through an anterior transverse incision, w’hich will leave only the
lateral walls intact.
No attempt is made to remove the cervix. The tubes and ovaries are
conserved in the usual manner and, after securing the vascular bundles on
either side of the isthmus, the part of the lower segment which remains
intact is divided between clamps and oversewn. Having secured haemostasis
in the broad ligaments in the manner already described, a continuous suture
across the pelvis closes both the visceral peritoneum and the stump of the
uterus in a single layer.
Occasionally, the bladder is also torn when the anterior wall of the lower
segment is ruptured. Handling the bruised and haemorrhagic tissues very
gently, the bladder should be dissected down off the lower segment, taking
care not to extend the rent in the bladder wall. When its margins are
mobilized and clearly exposed, the defect is accurately repaired in two layers.
The bladder should subsequently be drained with an indwelling catheter for
14 days. If the patient is in poor condition and the need to complete the
operation is therefore urgent, it may be necessary to close the bladder
quickly over a wide-bore suprapubic tube. A vesico-uterine fistula is quite
likely to follow, which will have to be dealt with later {see Chapter 29).
There appears to be little advantage in draining the peritoneal cavity after
dealing with a ruptured uterus. Peritonitis is almost inevitable, but is
usually controlled by systemic antibiotics. Localized collections of pus can
be drained later as they occur, although they are surprisingly infrequent.
Extraperitoneal drainage into the vaginal vault is a different matter, and the
method described above is strongly advocated for all cases in which the
broad ligaments have been opened up. Otherwise, a pelvic abscess from an
infected haematoma is very likely.
After-treatment. In the first few hours after operation it is most
important to complete the correction of shock by liberal blood transfusion.

r

*

?
2

Ruriunc vr int uitKLd

209

^H
hb^ ■
9H|B>When the blood volume has been made up, the keto-acidosis resulting from
the preceding obstructed labour must be dealt with along the lines described
in Chapter 11, whilst keeping a careful watch for circulator}- overload and
pulmonary’ oedema.
Sepsis must be energetically treated from the outset by broad-spectrum
antibiotics by the intravenous or intramuscular route. On the assumption
that peritonitis is certain to develop, it is wise to pass a Ryle’s tube soon after
recovery from the anaesthetic, so that continuous gastric aspiration can be
begun immediately. This is maintained for at least the first 48 hours, with
appropriate intravenous replacement of fluid and electrolytes, until the
bowel sounds return and the aspirate is normal. The gastric tube can then be
removed and oral feeding cautiously begun.

Prognosis
The prospects for survival depend firstly on the duration of the interval
between the rupture of the uterus and the start of effective treatment, which
emphasizes the importance of early diagnosis. Secondly, the speed and
effectiveness of resuscitation is decisive, particularly the provision of large
quantities of blood for transfusion. Thirdly, competent surgery and
anaesthesia are important, although less so than the first two factors
mentioned. Fourthly, energetic after-treatment will preserve the life of
patients who have survived the initial surgical phase: skilled nursing plays a
large part in this.
An analysis of 91 consecutive cases treated in Ibadan between 1953 and
1959, 37 of whom died, emphasizes these points. Of 11 patients who died
without surgical treatment, in 3 the rupture was not diagnosed until after
death and 8 others could not be made fit for operation. Eighteen patients
who were operated upon died within 24 hours, 8 of them on the table:
incomplete correction of shock was the main factor in these deaths. Eight
more patients died between 24 and 72 hours after operation, of general
peritonitis. There were no deaths from reactionary haemorrhage or pul­
monary complications. None of those who were still alive after 72 hours
died subsequently, although their convalescence was often very stormy.
Seventy-two per cent of the patients survived in the last two years of the
period reviewed, compared with only 56 per cent in the first five years. The
improvement was largely due to much more stored blood being available.
More recent experience has shown that the salvage can be raised to about
90 per cent, but patients who are already moribund when admitted make
further improvement beyond this figure unlikely.
As far as remote prognosis is concerned, intestinal obstruction from
adhesions is the main hazard. If those treated by repair of the rupture are not
sterilized, however, there is a grave danger of death from rupture in a

. -,- ^,^"-7*s

210

SEQUELAE OF OBSTRUCTED LABOUR

subsequent pregnancy or labour. The forceful statement of Etienne
Tarnier at the Paris Congress of Obstetrics in 1897 cannot be improved
upon: “If a woman in the battle to reproduce her race has ruptured her
uterus, she should be invalided from the service, for it is not with cripples
that an army takes the field.”
Suggested further reading
boulle, P. and CRICHTON, D. (i 964). Rupture of the unscarred uterus. Lancet, i, 360.
Krishna MENON, M. K. (1962). Rupture of the Uterus. J. Obstet. Gynaec. Brit. Emp.,

£

HASPELS, A. A. (1961). Uterine rupture in Central Java. Oosterbaan and Le Cointre
N. V., Goes, Netherlands.
MOIR, j. c. (1964) Munro Kerr's Operative Obstetrics. 7th Edition. Bailliere, Tindall
7.
and Cox, London. Chapter 32.
J
Eastman, N. j. and hellman, L. m. (1966) Williams' Obstetrics. (13th Edition).
:j
Appleton-Century-Crofts, New York. Chapter 34.

LATER SEQUELAE OF OBSTRUCTED LABOUR

£

Lower genital tract injuries
LTnrelieved obstructed labour may cause ver}’ severe damage to the lower
genital tract. This results from necrosis of the soft tissues subjected to ;
prolonged compression between the maternal pelvis and the presenting part
of the fetus. Sepsis increases the area and depth of the sloughing which
follows. In many cases, the full thickness of the vaginal wall is involved, i
Sometimes the whole vagina may be sloughed piecemeal, and ver}' occasion- *
ally it comes away as a complete cast.
1
Necrosis deep to the anterior vaginal wall may involve the posterior wall f
of the bladder and urethra, in which case a vesico-vaginal fistula will^
result when the sloughs separate (see Fig. 12.i Plate 2). This usually j
occurs between the second and tenth day after delivery, but occasionally 3
a fistula may be already present at the time of delivery if labour has been 4
exceptionally prolonged. Posteriorly, sloughing may result in a recto-vaginal 1
fistula. In severe cases, the whole of the recto-vaginal septum and the perineum may be destroyed. In these circumstances, the unfortunate sufferer is 4
totally incontinent of urine and faeces through a gaping cloaca.
1
Although continuous drainage of the bladder after an obstructed labour 1
may prevent some fistulae from forming (see Chapter 29), once full thickness
sloughing has occurred, urine tends to drain through the defect rather than 3
through the catheter unless the fistula is ver}’ small. It is sometimes possible
to drain all the urine through the catheter by applying continuous suction
with an electric pump: this not only keeps the patient dr}' but also encourages

British Journal of Obstetrics and Gynaecology
1985, Suppl. 5. pp. 100-115

12. Maternal mortality
In collaboration with C. E. ROSSITER

Summary. There were 238 maternal deaths. Five deaths occurred after
delivery among booked women who had no antenatal complications
(0-4 deaths per 1000 deliveries); 14 deaths were among booked women
who developed complications during pregnancy (3-7 per 1000); and 219
deaths were in the emergency admissions (28-6 per 1000). Bacterial
infections, eclampsia, anaemia, haemorrhage and disproportion
together with its consequences, were the leading causes. The principal
high-risk factors were lack of antenatal care, early teenage pregnancy,
high parity and high child mortality rate from previous births. In the
emergency admissions the operative delivery rate was 25% in the
women who survived and 49% in those who died. In severe eclampsia
and in neglected obstructed labour, a high haematocrit (^0-45) and, to
a lesser extent, a low haematocrit (^0-14) were of ominous significance,
mortality rate being 25-60% in such cases compared w ith <10% in most
other obstetric complications. Measures to reduce maternal mortality
should aim to lower the proportion of high-risk w omen (40% at present)
and also make it possible for operative deliveries, especially caesarean
section, to be performed as soon as the need arises.
Since the early sixties, attention has increasingly
been drawn to the high maternal mortality rates
of around 10 deaths per 1000 deliveries in many
areas in West Africa (Lawson 1962; Waboso
1973; Ojo & Savage 1974; Megafu 1975; Caffrey
1979; Ademowore 1980; Okojie 1980). Cannon
& Hartfield (1964). then working under difficult
conditions in a rural and semi-urban environ­
ment in Western Nigeria, showed how improved
standards of health care, together with better
organization of basic health services, helped to
bring about a reduction in both maternal and
fetal mortality and morbidity. Despite the suc­
cess which Cannon & Hartfield (1964) achieved,
they were quick to point out that wherever socio­
economic deprivation exists on a large scale and
maternal mortality rates are high, the enormous
loss of maternal lives cannot be brought under
control permanently until living conditions for
the general population are improved.
This chapter presents the cold facts about the
maternal health situation in the Zaria area and at
the same time it attempts to bring to people’s
notice the underlying human tragedy and suffer­
ing. Knowing that the principal conditions
associated with maternal death are bacterial
100

infections, anaemia, eclampsia, haemorrhage,
disproportion and its consequences, our
material was further explored to throw up addi­
tional factors which often combined with the
obstetric complications in worsening maternal
prognosis.
Decision on avoidable factors in maternal death

At the time this material was being collected, it
was felt that in the case of the maternal deaths,
further information about them was necessary.
Therefore, the circumstances surrounding each
maternal death were an important part of the
discussions at weekly meetings attended by all
medical and nursing staff of the Department of
Obstetrics and Gynaecology. A consensus was
reached as to the clinical causes of death and
what avoidable factors there might have been.
Such avoidable circumstances were categorized
as predominantly due to the patient, the
environment, cultural beliefs or to defects in the
health services.
An important patient factor was the failure of
the women with high-risk pregnancies to report
for antenatal care when they should have done

101

5

■

so. Another was the fact that some people did
not readily submit themselves to emergency
operations, preferring instead to withhold con­
sent for operations for as long as possible. For
example, rather than give consent for an emer­
gency caesarean section, some of those admitted
in early labour with disproportion took their
own discharge at that stage, only to return to us
later, still undelivered but gravely ill. Cultural
practices detrimental to maternal health have
already been described (chapter 2). Environ­
mental factors were those connected with diffi­
culties in transportation experienced by women
who lived in remote rural areas. In doubtful
cases, confirmation was sought from medical
social workers and such information enabled us
to determine the distance that the woman who
died had had to travel to reach the Zaria hos­
pital. Errors of clinical judgement and problems
with shortages of, or interruptions to, the supply
of piped water, electric power, telecommuni­
cations systems, fuel, drugs, linen and other
items of essential equipment, all came under
avoidable factors within the hospital health
services.

The death rate was highest in girls aged ^15
years, falling to the lowest level between the
ages of 20 and 24 years, then rising again to
another peak in the oldest women aged ^30
years.
Considering just those women who did not
receive antenatal care, maternal mortality was
still significantly related to both ethnic group
(X; = 9-72, d.f. = 1, P<0 01) and religion
(X: = 7-87. d.f. = 1. P<0 01) after allowing for
maternal age. It is impossible statistically to
determine which of these two factors—ethnic
group and religion—predicts maternal mortality
better, as they are so closely related to each
other. Among the unbooked, education,
address and parity did not contribute much more
to the prediction of maternal mortality after
allowing for the other variables, but of course
educated mothers rarely failed to receive ante­
natal care.
Because the distribution of maternal charac­
teristics differed between the booked and
unbooked groups (Appendices 12.1 and 12.2),
maternal deaths in the two groups will be con­
sidered separately.

Preliminary statistical analysis and social
factors
There were 238 maternal deaths (10-5 deaths per
1000 deliveries) of whom 49 died undelivered
and 68 gave birth at home before their arrival at
hospital. Throughout this text, maternal mor­
tality refers to deaths during pregnancy and up
to 42 days afterwards.
Appendices 12.1 and 12.2 show the relation
between maternal mortality and antenatal care,
address, religion, ethnic group, education, age
and parity.The most striking observation is that
219 of the 7654 women who did not receive
antenatal care died (28-6 deaths per 1000
deliveries) compared with only 19 of the 15 020
booked delivered women (1-3 deaths per 1000
deliveries).
Of course many women who did not have
antenatal care never came to the hospital and
were delivered at home. Their number is
unknown but must have been at least two to
three times the number admitted as an emer­
gency. Even so, the perinatal and maternal mor­
tality among the excluded population of
unbooked women is still very much higher than
those who received antenatal care.
Another striking observation is that maternal
mortality is greatly affected by maternal age.

Maternal mortality in booked patients

(

There were 19 deaths in the booked group and
five of them had no antenatal complications
whatsoever and all died after delivery. Two
of these five had been discharged within 24 h of
normal delivery only to be readmitted 10 days
later, when they both died within the first hour of
admission, one from severe bronchopneumonia
and the other from cerebrovascular accident due
to severe postpartum hypertension. Of the other
three, one was an anaesthetic death during the
repair of a third degree perineal tear; another
had severe postpartum haemorrhage and died
within minutes before blood for transfusion was
available; the third patient in this group had
torrential postpartum haemorrhage successfully
arrested by hysterectomy, only to succumb to
septicaemia later.
Two further deaths were related to HbSS with
one dying from pseudotoxaemia (Hendrickse et
al. 1972) after puerperal sterilization by tubal
ligation. The other had active tuberculosis early
in the antenatal period. Good progress was
being made on drug therapy until pre-eclampsia
with sickle cell painful crises developed at 30
weeks, followed 2 weeks later by fatal pericar­

ditis. Being HbSS. her haematocrit, despite mul­
tiple transfusions, was still only 0-26.

102

Nine of the remaining 12 deaths were associ­
ated with the following complications:
eclampsia (6). post-caesarean section genital
sepsis andzor septicaemia (4) and post-caesarean
section bronchopneumonia (1). Of these nine
women, only one, an eclamptic, died
undelivered, two gave birth spontaneously and
the remaining six by caesarean section. Spon­
taneous delivery took place in two eclamptics,
one of whom had a twin pregnancy and placental
abruption. The indications for caesarean section
were severe eclampsia (2), severe eclampsia and
twin pregnancy (1), disproportion and intrapar­
tum pre-eclampsia (1). breech presentation and
contracted pelvis (1), transverse lie and uterine
fibroids (1). In all, six in this group of nine deaths
also had mild anaemia (haematocrit 0-24-0-29)
in early puerperium as a result of uncorrected
blood loss. As is so often the case case in West
Africa (Ekwempu 1980), eclampsia was
fulminating in five of the six deaths. All five had
already had eclamptic fits at home before their
arrival in hospital; there had been no warning
signs during their weekly antenatal visits.
The remaining three deaths were from acute
pyelonephritis in pregnancy (1), pyrexia of
unknown origin (1) and postpartum haemor­
rhage followed by septicaemia after a normal
delivery in a patient who had pre-eclampsia.
By the prevailing standards in Zaria, death
was reckoned to be unavoidable only in three of
the 19 booked deaths. In the rest, the avoidable
circumstances were early teenage marriage in
three (aged 13-15 years) and defects in our
health services in the rest. Of the 19 deaths. 17
were Zaria residents and the other two lived
within 50 km. Six had received formal educa­
tion, although none had had more than primary
education. Particularly noteworthy was the fact
that in the booked group the death rate was
much lower in the women who were healthy
throughout pregnancy (five deaths in 11261
deliveries; 0-44 1000) than in those who devel­
oped antenatal complications (14 deaths in 3759
deliveries; 3-7/1000).
Maternal mortality in the unbooked women

Genera! considerations and social
characteristics

The unbooked women who died constituted the
crux of the problem. Because they were so many
(219) and because nearly all presented very late
for treatment, they made enormous demands on

our limited resources.
Their distribution according to ethnic group,
address, religion, age and parity is shown in
Appendices 12.1 and 12.2. All the Zaria resi­
dents who died and 82% of those whose homes
were outside Zaria (range of distances from
Zaria, 20-200 km) had access to good roads, by
which it is meant that their homes were within
2 km of an all-season road. None of the
unbooked women who died had received any
formal education.

Table 12.1. Major obstetric and medical complications
in 7654 unbooked women of whom 219 died
No. of deaths
Obstetric
and medical
complications

Obstetric
Genital infection
Eclampsia
Anaemia
In pregnancy
In puerperium
Haemorrhage
APH
PPH
Rupture of uterus
Disproportion
Prolonged labour
Retained placenta
Pre-eclampsia
Postpartum
hypertension
Advanced abdominal
pregnancy
Non induced
abortion
Medical or surgical
Sickle cell disease
Specific infections
Other infective
conditions
Surgical conditions
Malignancies
Other noninfective
Traditional surgery
Gishiri cut
Breast gangrene
Uvulectomy

One
Two
No. of abnormal or more
affected condition abnormal
only
women
conditions

609
390

26
18

28
24

917
988

11

50
48

544

2
2

3

177
1070
988
894
492

1
0
1

2
0

21
25
28
56
55
24
12

559

0

11

4

1

0

NK

0

2

5

0
14

1
4

0
3
1

22
0
0

53

0

7

88
7
12

1
0
0

7

391

31

73
NK
6

1
1

NK. Not known.
APH, antepartum haemorrhage; PPH, postpartum
haemorrhage.

103

K

There were avoidable factors related to the
patients themselves in 89% of the deaths, to the
health services in 53%, to cultural practices in
27% and to the environment in only 9%.
Obstetric and medical complications

i

I-

i

Table 12.1 lists the main obstetric and medical
complications found in the unbooked women,
including those who died. Those cases where
death occurred in the presence of a single
obstetric or medical complication were in the
minority. The more common pattern was death
from the combined effect of two or more
obstetric complications. For example, while
eclampsia was the sole obstetric disease present
in 18 deaths, it was accompanied by one other
complication in a further 15 deaths, and by two
or more other complications in nine. Anaemia
was due to the combined effects of malaria and
dietary deficiencies of iron and folic acid, some­
times to blood loss and occasionally to sickle cell
disease. Where severe anaemia was the only
cause of death, the haematocrit ranged from
0 05 to 019. Uterine rupture resulted chiefly
from disproportion and fetal malpresentations,
mostly transverse lie of the fetus.
The surgical conditions associated with mater­
nal deaths were intussusception, volvulus and
appendix abscess. Specific infections associated
with maternal deaths were typhoid (7), men­
ingitis (5), tetanus (4), hepatitis (1), and
amoebiasis (1). Under ‘other infective condi­
tions’ were 22 fatalities, mostly from pyrexia of
unknown origin. Under ‘other non-infective’
were deaths from peripartal cardiac disease (3),
severe bronchial asthma (1), renal failure of
unknown cause (1), circulatory overload during
intravenous infusion (1) and renal failure from
mismatched blood transfusion (1). There were
six malignant neoplasms: primary hepatoma (1),
malignant trophoblastic disease in remission (4),
and goitre (1). The only death was from respira­
tory obstruction caused by the huge goitre. ‘Tra­
ditional’ surgery describes three procedures.
One was the gishiri cut (chapter 2), followed by
fatal haemorrhage. Another was burns on the
breasts complicated by fatal gangrene and sep­
ticaemia. These breast injuries were sustained at
home, and followed a traditional form of treat­
ment for breast discomfort in early puerperium.
What happened was that a heated metal rod was

directly applied to the breasts while it was still
very hot. The last procedure was the excision of
the uvula by traditional healers complicated by

haemorrhage, respiratory infection and sep­
ticaemia. This operation, carried out under
unhygienic conditions, is for the treatment of
sore throat; the practice is widespread but
fatalities are apparently rare.
Another matter of interest was a multiple
pregnancy rate of one in 19 among the survivors
and one in 11 among the deaths. There were 379
twin, nine triplet and two quadruplet pregnan­
cies among the survivors and 14 twin and one
triplet pregnancies among the deaths.
The 46 women who died undelivered, were
associated with the following conditions:
eclampsia (9), severe anaemia with haematocrit
of 0 08-014 (8), uterine rupture (8), prolonged
labour (9), placenta praevia (1), placental
abruption (1), haemonhage from a gishiri cut
(2), meningitis (2), pre-eclampsia with anaemia,
haematocrit 0-15-0-19 (2), pyrexia of unknown
origin (5) and cardiac failure of unknown cause
(1).
Of the 1638 women who gave birth at home
before arrival in hospital 64 died. The number of
survivors and deaths for each complication in
this group of emergency admissions was as fol­
lows: retained placenta (804,23), postpartum
haemorrhage (181, 16), eclampsia (45, 3), rup­
ture of uterus (14,2), anaemia in puerperium
(530,27), genital sepsis (223,21), septicaemia
(5, 10), vesicovaginal fistula (10, 3) and second­
ary PPH (46, 1)
Knowing the principal obstetric and other
conditions associated with maternal deaths,
analysis was carried a stage further. Attempts
were made to determine the relation between
each obstetric complication and several other
important variables pertinent to local
conditions.
Time of death and duration of stay in hospital

Examined in this way, 30 (13-7%) of the deaths
took place within the first hour of arrival in hos­
pital, 47 (21-5%) within 1-6 h, 57 (26 0%)
within 7-24 h, 55 (25-1%) within 1-7 days and 30
died after more than a week in hospital. How­
ever, the pattern varied according to the nature
of the underlying obstetric complication. Over
25% of those who died from retained placenta
and postpartum haemorrhage but nearly 15% of
those with ruptured uterus, anaemia in preg­
nancy and prolonged labour, died within the first
hour in hospital. By contrast, in genital sepsis
and anaemia in the puerperium, most deaths
(65%) occurred towards the end of the first week

104

of the puerperium. As many as 72% of all who
died after embryotomy (13 out of 18) but only
19% of those who died after caesarean section
(five out of 27) did so within 24 h after these
procedures. This difference in prognosis reflects
the difference in severity of the underlying
obstetric complications rather than in the opera­
tions themselves (see below).
Events of labour and maternal mortality

In the group of emergency admissions, 90%
(6719) of the 7481 women who sun ived and 65%
(143) of the 219 who died were in labour when
they arrived at hospital for the first time; 25% of
the survivors but 44% of those that died were
already either in the second or in the third stage
of labour. The operative delivery rate for
singleton births among the sun ivors was 24-6%
(vaginal breech delivery 2-2%. forceps 3-5%.
caesarean section 14-6%, embryotomy 2-7%
and other deliveries 1-6%) compared with
53-2% for those who died (vaginal breech
delivery 1 -9%, forceps 9-5%. caesarean section
16 5%, embryotomy 11-4% and other deliveries
13-9%). The difference in distribution was very
significant (x2 = 188, d.f. =5. P<0-001). Par­
ticularly striking were the differences in the pro­
portions of women who required embryotomy
and ‘other’ deliveries between those who died
and those who survivied. The term ‘other
deliveries’ throughout this text refers to a total of
144 patients shown in Table 12.2. 113 of whom
were abdominal deliveries following uterine

rupture. Not shown in Table 12.2 are those
women in whom uterine rupture was detected
only after vaginal delivery, some in hospital (64)
and some at their homes (16). Also not shown
are eight others who died undelivered, with uter­
ine rupture present.
Maternal age, obstetric complications and
mortality
Fig. 12.1 shows that within each of the seven agegroups. the case fatality rate was always less than
1095: of all those affected by pre-eclampsia. The
same was true for disproportion and prolonged
labour. Case fatality rate exceeding 10%
occurred in the following groups: early teenage
girls with gross anaemia (haematocrit <014).
eclampsia and postpartum haemorrhage; the
over 30s with eclampsia; in all age-groups where
uterine rupture took place, that is multiparous
mothers aged ^17 years. In the 17-19 agegroup. where mortality associated with rupture
of the uterus was highest (37-5%), the number of
women affected (three deaths out of eight) was
small compared with 6-8 deaths out of 32-74
women affected in the age-groups 20-24. 25-29
and 30 years.

Modes of delivery, obstetric complications and
mortality
Maternal death rate varied according to the
mode of delivery and the nature of the obstetric
complications (Fig. 12.2). As expected, spon­
taneous delivery was associated with the lowest

Table 12.2. Mode of •other delivery' in 144 emergency admissions

No. of women
Diagnosis

Mode of delivery

Survivors

Deaths

Laparotomy, repair of ruptured uterus
Laparotomy, hysterectomy
Laparotomy followed by spontaneous delivery
Laparotomy followed by breech delivery with
perforation of aftercoming head
Placenta praevia intrauterine death Braxton Hicks bipolar podalic \ersion
followed by spontaneous delivery
Obstructed labour
Symphysiotomy
Hydrocephaly
Spontaneous delivery after aspiration
Advanced extra-uterine pregnancy Laparotomy and delivery
Transverse lie and intrauterine
death
Internal version and breech extraction
Constriction ring
Assisted vaginal delivery
Others
Not stated

79
14
1

13
7
0

1

0

6
9
2

0
2
0
1

Total

Rupture of uterus
Rupture of uterus
Suspected rupture of uterus
Suspected rupture of uterus

3
2
1

0

3

0

121

23

0

<

s

105

T

50

50

40

40

£

£
7‘ 6
E

s 30

30 -

<9
E

O

O

E
76 20
c

E
76 20
c

<6

<6
5
10

10

f

0

0u
<15

. t

-L

15

16

£

1-- 1 '' !

17-19 20-24 25-29 >30

Maternal age (years)

Fig. 12.1. Maternal age. obstetric complications and
maternal mortality. V Severe anaemia (haematocrit
<014); • pre-eclampsia; Oeclampsia; ▼ dispropor­
tion; O prolonged labour; PPH; ■ uterine rupture.

<•12. J—I—J—1__ I____1
<15

15

16

17-19 20-24 25-29 >30

Maternal age (years)

Fig. 12.3. Maternal age. mode of delivery and mater­
nal mortality. Ospontaneous delivery; •breech;
A forceps; A caesarean section; ■embryotomy;
other deliveries.

eclampsia were associated with the highest case
fatality rate of around 20%.

20

Maternal age, mode of delivery and mortality

£

Fig. 12.3 shows that spontaneous delivery,
vaginal breech delivery and caesarean section
were comparatively safe as they were all associ­
ated with a case fatality rate of under 8%. By
contrast, in early teenage girls treated by
embryotomy, in the over 30s delivered by
forceps (mostly in the presence of eclampsia)
and in those who underwent ‘other deliveries’,
irrespective of their age, case fatality rates
exceeded 10%.

E
O

E 10
76
c
®
<Q

s

0^?

,

___ L

/

J /

I_______ I
*■

//•

I

/

- 4-/
Mode of delivery

Fig. 12.2. Mode of delivery, obstetric complications
and maternal mortality. A APH; O pre-eclampsia; •
eclampsia; disproportion; ■ prolonged labour.
case fatality rates (<3%) except among the
eclamptics (mortality 11%). Caesarean section
performed for whatever reason was always com­
paratively safe (case fatality <5%). Both
embryotomy and other deliveries carried out in
the presence of prolonged obstructed labour and

Maternal haematocrit values and maternal
death

Although the aim was to know the venous
haematocrit value of every emergency on admis­
sion into the labour ward, operational reasons
(see discussion) permitted this in only 22% of
those who survived and in 55% of those who
died.
Haematocrit at delivery (or admission) and time
of death. In Fig. 12.4 the haematocrit values on
admission were all plotted against the interval
between admission to hospital and death. No
woman with haematocrit of 2s 0-40 died within
the first 6 h after admission, even though there

106
0.6

>

so, in women with PPH, maternal death rate was
as high as 40% when the haematocrit was 0-35—
0-39 compared with between 5% and 16% when
the haematocrit was <0-35.

0.5 -

■o

15
o
c 0.4
.0
'5>
tn

E

5

•O
O

c 0.3
o

E 0.2 -

o
o

tu

00

o

YT

I:

a

o

c

ol
j
15 0.1 -

5

0L

O

8

•o

-J------- 1------- 1--------1
1-6 7-24 1-7
8
Hours
Days
Duration of stay in hospital

Fig. 12.4. Maternal haematocrit at delivery and inter­
val between admission and death. O Born before
arrival: •delivery in hospital.

were among them seven with retained placenta.
21 with ruptured uterus, five with postpartum
haemorrhage and one with severe APH. Of
those with haematocrit <0-25 and who died
within the first 6 h of arrival in hospital. 10 out of
the 19 had already given birth at home before
arriving in hospital with acute blood loss from
retained placenta or postpartum haemorrhage
or both combined.

Maternal haematocrit at delivery, obstetric com­
plications and maternal death rate. Table 12.3
relates maternal mortality not only to the
haematocrit values at delivery but also to various
abnormal conditions. Two groups of women
were at the highest risk, those with dispropor­
tion. prolonged labour and rupture of uterus in
whom the haematocrit value was ^0-14, and
those with haematocrit values of >0-45, irre­
spective of the underlying disease. As expected,
where there had been haemorrhage, maternal
haematocrit reading rarely exceeded 0-39. Even

Blood transfusion, haematocrit values and
maternal mortality rates. Table 12.4 shows that
major operative deliveries had to be performed
in the presence of gross maternal anaemia. More
importantly, blood transfusion was associated
with improved maternal outcome when mater­
nal haematocrit at delivery was <0-20, but not
necessarily when the haematocrit was between
0-20 and 0-34, and quite definitely not when the
haematocrit was >0-35.

High maternal haematocrit and pregnancy out­
come. Eight of the 36 women admitted with a
haematocrit of ^0-45 died compared w ith death
rates of 4-9% in the other seven haematocrit
groups (Table 12-3). Table 12.5 compares the
survivors with the deaths among women with
high haematocrit values. Antepartum haemor­
rhage. pre-edampsia and retained placenta
occurred in the sunivors but not in those who
died. About half of those with eclampsia and
ruptured uterus died. All women delivered by
caesarean section survived and at the time the
operation was performed, the fetus was alive in
every case.

Major puerperal complications and maternal
deaths. Of all the complications in the puerperium. genital sepsis and anaemia (haematocrit
^0-29) were the most frequent, both in the sur­
vivors and in the fatalities. Of a total of 7481
survivors. 721 (10%) had anaemia. 339 (4%)
had genital sepsis. 216 (3%) had both and 6205
(83%) had neither anaemia nor genital sepsis.
Of the 219 deaths, 24 (11%) had anaemia, 27
(12%) had genital sepsis, 27 had both and 141
(65%) had neither. The difference in distribu­
tion between the sunivors and deaths was versignificant (P<0-01). The other major puer­
peral complications, together with the number
of sunivors and number of deaths, were as fol­
lows: wound infection (142.9); obstetric fistula
(64.12); obstetric palsy (27.2); secondary post­
partum haemorrhage (58.3); septicaemia
(12,25); and puerperal psychosis (32,2). As well
as having the above complications, nearly 62%
of the sunivors and just over 80% of the deaths
(90% in the case of vesicovaginal fistula and
palsy) had genital sepsis, anaemia, or both.

s

107
Table 12.3. Maternal mortality (%) by haematocrit at delivery and obstetric complications in the unbooked women

Maternal mortality (%) in relation to haematocrit at delivery

i-

Abnormal conditions

^014

General anaesthesia

3
(32)
0
(I)
33
(3)
38
(8)
33
(3)
3
(35)
7
(15)
6
(35)

3
(61)
17
(6)
0
(9)
0
(12)
0
(4)
10
(50)
5
(19)
2
(86)

10
(83)
14
(7)
21
(30)
20
(30)
67
(6)
3
(64)
16
(31)
6
(102)

8
(140)
9
(ID
10
(39)
12
(41)
8
(13)
9
(47)
9
(43)
4
(112)

8
(138)
23
(40)
8
(83)
10
(71)
8
(12)
7
(30)
15
(20)
2
(45)

8
(171)
18
(50)
8
(119)
10
(89)
18
(17)
17
(12)
40
(10)
10
(21)

12
(91)
25
(28)
10
(63)
15
(48)
21
(14)
0
(9)
0
(5)
0
(4)

23
(22)
50
(6)
29
(17)
25
(12)
57
(7)
0
(1)

No. of all women

122

185

241

331

346

346

170

36

No. of deaths

11

10

20

14

19

24

15

8

deaths

9

5

8

4

5

7

9

22

Eclampsia

i

k

Disproportion
Prolonged labour

Uterine rupture

!

Antepartum haemorrhage
Postpartum haemorrhage

Retained placenta

1
l

0 15-0-19 0-20-0-24 0-25-0-29 0-30-0-34 0-35-O-39 0-40-0-44

2J0-45

(0)
0
(3)

Results are percentages with the numbers of affected uomen shown in parentheses.

Table 12.4. Maternal mortality

) by mode of delivery and blood transfusion in the unbooked women-'

Spontaneous delivery
breech and forceps

Haematocrit
at
delivery
^014

0-15-0-19

0-20-0-24
0-25-0-29
0-30-0-34

0-35-0-39
0-40-0-44

2s0-45
Not recorded

Caesarean section

Other deliveries

Trans­
Not
fused transfused

Trans­
fused

Not
transfused

0
(3)
0
(6)
22
(9)
18
(17)
11
(19)
13
(23)
20
(15)
43
(7)
11
(119)

5-9
(118)
1- 8
(168)
70
(172)
2- 8
(176)
7- 6
(92)
8- 8
(91)
13-3
(45)
33-3
(12)
3- 2
(1092)

100
(1)
28-6
(14)
7-6
(66)
1-9
(154)
3- 6
(251)
4- 1
(246)
5- 6
(124)
12-5
(24)
0-6
(4822)

Trans­
fused

Not
transfused

Trans­
fused

6
(100)
1
(142)
6
(133)
1
(118)
2
(40)
18
(17)
25
(8)
33
(3)
3
(623)

100
(1)
29
(14)
6
(54)
2
(139)
2
(207)
3
(183)
6
(81)
13
(15)
0-5
(4339)

7
(15)
10
(20)
7
(30)
2
(41)
12
(33)
4
(M)

(22)
0
(2)
1
(350)

Not
transfused

(0)

(0)
18
(H)
0
(12)
3
(34)
2
(44)
3
(35)
0
(7)
0-5
(379)

(0)

(0)
0
(1)
0
(3)
30
(10)
21
(19)
13
(8)
50
(2)
3
(104)

'Excludes data on women who died undelivered.
Results are percentages with the numbers of affected women shown in parentheses.

All deliveries

108
Table 12.5. Maternal outcome and intrauterine fetal
death by obstetric complications and mode of delivery
(unbooked women with haematocrit of >0-45 only)

Factors

Total Survivors Deaths

Total no.
Obstetric complications

36

Antepartum haemorrhage 1
Pre-eclampsia
4
Postpartum haemorrhage 0
Retained placenta
3
Eclampsia
6
Disproportion
17
Prolonged labour
12
Uterine rupture
7
Genital sepsis
5
Treatments
General anaesthesia
22
Blood transfusion
12
Mode of delivery
Spontaneous delivery
12(6)
Vaginal breech
1 (1)
Forceps
5d)
Caesarean section
9(0)
Destructive operations
2(2)
Laparotomy for ruptured
uterus
7(6)

28

8

1
4
0
3
3

9
3
1

0
0
0
0
3
5
3
4
4

17
7

5
5

10(5)
0(0)
4(0)
9(0)
1 (1)

2(1)
1 (1)
1 (1)
0(0)
1 (1)

4(3)

3(3)

12

tually survived or died, the risk of puerperal
anaemia and genital sepsis was high: more than
40% of them had either anaemia or genital sepsis
or both combined. Following hospital delivery,
however, the prevalence of anaemia and genital
sepsis varied depending on the mode of delivery
between 4% and 43% in those who survived and
between 20% and 76% in those w-ho died.
Fetal outcome

Table 12.7 shows the fetal outcome in the
women who died and in those who survived:
fetal survival for hospital and home births was
similar in the surviving women (x2 = 4-8,
d.f. = 2, P - 0 09), whereas in the women who
died, the proportion of stillbirths in hospital
deliveries was 2-5 times greater than among
home births (x2 = 38, d.f. = 2, P<0 01). Table
12.8 shows that 38% of babies born to the
women who died were normal-birthweight
babies who had all died in utero before arrival at
the hospital. Only 8% of the babies of the surviv­
ing women were in that category. A total of 2317
babies, including those that were motherless,
were weighed on discharge from hospital; 39%
of the babies of the surviving mothers and 40%
of the motherless babies weighed 2-5 kg at that
time.

Results are numbers of women, with numbers of intra­
uterine fetal deaths on admission shown in
parentheses

Mode of delivery', genital sepsis, puerperal
anaemia and mortality
Table 12.6 shows that among those admitted to
hospital after home delivery, whether they even-

Past obstetric history and maternal mortality
Information about previous children was avail­
able from 7006 of the surviving women and 202
of those who died. The women who survived

Table 12.6. Puerperal anaemia, genital sepsis, mode of delivery and maternal deaths (unbooked women)
with anaemia, genital sepsis,
or both

No. of women
Mode of delivery

Home births*
Hospital births
Spontaneous delivery
Vaginal breech
Forceps
Caesarean section
Embryotomy
Other deliveries
Major operations
Repair of ruptured uterus
Hysterectomy
’Born before arrival at hospital

Survivors

Deaths

Survivors

Deaths

1574

64

40

52

4079
197
257

24
3
15
27
18
23

4
8

33
31
43

76
33
20
52
50
66

18
5

51
57

72
60

1049

203
121

113

30

14

I

f

109
Table 12.7. Maternal and fetal outcome by place of birth (unbooked women, singleton and multiple births
combined)

Survivors
Fetal outcome

F

I

Hospital births
Live birth
IUD on admission
IUD after admission
Neonatal death
Home births*
Live birth
Stillbirths
Neonatal death

Maternal deaths

Number

% of group

% of total

Number

% of group

% of total

4696
838
224
477

75-3
13 4
3-6
7-7

59-6
10 6
2-8
61

28
66
10
15

23-5
55-5
8-4
12 6

150
35-2
5-3
80

1226
266
152

74-6
16-2
9-2

15 6
3-4
1-9

47
16
5

69 1
23-5
7-4

25-1
8-5
2-7

‘Born before arrival at hospital.
IUD, Intrauterine death

Table 12.8. Fetal birthweight by fetal outcome in
babies of surviving women and in motherless babies
Fetal
birth­
weight
(kg) Fetal outcome
^2-5

>2-5

Total

Live
IUD on admission
IUD after
admission
Neonatal death

10

Babies of
surviving
women

Motherless
babies

$

759 (13-5)
286 (5-1)

9 (8-7)
16 (15-5)

75
t
o 5 E
75
c

79
169

(1-4)
4 (3-9)
(30)
9 (8-7)
Live
3584 (63-9) 16 (15-5)
IUD on admission 451 (80) 39 (37-9)
IUD after
admission
98 (1-7)
5 (4-9)
Neonatal death
187 (3-3)
5 (4-9)
5613 (100 0) 103 (100 0)

Results are numbers, with percentages in parentheses.
IUD, Intrauterine death

had a total of 15 788 children from previous
births (parity 0-16) and the women who died had
a total of 505 children (parity 0-13), giving a
mean number of 2*25 children per woman who
survived and 2-50 for those who died. However,
the proportion of surviving children was 71%
(1-61 per woman) for the surviving women and
only 58% (1-44 per woman) for the dead
women.
Fig. 12.5 and Table 12.9 present the propor­
tion of live children from previous births against
maternal mortality rate at each parity group.
Maternal mortality rate rose as child survival

£

to

2

0L
^50

>50
too
Child survival from
previous births (%)

Fig. 12.5. Child survival, previous births (01-2;
• 3-6; 7-9; ■ 10) and maternal mortality.

worsened, particularly in those with seven or
more previous births.
Discussion
The statistics reported here display a maternal
death rate which is high even by Third World
standards (Makokha 1980; Mtimavalye et al.
1980; Chi et al. 1981; MacPherson 1981). Too
many women died because they reported for
treatment when their illness had reached an
advanced stage. The nature of the obstetric com­
plications, the frequent resort to embryotomy,

110

Table 12.9. Child mortality, previous births and maternal death rate (unbooked women)

St of children alive from previous births

^50

>50

Maternal deaths

Parity
1-2
3-6
7-9
^10

Total
no.

n

506
509

15
9

192
116

10
10

KM)

Maternal deaths

%

Total
no.

n

30
1-8
5-2
8-6

599
311
84

20
7
2

the large proportion of women dying within the
first 6 h of arrival in hospital, and the high still­
birth rate among babies of normal birthweight,
all attest to this fact.
Data analysis on maternal mortality in Zaria is
compounded by there being more than one
obstetric or medical cause for most maternal
deaths. Faced with this problem, the practice in
many places is to ascribe death to one principal
or major cause, regarding all others as merely
secondary or contributory. This system of classi­
fication is difficult to justify in Zaria where, for
cultural and religious reasons, autopsies are not
performed. In presenting the principal diseases
or complications associated with maternal mor­
tality, the system adopted both here and in a
previous report (Harrison 1980) is one which
reflects all the major circumstances surrounding
each maternal death. In this way. the contri­
bution each obstetric complication made to the
overall spectrum of maternal death becomes
clearer, a point borne out by the analysis of the
relation between maternal haematocrit levels
and other obstetric complications (Table 12.3).
An intriguing finding is the association of high
maternal haematocrit level with poor maternal
prognosis in eclampsia. Koller (1982) argued
that by its effect in increasing blood viscosity and
hence causing placental infarcts, a high maternal
haemoglobin level is associated with intra­
uterine fetal death. In chapter 11, fetal outcome
was shown to be best when maternal
haematocrit level at delivery was between 0-30
and 0-40, and that at higher haematocrit levels,
the risk of low birthweight and perinatal death
increased. Apparently the relation between
maternal haematocrit level and pregnancy out­
come does not end there. According to this
report, there are at least two obstetric condi­

3-3
2-3
2-4

Maternal deaths

Total
no.

n

c/f

1055
654
57
6

20
16
0
0

1- 9
2- 4
00
00

tions—anaemia and eclampsia—where mater­
nal haematocrit level gives some indication of
the severity of the underlying condition. In
anaemia, maternal risk increases as the
haematocrit value drops, whereas in eclampsia,
mortality rises as the haematocrit level rises
above 0-40. A very high haematocrit level was
also a feature of prolonged obstructed labour,
especially when this culminated in uterine rup­
ture (Table 12.5). In this case, haemoconcentra­
tion from dehydration was the probable cause,
whereas in eclampsia, the mechanism for the rise
in haematocrit was volume depletion through
vasoconstriction and altered vascular per­
meability (Gallery et al. 1979). In either case,
resort to delivery by caesarean section was
apparently safer than any other form of delivery
(Table 12.5). But such a result was principally
due to selection. In women with less advanced
disease, the fetus was still alive, and so delivery
by caesarean section was favoured but in those
with more severe disease, the fetus was often
already dead in utero, and so the practice was to
avoid caesarean section.
There is no doubt that fetal and maternal sur­
vival are strongly linked. For example, in this
series, even among the women who failed to
receive antenatal care, the survivors achieved
better fetal results than those who died. This
connection between maternal and fetal outcome
was also evident when child survival rates from
previous births were compared between the sur­
vivors and those who died. Again, the survivors
had a marked advantage over those who died.
Indeed in the situation in Zaria, there is some
justification for regarding a high child mortality
rate by itself as a maternal risk factor, perhaps on
a par with high parity and early teenage preg­
nancy. Of course all these three reproductive

Ill

?

patterns are closely linked and they are indica­
tive of the serious defects in the living conditions
in the area.
As many as 80% of deaths occurred in women
who had reasonable access to health care as they
lived less than 2 km from a main road. This was
an unexpected finding and it carries wider impli­
cations. For a full appreciation of what these
implications are, it is necessary first of all to
recall that the illness from which the emergency
admissions died always started at their homes.
When they realized that expert intervention was
required and because road communication for
them was comparatively easy, they eventually
reached hospital where they died. For those liv­
ing in remote areas without access to hospital
(perhaps 60% of the population) the deaths
from complications needing operative treatment
are almost certainly greater than the hospital
statistics would suggest. On the other hand, it
can also be argued that in places such as Zaria,
where vital statistics for the whole area are not
kept, it is not possible to say with confidence
whether maternal mortality rates among the
rural dwellers were any higher than those
observed at the hospital. There is another side to
this argument. Most of those with homes outside
Zaria, whether they live close by or far from a
main road, belong to the same socioculturalethnic group (Murphy & Baba Tukur 1981).
Their attitudes to life are the same (Lister 1980)
and so it is very likely that the obstetric condi­
tions needing operative intervention, such as
obstructed labour and eclampsia, will arise just
as often, and that in the absence of effective
treatment, the mortality among the rural
dwellers in remote areas must be very high
indeed. Exactly how many perish in this way is
not known, but the fact that each year there were
nearly 300 fresh cases of vesicovaginal fistula
attending the gynaecological clinic in Zaria,
many from the rural background, gives some
indication of the scale of maternal morbidity and
mortality among the rural dwellers.
A recent report of 20-24 maternal pregnancyrelated deaths per 1000 women of childbearing
age in certain isolated villages in Gambia, West
Africa (Lamb et al. 1984) nearly match the
maternal mortality rate in the unbooked women
in our Zaria study, 29 per 1000, these rates are
between 2 and 20 times the rates commonly
reported from urban hospitals in developing
countries. It may well be that the relatively ‘low’
maternal mortality rates from urban centres

grossly underestimate the magnitude of the
problem at least in the poverty stricken parts of
West Africa. All this is very reminiscent of the
situation in 18th century rural England with its
mortality rates of 24-29 maternal deaths per
1000 baptisms (Dobbie 1982).
As in many societies, maternal mortality in
Zaria has now been shown to vary with maternal
age. Maternal mortality, high among the early
teenage mothers, fell to the lowest level between
the ages of 20 and 29 years, and rose again with
advancing age. This age effect on maternal sur­
vival was quite pronounced in eclampsia, severe
anaemia, prolonged labour, postpartum
haemorrhage, uterine rupture and even after
certain operative deliveries. In most cases,
except eclampsia in the over 30s, it was the teen­
age girl who came out worst of all. Exactly why
this is so is not clear from the analysis, but it may
be a consequence of one of the important tradi­
tional customs in the area. There is a strongly
held belief that home delivery is best for young
teenage girls. So it would seem likely that when
things went wrong during labour, people were
slow to seek expert help whenever teenage
expectant mothers were the ones affected.
Complications of early pregnancy featured
very little in these Zaria maternal deaths partly
because of the selective nature of this survey and
partly because of people’s attitude towards
health care. Although each year about 600
women with incomplete abortion and 90 with
ectopic gestation were being treated, the vast
majority, having been discharged early, were lost
to follow-up. Attitudes to child-bearing remain
very traditional among most people in Zaria.
Only a tiny minority practise effective con­
traception, and pregnancy termination—except
for medical reasons—is seldom resorted to.
Therefore, death from criminal abortion, which
is rampant in many Third World cities, did not
occur in Zaria during the period of this survey.
As many as 88% of those who reported to hos­
pital for antenatal care did so for the first time in
the second half of pregnancy. Therefore, in the
Zaria area, fatalities from early pregnancy com­
plications, such as spontaneous incomplete
abortion and molar pregnancy, might well have
taken place outside hospital without attracting
much notice.
Regarding the motherless babies, the results
presented here would suggest that while these
babies were in hospital, the support given to
them was probably no worse than that received

112

by the other babies. The proportion that
weighed ^2-50 kg on discharge was the same for
both the motherless babies and the rest. The
existing infant welfare clinic and a home for
motherless babies in Zaria both served to com­
plement whatever care these babies received at
their homes.
Attention must now be turned to the events in
the hospital itself and to some other results of
this study. These include the high prevalence of
bacterial infections, the circumstances sur­
rounding blood transfusion, and also the predic­
tive value of maternal haematocrit. Judging
from the maternal haematocrit values at
delivery, there is no doubt that many emergency
admissions were in poor health. They were
clearly high operative risks at the time they were
subjected to major operative procedures. Thir­
teen per cent of those who had general anaes­
thesia, 109t of those subjected to either repair of
ruptured uterus or hysterectomy and 307c with
retained placenta, some of whom subsequently
needed manual removal of the placenta, arrived
at hospital with venous haematocrit values
<0-20. Among the survivors were women
admitted in obstructed labour with haematocrit
of0 09,eclamptics with haematocrit of 0-13, and
those who, despite a haematocrit of only 0 07,
had to have manual removal of placenta carried
out shortly after admission.
Under these circumstances, blood transfusion
played a crucial role in the management of the
emergency admissions in Zaria, with as many as
247 of the survivors and 487 of those who died
receiving transfusion. However, there were
problems. Shortage of donor blood was not one
of them but delay in getting transfusion started
was. The major constraints were faults in the
transportation and in the telecommunication
systems and frequent interruptions to the elec­
tric power supply. Occasionally, the hospital ran
out of supplies of very basic items such as blood
giving sets, suitable serum needles and intra­
venous cannulae. Blood that was not cross­
matched was hardly ever transfused, although
the use of blood cross-matched for less than 2 h
was commonplace. In spite of these shortcom­
ings, there is no doubt that blood transfusion
proved life-saving when used to raise the circu­
lating haemoglobin level quickly in patients
found to be grossly anaemic during labour or
shortly afterwards (Harrison 1982). However,
there is no ready explanation as to why, when
maternal haematocrit was between 0-20 and

100

90
80

g 70•S 60 CD
t
O
E 50 c 40 a
to

5 30 20 -

10 0L
Ln1L——‘I —i____ I______ I
« 0.14
0.20-0.24
0.30-0.34
0.40-0 44
0 15-0.19
0.25-0.29
0.35-0.39
? 0.45

Maternal haematocrit at delivery

Fig. 12.6. Maternal haematocrit at delivery, blood
transfusion and maternal mortality. •Transfused:
C not transfused.

0-39, blood transfusion failed to show the same
sort of advantage as it did when the haematocrit
was <0-20. In the case of women with high
haematocrits (^0-45), the increased risks they
ran when they were transfused, as opposed to
when they were not transfused (Fig. 12.6) would
suggest that the use of blood transfusion was
inappropriate, or that such patients were being
over-transfused, or that insufficient attention
was being paid to coexisting biochemical distur­
bances. It may also be that those w ho were trans­
fused had the more severe disease. If so, the
poor results they had may be due not so much to
the transfusion itself as to the severity of the
underlying disease, be it eclampsia, prolonged
obstructed labour or rupture of the uterus.
Whatever the explanation, there is the whole
question of body fluid replacement during emer­
gency operative procedures in these volumedepleted patients w ith high haematocrit and pos­
sibly metabolic acidosis (Akinkugbe el al. 1977).
All these and other issues about these patients
deserve further study. Meanwhile, an important
step towards reducing maternal mortality in
places such as Zaria would be the installation of
a microhaematocrit centrifuge in the labour
ward itself, so that the haematocrit value of all
emergency admissions in labour could be deter­
mined on the spot and the results made known
promptly.

113

*

I

Further cause for concern in Zaria was the
mortality associated with bacterial infection in
the puerperium. Both booked and emergency
admissions who died were almost equally
affected. Faults in the hospital environment and
improper hygiene were chiefly the cause (see
chapter 2).
It is obvious that the majority of the obstetric
and medical complications associated with
maternal mortality, related as they were to the
prevailing social attitudes and severe depriva­
tion, were largely preventable. However, the
fact remains that the fundamental problem of
maternal health in the Zaria area is the existence
of an exceedingly large proportion of high-risk
women—early teenage primigravidae, high par­

ity, and lack of antenatal care—making up 40%
of the survey population compared with only
2%, for example, in Australia (Lumley 1980). If
maternal mortality is to fall, the long-term need
is a reduction in the proportion of women with
these three high-risk factors. The way to do so is
to improve living standards (Chalmers 1980),
never neglecting universal formal education in
the process (chapter 3). In the short-term, how­
ever, it should be possible to improve the health
of the pregnant women through antenatal care,
and then gain acceptance in the population so
that operative deliveries, especially caesarean
sections, can be performed as soon as the need
arises.

Appendix 12.1. Maternal mortality by social factors and booking status

Maternal mortality ratc/1000

Unbooked

I

HausaFulani

Other
Nigerians

27-8
(2593)
35-2
(3778)
19-5
(308)

(666)
15-0
(200)
0-0

Booked

Not stated

Not
stated

HausaFulani

Other
Nigerians

0-0
(16)
0-0
(2)
00
(29)

1-8
(3890)
3 9
(517)
0-0
(78)

10
(9989)
00
(335)
0-0
(69)

P)

00
(23)
00
(10)

00
(72)
2-1
(4365)

(5)
00
(90)

(0)
0-0

(0)
0-0

()•()

(2)

(14)

(48)

1-0
(7028)
0-7
(2773)
0-0
(12)
1-7
(580)

00
(40)
0-0
(27)
00
(9)
0-0
(36)

(6)
31-8
(6636)
0-0
(37)

00
(50)
9-4
(849)
00
(14)

0-0
(3)
0-0
(12)

(0)
0-0
(47)

2-3
(2193)
0-6

(0)

All deaths

211

8

All deliveries

6679

913

Social factors

Others

Others

Not
stated

HausaFulani

Other
Nigerians

00
(25)

0-0
(15)
0-0

(0)
00
(5)

Others

Not
stated

(21)

«>)

(0)

(II)
0-0

(0)

(<>)

(2)

(0)

(0)

(0)
0-0
(2)

0-0
(16)
0-0
(5)

(0)
00
(26)

(II)
0-0

(0)

(0)
0-0

(9)

(0)

(I)

(0)
0-0
(9)

(0)
0-0
(2)

(0)
0-0

(0)

(0)
0-0

(I)

(0)

(I)

00
(3)
0-0
(27)

(0)
0-0
(28)

(2)
0-0

(0)

(0)
0-0

(19)

(0)

(2)

(0)

(0)

Residence
Zaria

Outside Zaria

Not stated

Religion
Christian

Islam

00
(25)
31 9
(6617)

Others

Not stated

(0)
00
(37)

Formal education
Education
0-0
None

Not stated

7-5

00
(8)
00
(7)

(47)

(0)

8-9
(562)
11-7
(256)
0-0

0-0
(4)
0-0

(0)

0-0
(105)
0-0

(7)
(0)

00

00

(0)

96
(104)
1-8
(4347)
00
(34)

(8093)
0-0
(107)

0-0
(76)
0-0
(29)
00
(7)

(0)

(0)

(<>)

0

0

9

10

0

0

0

0

0

0

15

47

4485

10 393

112

30

28

21

0

2

Numbers of deliveries are shown in parentheses

00

(!

I

;r|i.

■ i|h!H

• ‘ .ll ’H

Appendix 12.2. Maternal mortality (deaths per 1000 deliveries) by age, parity and booking status

All deliveries
Maternal
age
(years)

< 15
15
16
17-19
20-24
25-29
>30
Not
recorded
Total no. of
deliveries
Total no. of
deaths
Deaths per
1000
deliveries

Unbooked Booked

Not
stated

All primignividae

Total

56 I
71-3
320
208
196
24 9
368

10-5
2-1
0-0
0-9
1-0
1-3
1-6

00

5-8
7-4
151

17

00

00

7654*

15 020*

219

19

286

1-3

37-8
39-7
14-3
7-6

UnNot
booked Booked stated

56-7

All multigravidac

Un­

Not
stated

Total

Unbooked Booked

0-0
0-0
0-0
0-0

0-0
12 7
12-7
7.1
5-9
7-3
15-3

0-0
22-0
27 I
45-3

00
0-0
1-5
15

00

00

1-6

0-0

0-0

Total

booked Booked

38-5
41 9
14-6
80
4-8
6 1
10-1

00
32-2
28-2
20-5
22-5
25-0
37-8

00
00
00
1-8
10
14
1-7

00

2-4

-i

Parity 2s 5

Not
staled

TI
Total

17-5
19-2

10-8
2-4
0-0
0-0
1-1
00
0-0

11

00

00

51*

22 725*

2767

3543

13

6323

4804

II 460

36

16 300

1693

2174

12

3879

0

238

92

4

0

96

125

15

0

140

59

3

0

62

10-5

33-2

15-2

26-0

34-8

1-4

()()

16- 0

00
0-0
0-0
00
0 0

73-6
32-9
21-0
11-8

1-1

0-0
0-0
00
0-0
00

1-3

0-0

0-0

8-6

• I

00
9-7
0-0
0-0

[ ■

11-6
20-1

00

>;
r

i ■

*Data on total deliveries include those on 102 mothers (83 unbooked with two deaths, 17 booked and two not stated) with missing parities

Mi

■

K
Tropical Doctor, April 1986

78 I OBSTETRICS AND GYNAECOLOGY

(1) Again the marking of the date of consultation
in the first column is sometimes difficult.
(2) The drawing in the second column reminds the
TBA to give appropriate nutritional advice.
(3) A cross (X) should be placed in the third column
if malaria prophylaxis has been given.
(4) The dispensing of iron tablets (with folic acid)
and/or multi-vitamin tablets should be marked in
the fourth column
(5) The dates for tetanus toxoid immunization,
which has to be given by the supervisor twice during
pregnancy, should be noted in the fifth column.
(6) Finally, referrals to higher levels of health care
(health centre or hospital) should be noted in the
last column.
DISCUSSION

i

I

I

The introduction of an antenatal card for use by
illiterate TBAs is not easy. First of all, it is import­
ant to discuss the idea with whoever will later teach
the TBAs and use it during their supervision in the
villages. Their commitment to using the card to
improve the performance of the TBAs is essential.
Therefore they should also be involved in the design
and content of the various drawings on the card.
Once the card has been designed, it should be field
tested by the nurses and midwives who will train
and subsequently supervise the TBAs. Changes
should be made according to their experiences.
Secondly, the antenatal card will only prove use­
ful if it can: identify the problems of a particular
pregnancy (e g. anaemia, insufficient fetal growth,
wrong presentation), and the at-risk cases; help
TBAs interpret their findings, by giving a clear set
of instructions and standing orders on what to do
in each situation (these instructions should be
taught together with the use of the card).
These two conditions can only be met by con­
tinuously supervising the TBA's performance and
discussing the findings with her. In this way the
card can provide a framework of things to look for
during the supervision, and so enhance its effec­
tiveness. At the same time it can improve the TBA’s
performance by showing her any deficiences in her
skills or knowledge. Thus it can help to direct TBAs’
training, and supervisors’ attention, more specifi­
cally towards aspects that need emphasis.
Finally, the card described here needs evaluation.
At present, it is too early to suggest how successfully
it identifies women with a high obstetric risk through
the regular control of TBAs. Studies of the sensi­
tivity of the antenatal cards (i.e. the proportion of
pregnant women with high-risk factors detected

through their use) are badly needed, not only for
the card presented in this paper, but also for the
antenatal cards used by nurses, midwives and
doctor.
This antenatal card has been a considerable help
in the process of improving the performance of
TBAs and represents a first step in making their
antenatal care more effective. Detailed studies of
this aspect of primary health care are urgently
needed to make the training of TBAs a more worth­
while activity.
REFERENCES
Chabot H T J & Savage F (1984) A Community Health Project
in Africa (slide set). Teaching Aids at Low Cost (TALC). PO
Box 49. St Albans, Herts. AL1 4AX, UK
Essex B J 4 Everett V J (1977) Tropical Doctor 7, 134-138
Sims P (1978) Tropical Doctor 8, 137-140
Watson D S (1984) Tropical Doctor 14, 133-135
Williams M (1980) The Training of TBAs: Guidelines for
Midwives Working in Developing Countries. Catholic Institute
for International Relations, 22 Coleman Fields, London N1
7AF. (New edition in preparation)

Should vesicovaginal fistula
be treated only by
specialists?
J G Thornton mb dtm&h*
Chogoria Hospital, Chogoria. Meru, Kenya
TROPICAL DOCTOR, 1986, 16, 78-79

INTRODUCTION

Vesico-vaginal fistulas (VVFs) are still common in
many developing countries. The largest series ever
reported (Tahzib 1983) refers to the period
1969-80, during which no less than 1443 cases were
collected at one university hospital in Northern
Nigeria. This report is typical of most papers from
developing countries on VVFs which are usually
derived from large numbers referred to teaching
hospitals, including many difficult cases referred
after treatment elsewhere has failed. Such reports
give the impression that fistula repairs should only
be attempted by expert vaginal surgeons. Unfor­
tunately experienced gynacologists are rare in the
rural areas of developing countries where most
fistulas occur, so the majority of repairs are probably
•Present address: Department of Obstetrics and Gynaecology.
Leeds General Infirmary, Leeds LS2 9NS, UK

Tropical Doctor, April 1986

carried out there by doctors without this special
training, who individually operate on only a few
cases in a lifetime. This paper reports the results of
a small series of 13 cases from Chogoria, a small
hospital in rural Kenya, which indicates that with
orthodox methods good results can be obtained by
nonspecialist doctors.
Before 1970, VVF repair was rarely attempted
at Chogoria. Patients were either referred, or if that
proved impossible, the ureters were transplanted
into the colon. This was an unsatisfactory proce­
dure, and although the author has seen two patients
still living between 10 and 15 years follow ing opera­
tion, most eventually die of renal failure (Preston
1951). Since then, with the initial encouragement
of Professor Donald Gebbie, at that time Professor
of Obstetrics and Gynaecology at Nairobi, vaginal
repair has been attempted in all suitable cases. This
experience has provided the material for this paper.

OBSTETRICS AND GYNAECOLOGY | 79

difficult for inexperienced theatre staff and
anaesthetists to maintain, and we have therefore
not attempted it.
A two- or three-layer repair technique was used.
Chromic catgut was used for a one- or two-layer
bladder repair, and monofilament nylon for the
vagina. In one case a Martius pedicle graft of fat
from the labium majus was used. Postoperative
drainage was always via an indwelling urethral
catheter, held in place with a suture rather than a
balloon, which was left in for a minimum of 14
days. Continuous catheter suction was never used.
RESULTS

Twelve of the thirteen fistulas were successfully
closed at the first attempt. In the remaining case a
pinhole residual fistula was successfully repaired
three months after the primary operation.
DISCUSSION

MATERIALS AND METHODS

Fourteen fistulas were seen over a period of twelve
years, 1972-83. All were of obstetric origin, and all
but two had followed delivery outside Chogoria
Hospital, mostly at home. Of the two fistulas which
occurred in this hospital, one was caused by Cae­
sarean section and the other followed vaginal deliv­
ery in a woman who had previously had a Caesarean
section. No fistula following gynaecological surgery
was seen, and malignant fistulas have not been
included.
One large fistula was referred without any
attempt at repair, and no further details are avail­
able. Of the 13 fistulas repaired, 5 were juxtaurethral, 3 mid-vaginal and 5 juxta-cervical. Only
one patient had undergone a previous attempt at
repair (elsewhere, per abdomen). A minimum
period of three months was allowed between delivery
and repair. The longest period of delay was ten
years.
Eight different doctors carried out the repair
operations, none of them being responsible for more
than three. The only specialist gynaecologist
amongst them, Professor Gebbie, operated on two
of the patients.
Al! fistulas were repaired per vaginam, except for
one vesicocervical fistula which was repaired per
abdomen. The vaginal operations were performed
with the patient in the lithotomy position. This
makes the repair of low fistulas adherent to the
back of the pubis rather difficult, and some authors,
notably Lawson & Stewart (1968), have advocated
the knee-chest position for them. However, this is

Although this is a small series of VVF repairs, it
is of interest because it shows that good results can
be obtained by relatively inexperienced doctors in
a small hospital. It could be argued that we were
just lucky and that nonspecialists should not attempt
VVF repair. However, travel and accommodation
costs and long waiting-lists often prevent successful
referral. Patients may misunderstand a specialist
from far away when he tells them to return in a
few months for surgery. They may believe he cannot
cure them but is unwilling to say so.
In these situations we would advise nonspecialists
to study the writings of Chassar Moir (1967) and
Lawson & Stewart (1968), whose books have
guided us. They should refer circumferential fis­
tulas, massive fistulas extending from cervix to
urethra, cases complicated by severe scarring, cases
with a concurrent rectovaginal fistula, and failures
which have not been cured by simple second opera­
tion. They can attempt the rest with a good chance
of success.
ACKNOWLEDGEMENT

The author is indebted to Mr J B Lawson for help in the
preparation of this paper.
REFERENCES

Chassar Moir J (1967) The Vesicovaginal Fistula, 2nd edn.
Bailliere, Tindall & Cassell, London
Lawson J B & Stewart D B (1968) Obstetrics and Gynaecology
in the Tropics and Developing Countries. Edward Arnold,
London
Preston PC (1951) Journal of Obstetrics and Gynaecology of
the British Empire 58, 282-290
Tahzib F (198 3) British Journal of Obstetrics and Gynaecology

I

Tropical Doctor, July 1980

OBSTRUCTED LABOUR IN BENIN CITY. NIGERIA I 113

Obstetrics and Gynaecology
Obstructed labour—a fouryear survey at the
University of Benin
Teaching Hospital, Benin
City, Nigeria*
A. U. Oronsaye, MB, BS, MRCOG
M. I. Asuen, MD, FACOG
University of Benin Teaching Hospital,
Department of Obstetrics and Gynaecology,
Benin City, Nigeria
TROPICAL DOCTOR, 1980, 10, 113-116

In spite of its rarity in the developed countries,
obstructed labour still contributes immensely to
high maternal and perinatal mortality and morbid­
ity in many developing countries of the world, like
Nigeria. Culture factors, poverty, and inadequate
health facilities are some of the contributing factors
to obstructed labour in our area. Malnutrition,
uncontrolled infectious diseases in childhood, and
the resultant poor skeletal development during the
reproductive age have also been blamed for
obstructed labour in the developing countries (Baird
1963; Gilles et al., 1956; Lister 1960). A multifaced
approach is therefore necessary if obstructed labour
is to be eliminated from all obstetrical units in
Nigeria. A review such as this aims at sustaining
interest in this problem.
MATERIALS AND METHOD

The post-partum records of all mothers delivered
at the University of Benin Teaching Hospital. Benin
City, Nigeria from April 1, 1973, to March 31,
1977, were reviewed. During this period, there were
6,369 deliveries recorded and 136 cases or 2% of
the patients had obstructed labour. During the same
period, 520 Caesarean sections were performed in
the unit, 102 of which were carried out because of
obstructed labour. One hundred and twenty-six case
notes were suitable for analysis and these form the
basis of this study. Most of the women were brought
moribund from neighbouring villages and homes,
• This paper was presented at the first International Conference
of the Nigerian Society of Obstetrics and Gynaecology at Ibadan,
Nigeria, in October 1977.

having been in labour for days. Some of them were
originally booked for antenatal care but absconded
to avoid possible Caesarean deliveries. A small
group was also referred belatedly from some mater­
nity homes, where either because of inadequate
facilities and/or due to ignorance the patients had
waited too long. Table 1 shows the distribution of
causes of obstructed labour.
In the three cases of fetal abnormality, two were
of anencephalus and one hydrocephalus. Table 2
shows the various presentations among the cases
and their relationship to parity. Details of these
cases were as follows:
Vertex presentation. Seventy-nine cases presented
with the head and 11 of these patients were booked
for antenatal care. There were 19 cases of occiput­
posterior, nine brow, three face and 48 undeter­
mined because of marked caput formation. Fetal
heart sounds were absent in 17 patients and queried
in two. The cervical dilatation was more than 8 cm
in 61 patients, less than 8 cm in seven and not
recorded in 11 patients. There were three cases of
Caesarean sections (for cephalo-pelvic dispropor­
tion) in this group. There were 10 laparotomies for
ruptured uterus, two of which were previous> sec­
tions,
t
, and nine craniotomies, two of which were for
malformed babies. Sixty emergency lower segment
Caesarean sections were performed. There was a
total of 21 stillbirths (four fresh and 17 macerated)
and 58 live births. Fetal weight was more than 2.5
kg in all 70 cases with operative deliveries.
Transverse lie. There were 12 cases with the fetus
in transverse lie. Two patients had had two previous
sections each for cephalo-pelvic disproportion and
one patient had had a previous Caesarean section
for an unknown cause. There were two cases of
ruptured uterus (both of which were previous Cae­
sarean sections). Fetal heart sounds were absent in
three patients and questionable in two others. The
Table 1. Causes of obstructed labour
Feto-pdvic disproportion
Impacted transverse lie
Mento-posterior/face presentation
Shoulder dystocia
Fetal abnormality

103
12
6
2
3

Total

126

1 14 I OBSTRUCTED LABOUR IN BENIN CITY. NIGERIA

Table 2. Parity incidence in relation to various causes of obstructed labour in Benin
Presentation

0

I

2

3

4

5

6

7

8

Total

Vertex

37
0
0

7
2
0

0
I
2

19
1
0

6
6

4
0
0

0
0
0

0
0
0

79
12

2

6
2
0

I

0

I

I

0

2

1

0

0

6

(twins)
Breech
Fetus lying free in peritoneal cavity
Not recorded

0
2
1
0

0
1
0
0

0
0
I
2

I
I
0
0

0
4
0
I

0
0
1
3

2
I
0
I

1
0
0
0

0
0
0
2

4
9
3
9

Total

41

10

7

23

19

14

9

I

2

126

Transverse lie
Compound presentation
Shoulder presentation with arm or hand prolapse
(singleton)
Shoulder presentation with arm or hand prolapse

I '

cervical dilatation was less than 8 cm in all 12 cases.
Abdominal delivery was carried out in all 12
patients resulting in four stillbirths (one macerated)
and eight live births. Eight babies weighed more
than 2.5 kg. All three previous sections were booked
patients who defaulted to avoid repeat section.
Compound presentation. There were four patients
with compound presentation. One patient had a
previous Caesarean section for unknown indication.
Fetal heart sounds were present in all patients and
the cervical dilatation on admission was less than
8 cm. Caesarean section was the mode of delivery
in all patients resulting in live births. Fetal weight
was more than 2.5 kg in all cases.
Shoulder presentation with arm or hand prolapse
{singleton) There were six patients in this group.
Fetal heart sounds were present in five and absent
in one. The cervical dilatation on admission was
more than 8 cm in four patients. Caesarean section
was carried out resulting in five live births and one
fresh stillbirth. All babies weighed more than
2.5 kg. Two of these six patients were booked for
antenatal care.
Shoulder presentation with arm or hand prolapse
(twins). Four patients with twin pregnancies, three
of whom delivered their first twin at home, were
admitted with arm prolapse of the second twin. The
fourth patient delivered her first twin in the hospital
and was subsequently obstructed with the second
twin. All four babies were delivered by Caesarean
section resulting in three live births and one still­
birth. Three of these patients were booked cases.
All four babies weighed less than 2.5 kg. The
cervical dilatation on admission was less than 8 cm
in two patients.
Breech presentation. There were nine such(cases in
the series. Fetal heart was queried in oqcl patient

4

and the cervix on admission was dilated more than
8 cm in five patients. Deliveries were by Caesarean
section in all cases resulting in one stillbirth and
eight live births. All babies weighed more than
2.5 kg. Seven of these patients were booked for
antenatal care, but defaulted in their last trimester
to avoid Caesarean section.
Fetus lying free in the peritoneal cavity. In three
patients, the fetuses were lying free in the peritoneal
cavity following uterine rupture. Laparotomy was
done resulting in three stillbirths (two macerated).
One baby weighed more than 2.5 kg. The macerated
fetuses were not weighed. The cervical dilatation
was also not recorded.
Presentation not recorded. In nine patients the
presentation was not recorded. Fetal heart sounds
were absent in three cases and the cervix dilated
more than 8 cm in six patients. Caesarean section
was performed and there were six live births and
three stillbirths (fresh). All babies weighed more
than 2.5 kg. One patient had had a previous section
for antepartum haemorrhage in this group.
MANAGEMENT

The management of our 126 cases is shown in
Table 3
Table 3. Summary of management
Emergency lower segment Caesarean section
Caesarean hysterectomy
Sub-total hysterectomy
Repair of uterus
Repair of uterus with bilateral tubal ligation
Craniotomy

102
4

Total

126

1

7
3
9

OBSTRUCTED LABOUR IN BENIN CITY. NIGERIA 1115

Table 4. Associated diagnosis on admission of
cases of obstructed labour
Ruptured uterus with intrauterine fetal death
Intrauterine death with intact uterus
? intrauterine death
Fetal heart beat present

95

Total

126

15
10
6

Obstruction was relieved in nine patients by
destructive operation; seven of these patients had
no fetal heart beat. The remaining two patients
with fetal heart beat present were one case of
anencephalus and one of hydrocephalus.
Fifteen patients with ruptured uteri were sub­
jected to laparotomy. Four had Caesarean hyster­
ectomy, one had subtotal hysterectomy, and 10 had Fig. I. Oedema of the vulva in obstructed labour.
uterine repair. Three of the patients with uterine
repair also had bilateral tubal ligation. Emergency births. The average hospital stay was 26 days with
lower segment Caesarean section was carried out a few patients staving
weeks,
staying ud
up to six weeks.
in 102 cases, resulting in 92 live births and 10
stillbirths.
DISCUSSION
At a time of increasing emphasis on good perinatal
COMPLICATIONS
medicine in the developed world, it is an irony that
The complications associated with the diagnosis of in Nigeria, as in other developing countries,
obstructed labour are shown in Table 4.
obstructed labour still remains a prominent feature
The complications arising from the mode of man- of our obstetric casualties. Many mothers and chilagement in each case are shown in Table 5.
dren either perish or are being maimed in the
Following craniotomy in nine cases, there were process of childbirth due to obstructed labour,
two vesico-vaginal fistulae, one of which eventually Ampofo (1969) attributed 61 maternal deaths dursuccumbed
due to sepsis. Fifteen laparotomies .per­■ ing a five-year period (1963-1967) in Ghana to
'
formed for ruptured uteri resulted only in stillbirths obstructed labour. Obstructed labour constitutes by
with eight of the mothers sacrificing their child- ffar the
’ commonest cause vof,‘ ruptured uteri in
bearing function. Emergency lower segment Cae- Nigeria. In our series the cause of obstruction
sarean section was carried out in 102 cases resulting due to disproportion in 103 cases out of 126 or
in one vesico-vaginal fistula, 10 fresh stillbirths,I and 81.4%. This
is similar to the experience
of other
- ----------•*
IVIIVV VJI
five neonatal deaths. The overall perinatal mortality workers from the developing countries (Ampofo
was 309.5/1,000 and maternal mortality of 8/1,000 1969; Cannon and Hartfield°l 964; Lawson* 1965)°
Table 5. Results of management of 126 cases of obstructed labour
Loss of

Procedure

Number

child­
bearing
function

Craniotomy

9
15

8

Laparotomy
Emergency lower segment
Caesarean section

102

Total

126

Vesico­
vaginal
fistula
2

Stillbirth

Fresh

Neonatal
Macerated death

Perinatal
death

Maternal
death

2
2

7

9
15

I

13

10

8

2

14

20

5

15

5

39

I

116 1 OBSTRUCTED LABOUR IN BENIN CITY. NIGERIA

There was a total of eight previous Caesarean
sections (six primary and two previous sections),
four of which ruptured following obstruction. Four
out of eight previous sections rupturing would tend
to support the views of those who deplore Caesarean
section for dead babies and hesitate to perform it
in the presence of live babies in order to avoid
scarring of the uterus that may be subsequently
obstructed (Lawson 1965). In one of the largest
incidences of ruptured uteri yet described (171 cases
out of 15,908 deliveries) only 38 cases or 22% were
ruptured scars (Rendle Short 1966). It thus seems
that in the planned management of obstructed
labour it does not matter much if a scar is left in
the uterus. Many destructive operations have been
performed on apparently well chosen patients with
disastrous results (Mokgokong and Crichton 1974).
Crichton and Bouile (1964) reporting on rupture
of the unscarred uterus gave five criteria for intrauterine manipulation of neglected obstructed
labour. These are as follows: (1) full dilatation of
the cervix; (2) relaxed uterus; (3) intact or recently
ruptured membranes; (4) reasonably sized pelvis;
and (5) baby whose weight is not above average.
Rarely are those criteria met in most cases of
obstructed labour so that fistulae formation and
ruptured uteri are apt to occur with increasing
frequency particularly in inexperienced hands.

Tropical Doctor, July 1980

Thus we believe that in the management of
neglected obstructed labour. Caesarean section
should be liberalized. Extraperitoneal Caesarean
section may find practical application in our set-up
as shown by Mokgokong and Crichton (1974).
There may still be a limited place for destructive
operation in obstructed labour but this must be
done by those skilled in the art, if the associated
mortality and morbidity are to be avoided.
ACKNOWLEDGEMENT

We are grateful to our consultant colleagues who allowed us to
include their patients in the series and also to Mr James Okhakhu
for secretarial services.
references

Ampofo, A. D. (1969). West Afr. med. J.. 18, 17.
Baird, D. (1963). British Obstetrics and Gynaecological Prac­
tice, 2nd edition, p. 753. London: Heinemann.
^,1„1W11 D., Oltw
Crichton,
and Bouile, P. (1964). Lancet, 1, 369.

Cannon, D.S. H.,and Hartfield, V. J. (1964). J. Obstet. Gynaec.
Brit. Cwlth. 71, 940.
Gilles, H. M. et al. (1956). Brit. med. J..2, 686.
Lawson, J. B. (I965).Z Obstet. Gynaec. Brit. Cwlth, 72, 877).
Lister U. G. (1960) J. Obstet. Gynaec. Brit. Emp., 67, 188.
Mphahlele, M., and Van Der Meulen, A. J. (1975). S. Afr. med.
J.. 46, 1204.
Mokgokong, E. T., and Crichton, D. (1974). S’. Afr. med. J.. 48,
788.
Rendle Short, C. W. (1964). J. Obstet. Gynaec. Brit. Cwlth. 68,
44).
Rendle Short, C. W. (1966). Amer. J. Obstet. Gynaec., 79, 1114.

Notes and News
Audio-visual training in tropical diseases

The first two of these series (malaria and schis­
tosomiasis) are completed; the others are in
In cooperation with the World Health Organiza­ preparation.
tion, the Royal Tropical Institute in Amsterdam
These teaching aids are not only aimed at medical
and the Liverpool/London Schools of Tropical schools, but also at institutions which train inter­
Medicine are developing audio-visual packages in mediate and primary health workers. We would be
the form of microfiches and strips of unmounted very grateful to you if you could provide us with an
slides (84 slides each) on eight important groups of estimate of how many sets you might be interested
tropical diseases (malaria, schistosomiasis, leish­ in purchasing and a list of institutions which would
maniasis, leoprosy, geohelminthic infections, amoe- be specially interested in the slide series on malaria.
biasis, other helminthic infections, other protozoal The cost of each of these sets is $9.00 for microfiche
infections). These slides/microfiches are intended edition and $15.00 for unmounted 35mm slides
for the training of health workers of various levels edition. Orders should be addressed to Royal Trop­
and categories engaged in the control and study of ical Institute, Dept. Tropical Hygiene, Section
tropical diseases. The accompanying texts are avail­ M.O.N., Mauritskade 63, 1092 AD Amsterdam,
able in English, French, and Spanish.
Netherlands.

M
124

I

OBSTETRICS AND GYNAECOLOGY

Antenatal care in a district
hospital*
Dr John Chiphangwi
Queen Elizabeth Central Hospital
PO Box 95, Blantyre, Malawi

Tropical Doctor. July 1987

It is wise to sit down with the hospital administrator
or accountant to find out what money is available
for maternal and child health work. It is only when
one appreciates financial constraints that one can
tailor one’s plans to the resources available. The
drugs to be stocked and the type of laboratory tests
to be done during antenatal care will depend on the
facilities and money available.

tropical DOCTOR, 1987, 17, 124-127

AT-RISK APPROACH

Although the aims of antenatal care are the same
all over the world, the actual practice varies from
one region to another. Even in the same country
the practice in a teaching hospital in an urban area
very often differs from that in a district hospital in
a rural area. It is important, therefore, that the
provider of antenatal care in a rural setting should
understand the circumstances and peculiarities of
the district, including the human, technical and
financial resources available, in order to give a
service that is responsive to the needs of the people
served. For this reason I intend to dwell more on
the administrative and organizational aspects than
on the routine aspects of antenatal care.

A typical district hospital in the tropics is unlikely
to have all the staff and facilities that one might
wish in order to reduce maternal and perinatal
mortality and morbidity. The at-risk approach
essentially involves: (a) taking an inventory of the
main causes of maternal and perinatal mortality
and morbidity in the hospital and the district it
serves; (b) arranging these problems in some order
of priority or importance, based on figures obtained
in the district; (c) working out a system that ensures
that women at high risk are picked out easily, so
that they can be given the special care that they
need. How can this be done?

ACTION-ORIENTATED ANTENATAL CARD
There is no substitute for a well-motivated,
inventory of resources
properly-trained and alert midwife when it comes
Let us assume that a doctor has just been posted to identifying the at-risk antenatal patient, but the
to a district where he has never worked before. One majority of district hospitals do not have this ideal
of the first things he should do is to make a quick midwife. Firstly, there are never enough midwives,
inventory of the resources available in the district and secondly the workload is such that in any one
for antenatal care.
day the same midwife is expected to run clinics
Human resources: The numbers, and levels of train- (nutrition surveillance, immunization and antenatal
ing, of nurse/midwives, paramedical staff, and tra- care), look after a ward, and perform a number of
ditional birth attendants (TBAs) who provide other tasks. In these circumstances, the traditional
antenatal care in both static and mobile clinics antenatal cards, which expect the midwife to look
should be ascertained. The standard (and type) of for each of the risk factors and then decide what to
coverage of antenatal services will, to some extent, do in any particular case, might lead to some atdepend on the number and calibre of the staff who risk patients being missed and not being referred
give this
’ ’ ’ service.
.
on time.
Infrastructure: This includes the number, quality,
On the other hand, no antenatal card, however
and location of health centres and maternity units well-designed, can possibly display all the known
(both static and mobile) that are available in the risk factors - with appropriate action inserted. To
district. The quality of roads during the rainy season do this it would have to stop being a card and
and the availabilty of transport are equally become a manual of obstetrics. Clearly one needs
important.
to look at the most important risk factors in each
Financial resources: Although most medical schools country or district and then design a card that
do not prepare medical students in aspects of fin- ensures that these conditions are not missed. It
ancial management, the new district medical officer follows that no one card can be ideal for every
will find himself being confronted with this problem. country.
Many action-orientated cards have been
•Second in a series on Essentials of an Obstetric Service. See:

designed. Some have a check-list of the most import-

April 1987, p. 77

ant risk factors regarding the past history as well

Tropical Doctor, July 1987

1

A

I
i

)

<

j

OBSTETRICS AND GYNAECOLOGY

I

125

Recently, there has been a lot of excitement about
as the current pregnancy, and the midwife need
only tick off the appropriate column. In some, no the use of an ordinary measuring-tape to assess the
attempt is made to give a score for each risk factor; symphysis-fundus (SF) measurement, which can be
others give a score for each risk factor so that one plotted on a graph on the antenatal card so that
can grade the degree of risk that a particular mother progress can be seen at a glance. Various workers
has. Yet another approach is to use a colour scheme: have shown that when the last menstrual period is
the card is designed in such a way that the answers known accurately, a normal propagation of the SF
to questions on the check-list arc ticked in either measurement is almost always associated with a
white or red squares; all ticks in red squares denote normal-for-dates baby. When the growth curve goes
risk, and all the midwife has to remember is that below the 10th percentile the result is usually a
any mother with a tick in a red square needs to be small-for-dates baby. When the growth curve is
referred for further attention by a clinical officer, above the 90th percentile one should suspect mul­
doctor or specialist, as the nature of the case tiple pregnancy, polyhdramnios or a large-for-dates
baby.
demands. This is the card used in Malawi.
Some authors8 have suggested that the SF meas­
Whichever method is used, the important thing
is that the midwife should be able to decide at a urement is not as accurate as was once thought.
glance which mother needs to be referred higher up From a study they concluded that SF measurement
the health service hierarchy for further attention by was so variable that it was not very reliable as an
someone with the necessary expertise to handle the index of uterine (and fetal) growth. Not only were
there variations between observers, but also varia­
problem.
In our experience the main problem with the tions by the same observer on the same patients.
action-orientated card is that the less motivated and This was attributed to a number of factors, such as
less able midwife will tend to stop at the conditions movements of the fetus and position changes (at
listed on the card. Such nurses fail to ask questions close intervals of time), and intermittent physiorelated to conditions not listed. This is not a con­ logical contractions of the uterus.
Where does this leave us? I think that no one
demnation of the at-risk card but a reminder that
constantly
be
reminded,
through
method
of clinical assessment of fetal growth. and
midwives must c--------- j -- ----------- .
w
refresher courses, about other risk factors that need condition is perfect. Even the most enthusiastic SF
’ ■ which
■ ' an exhaustive check listt measurers accept this. In our community in rural
to be looked for (of
is contained in explanatory notes that accompany Malawi only 50% of women remember their last
the card). However, if the card is designed after menstrual period with any degree of certainty.
looking at the commonest and most important risk Reliance on SF measurement alone in these circum­
factors in each community, then it will ensure that stances is likely to be inaccurate. Supplementation
the majority of women with the most common risk with abdominal palpation is probably still necessary.
factors will be identified by use of the card. This is
good enough; even in countries with the best facil­ DATA TO BE RECORDED
ities, detection of risk factors is not 100% ON ANTENATAL CARD
What are the acceptable minimum of measurements
accurate1,2.
that should appear on the antenatal card?
authors nave
have snown
shown mat
that neigni
height is
assessment of fetal growth
Height: Many autnors
The assessment of fetal growth and condition is one related to obstetric per forma nee4**. Women of short
;
are more likely
of the most important
aspects of‘ antenatal* care, yeti stature
J
. than taller ones to have
it is often no^done
not done very well. The time-honoured problems related to cephalopelvic disproportion. It
method of assessment by abdominal palpation and is important for each country or district to work
use of fingerbreadths above and/or below the out, on the basis of local experience, what is the
umbilicus (or above the symphsis pubis, or below minimum “acceptable” height. All women below
the xiphisternum) has been shown to be inaccurate, that height should then be referred.
It is easy to put a mark on a wall of the antenatal
even in the hands of senior obstetricians. The more
modern and accurate methods (e.g. serial measure- clinic. All women below that mark will need to be
ment of the biparietal diameter using ultrasound, referred. Another method suggested5 is to put a bar
or the serial measurement of serum or urinary at an appropriate height at the entrance to the
oestriols) arc not available in most developing clinic. Those who can walk in without stooping need
countries.
to be referred.

!

i

I
N

i

I

126

i
!

I

OBSTETRICS AND GYNAECOLOGY

Tropical Doctor, July 1987

Fundal height (SF) has been discussed above.
pregnant women in the tropics. In Malawi anaemia
Weight: Weight gain in pregnancy may be negli- (Hb<10g/dl) appears in about 10% of the obstetric
gible or high, depending on circumstances. In many population. Nutritional factors appear to be importAfrican women weight gain is minimal and often ant in the aetiology of anaemia in the tropics,
erratic, even when the fetus is growing well. This although worm infestation, bilharzia and malaria
makes it difficult to ascertain how important weight are also important factors.
gain is. However, if there is no weight gain, or if
The prophylactic use of iron in pregnancy is
weight is lost, in the presence of other warning signs widely accepted and there is a strong case for
(e.g. static SF measurements and reduction of training traditional birth attendants to distribute
abdominal girth) then one should suspect intra- iron to their clients. In addition, folic acid should
uterine growth retardation. Excessive weight gain be given in cases of multiple pregnancy and severe
would make one suspect the possibility of multiple anaemia. If worms are present they should be eradipregnancy or impending preeclampsia.
cated by appropriate treatment. With regard to
As with SF measurements, the weight can either bilharzia, we recommend that treatment be deferred
be tabulated or shown in graphic form on the until after delivery.
antenatal card.
Blood pressure: In some parts of the tropical world Antimalarials
preeclampsia/hypertension is an important con- To all intents and purposes chloroquine is the only
tributor to maternal and perinatal mortality and safe drug to be used for malaria suppression in
morbidity. A country where such a problem exists pregnancy. Its efficacy is, however, considerably
should ensure that no antenatal clinic is without a reduced in those countries of Asia and East Africa
sphygmomanometer. It is best to invest in a wall’
where resistant strains of P. falciparum to chloromounted type (or one with a tripod), which is quine have been reported. Pregnant women should
expensive initially but lasts longer and is therefore be advised to avoid, where possible, being bitten by
eventually. cheaper.
mosquitoes
(e.g.ofby
use of mosquito
. •
mosquitoes
(e g. by use
mosquito
nets, coilsnets,
and coils and
Haemoglobin and blood group: Most district hos- repellents), in addition to taking their weekly
pitals can perform haemoglobin (Hb) examinations, chloroquine (300 mg base per week). They should
As anaemia is a major contributor to maternal and also be advised to drain stagnant water and clear
perinatal mortality and morbidity, each mother the bush near their houses. They should seek medishould have at least one examination at the begin- cal advice immediately in case of fever, and in those
ning
with
can
i where
* resistance to chloroquine
> is a big prob.
. • of antenatal
' care.~Those
’
1-1 a1-low
. 2_Hb
’------Jbe areas
given appropriate treatment and prophylaxis; it is lem alternative drugs should be available for• use
also prudent to repeat the Hb examination, in those (e.g. amodiaquine, quinine, Fansidar).
cases, towards term. Most rural clinics cannot
afford to do more than two Hb estimations.
Tetanus toxoid
Urine/stool testing: In communities with heavy Neonatal tetanus is a problem in most tropical
worm infestations it is prudent to examine the stool countries (Table 1). It is also one that can easily be
in those with anaemia. Testing urine for the pres­
ence of protein is useful in those areas with problems Table 1. Neonatal tetanus deaths (modified from
of hypertension. Testing for glucose is also advisable Population Reports, Series L, No. 5,
if diabetes is a common problem; it may not be March-April 1986)
necessary as a routine examination in all areas.
Sexually transmitted diseases (e.g. syphilis): These
Deaths per
are increasingly important in most countries. If this
1000 live births
is a problem in your community, then VDRL (or
3
some other test) should be performed at the start Yemen Arab Republic
Thailand
5
of antenatal care.
Sudan
9
Sickling: This test is not done routinely, but only Malawi
10
when indicated in appropriate populations.
Somalia
21
Nepal
15
Bangladesh (various surveys)
22-37
MEDICATION DURING PREGNANCY
Pakistan
31
Iron and folic acid
India (various surveys)
5-67
Anaemia is one of the most important problems in

Tropical Doctor, July 1987

I

eradicated if nonimmune pregnant women are given
tetanus toxoid at least twice during pregnancy. The
two doses of tetanus toxoid must be given at least
4 weeks apart, the first dose early in the second
trimester and the last at least 2 weeks before deliv­
ery. Not only does this protect the baby against
tetanus (for 5 months after delivery) but it also
protects the mother should she happen to deliver in
unhygienic conditions. Two doses protect the
woman for at least 2 years. After the third dose
protection lasts for 5 years.
Other immunization
There is very little justification for other forms of
immunization during pregnancy. Prophylaxis
against other diseases, e.g. rubella, should be given
before pregnancy. (This is rarely done in the
tropics.)

obstetrics and gynaecology

!

127

health care. Tact and explanation are important in
order to avoid people leaving the waiting area for
home. This usually happens where the quality of
the building and/or the food is poor.

i

CONCLUSION

Antenatal care at a district hospital should start
with an understanding of the problems peculiar to
that district, followed by an inventory of resources
available. On this basis a rational approach can be
made to the level of care that is possible and
appropriate within the constraints of manpower,
money, and infrastructure. Health for all by the
year 2000 can never mean the same thing to every
country. Priorities in each country have to be iden­
tified and steps taken to provide services that will
address those priorities.

1

REFERENCES
PLACE OF DELIVERY

As indicated above, the aim of the at-risk approach
is to identify those at high risk so that they can
deliver at a unit which is best suited to deal with
their problem. For instance, a mother with one
previous caesarean section should be referred to a
unit with facilities to perform a repeat caesarean
section should this become necessary.
Some multigravidas (gravida 2-4) with no risk
factors can be delivered in a rural clinic or by the
traditional birth attendant. Women of high parity
(gravida 5 and above) are best delivered in better
staffed and equipped units, as the likelihood of
problems is greater in this group.
waiting mothers’ area at district hospitals

1

I

1
2

3

4

5

Hall MH, Chng P, MacGillivray I. Is routine antenatal care
worth while? Lancet 198O,ii:78-8O
Chng PK, Hall MH, MacGillivray I. An audit of antenatal
care, the value of the first antenatal visit. Br Med J
1980;281:1184-6
Bagger PV, Eriksen PS, Secher NJ, Thisted J, Westergaard
L. The precision and accuracy of symphysis-fundus distance
measurements during pregnancy. Acta Obstet Gynaec Scand
1985;64:371-4
Baird D. Combined textbook of obstetrics and gynaecology.
Edinburgh: Livingstone, 1962
Everett VJ. The relationship between maternal height and
cephalopelvic disproportion in Dar es Salaam. East Afr Med
J 1975;52:251-6

Complication of cervical
cerclage

The majority of district hospitals in rural areas do
not have enough ward space to accommodate all
waiting mothers. It has been found appropriate in
some district hospitals to construct low-cost build­ Timothy E E Goodacre bsc frcs*
ings to accommodate those mothers who do not Mvumi Hospital, Dodoma, Tanzania
require close observation. This enables the nurses
to give closer attention to those who are sick and TROPICAL DOCTOR, 1987, 17, 127-128
in greater need of assistance. The mothers in the
waiting area can be reviewed regularly in the clinic Cervical cerclage is a well established and fre­
as their cases demand, and they are close enough quently performed procedure for the treatment of
to the ward to come for immediate admission should incompetence of the internal cervical os. With due
labour start.
care in management, the procedure has a low com­
It is important that a midwife be assigned to plication rate and can be regarded as relatively
check on these women regularly, because some, harmless. Neglect, however, can result in major
women may not recognize signs of impending prob­ complications, which might be expected to occur
lems. It would be sad if tragedies were to arise
when the mother is so close to the hospital. Some • Present address: Department of Plastic Surgery, Mount Vernon
women regard the waiting area as second-rate Hospital, Northwood, Middlesex

ii

ii

d

I

Tropical Doctor, April 1986

Antenatal card for illiterate
traditional birth attendants
H T J Chabot md dtph*
K H Eggens md
TROPICAL DOCTOR, 1986, 16, 75-78

INTRODUCTION

The importance of accurate and regular recording
of the findings during antenatal consultations is
generally accepted. Various control cards for ante­
natal clinics exist (Sims 1978, Watson 1984, Ken­
nedy - personal communication, Essex & Everett
1977). All of them are designed for use by trained
(auxiliary) nurses and/or midwives. With various
degrees of complexity, the aim of these cards is to
identify important risk factors during pregnancy so
that appropriate action can be taken.
In Guinea Bissau, the training of traditional birth
attendants (TBAs) as part of the primary health
care (PHC) programme was started in 1977 (Cha­
bot & Savage 1984). Almost all TBAs are illiterate.
After some years it became apparent that their
work was not as effective as had been expected: for
example, neonatal tetanus was still occurring in
villages with TBAs. However, it was difficult to
supervise and support their work, as no records were
kept of what was actually done.
The need for an antenatal card that could be
used by illiterate TBAs became apparent. The card
presented in this paper was developed by the Min­
istry of Health and Social Affairs and has been
used nationally since 1982. A good description of
the contents and methods of TBA training has been
given elsewhere (Williams 1980), so this aspect will
not be discussed here.
AIMS OF ANTENATAL CARD

(1) To help the nurse-supervisor discover and cor­
rect mistakes made by TBAs. The card was there­
fore designed as a tool for teaching and supervision
by the health staff, enabling them to explain mis­
understandings to the TBAs, while caring for the
pregnant women in their villages.
(2) To help TBAs remember what to look for during
antenatal consultations, so as to take appropriate
action. The card should therefore help TBAs to
interpret results correctly.
•Correspondence to: Boite Postale 82, Segou, Mali

OBSTETRICS AND GYNAECOLOGY | 75

(3) To enable TBAs to identify pregnant women at
risk, and refer them when necessary. All essential
information for discovering women at risk should
therefore be included on the card.
HOW TO USE THE CARD

As the card is to be used by illiterates, the need for
writing has been minimized. All essential infor­
mation can be recorded by putting a cross (X) in
the appropriate box, or by putting the necessary
number of ticks in the correct place. Only the
administrative information (name, age and village)
needs to be written down, and this has to be done
only once during the whole pregnancy. A local
schoolteacher or literate member of the community
can help with this. The pictorial symbols used in
the card were developed and tested in the field with
the PHC nurses. They were designed by an artist
who had also participated in the national training
guide for VHW/TBAs.
As with the road-to-health chart, the woman
keeps her own antenatal card (in a plastic bag if
available), as a record for other health workers of
her present and previous pregnancies.
DESCRIPTION OF THE ANTENATAL CARD

The card is divided into three main parts: history,
examination, and care.
History
The front page (Figure 1) gives four pieces of
information on the reproductive life of the pregnant
woman:
(1) The number of her children alive.
(2) The number of her children that have died.
The total number of pregnancies can then easily be
obtained by adding the number of live and dead
children. In this way parity is known more accu­
rately than by asking “How many pregnancies have
you had?’’ as women tend not to mention their
children that have died. Furthermore, TBAs find it
difficult to ask pregnant women in a reliable way
how many of their children were born dead and
how many were born alive: it is much easier to ask
the number of children actually alive and dead.
Once this is known, the supervisor can find out the
number of stillbirths among the children that have
died by looking through the answers. The first
(empty) box is meant for women with no previous
pregnancies.
(3) The number of abortions can be indicated by
writing the number or by putting the right number
of ticks (lX) in the open space.

76 | OBSTETRICS AND GYNAECOLOGY

Tropical Doctor, April 1986

Republica da GuirW-Bissau
„
M.S.A.S.
,
Projscto
Sadde
Base

Piche n~
Fl CHA DG CONSULTA PR^-NATAL-

Notns 4a Gravida
Nom« da labanca

Idada ___

Sector de

HISTORIA

Utt | tftif

*

OS&SSb
<B38K&)

rrO

(K

'Wrfta >> <KS£)

omm

ox (me

Figure 1. Front of antenatal card: history
(4) The TBA indicates the number of “moons” the
pregnant woman thinks she is pregnant by putting
a cross (X) in the appropriate box during the first
antenatal visit.
Examination

using a measuring tape or a cord with knots at the
corresponding heights for every month of preg­
nancy, from 5 months (umbilicus) upwards. The
result of this finding can be noted down by the TBA
as the number of “moons” of pregnancy, or by the
corresponding number of ticks at the appropriate
date of consultation. In this way both the TBA and
the supervisor can check whether the uterus is
growing or not.
(6,7,8) These diagrams show the position of the
fetus. A cross (X) is put at the corresponding date
for the particular position found. A clear explana­
tion of what to do in various circumstances should
be given during training.
(9) This drawing will be changed, as its meaning
is not clear. The intention was to ask the pregnant
woman if she felt fetal movements. In case of a
positive reply, it should be marked down in the
“+” column.
(10) Here the TBA should mark whether she hears
fetal heart-sounds or not.
(11) Urine analysis with Albustix is only done if
the pregnant woman has signs of oedema. Results
indicated by the number of “+” signs are marked
down as shown on the tube itself.

The inside of the card (Figure 2) gives information
about the results of the physical examination done
by the TBA.
(1) The date of each consultation is a piece of
information that has to be written. If the TBA
cannot write (in Portuguese or Arabic), someone
else in the village should be asked for help.
(2) If there are no scales in the village or if the
TBA cannot read its numbers, this column should
be left empty. However, during the supervisory visit
the nurse or midwife can mark the correct weight
here.
(3,4) The important signs of anaemia and oedema
should be looked for by the TBA, and she can be
taught how to do this. Confusion might arise as to
what is meant by “+” and
While field-testing
the card we found that for most TBAs it was
“logical” to mark the presence of anaemia in the
” column. After all, to be anaemic is a negative
thing. In the same way, the absence of oedema
should be noted in the “+” column. Instructions for
the trainers of TBAs have since been changed Care
accordingly.
The back page (Figure 3) of the antenatal card
(5) The TBA can find the height of the uterus by helps the TBA to decide what care to give.

loO
COMMUNITY HEALTH CEU

326, V Main, I Block
Korambngala
Bangalore-5t0034
India

OBSTETRICS AND GYNAECOLOGY | 77

Tropical Doctor, April 1986

I

I
52

2
o'

g

o
3

I

S
N

2:
«o

>

s

0

o

I]flu*
SHp II

i

Skiis

3

g

£
a

3

3

a

3

a

2

3.
8

c

w

3

a

n

□
o

K
to

o
o

+

©

a.
oa

a.
sa

5
*•.
a
o'
3

i

i

□
a

fel
gg

« X
«■

Tropical Doctor, April 1986

78 | OBSTETRICS AND GYNAECOLOGY

(1) Again the marking of the date of consultation
in the first column is sometimes difficult.
(2) The drawing in the second column reminds the
TBA to give appropriate nutritional advice.
(3) A cross (X) should be placed in the third column
if malaria prophylaxis has been given.
(4) The dispensing of iron tablets (with folic acid)
and/or multi-vitamin tablets should be marked in
the fourth column.
(5) The dates for tetanus toxoid immunization,
which has to be given by the supervisor twice during
pregnancy, should be noted in the fifth column.
(6) Finally, referrals to higher levels of health care
(health centre or hospital) should be noted in the
last column.
DISCUSSION

The introduction of an antenatal card for use by
illiterate TBAs is not easy. First of all, it is import­
ant to
_ discuss the idea with whoever will later teach
the TBAs and use it during their supervision in the
villages. Their commitment to using the card to
improve the performance of the TBAs is essential.
Therefore they should also be involved in the design
and content of the various drawings on the card.
Once the card has been designed, it should be field
tested by the nurses and midwives who will train
and subsequently supervise the TBAs. Changes
should be made according to their experiences.
Secondly, the antenatal card will only prove use­
ful if it can: identify the problems of a particular
pregnancy (e.g. anaemia, insufficient fetal growth,
wrong presentation), and the at-risk cases; help
TBAs interpret their findings, by giving a clear set
of instructions and standing orders on what to do
in each situation (these instructions should be
taught together with the use of the card).
These two conditions can only be met by con­
tinuously supervising the TBA’s performance and
discussing the findings with her. In this way the
card can provide a framework of things to look for
during the supervision, and so enhance its effectiveness. At the same time it can improve the TBA’s
performance by showing her any deficiences in her
skills or knowledge. Thus it can help to direct TBAs’
training, and supervisors’ attention, more specifically towards aspects that need emphasis.
Finally, the card described here needs evaluation,
At present, it is too early to suggest how successfully
it identifies women with a high obstetric risk through
the regular control of TBAs. Studies of the sensitivity of the antenatal cards (i.e. the proportion of
pregnant women with high-risk factors detected

through their use) are badly needed, not only for
the card presented in this paper, but also for the
antenatal cards used by nurses, midwives and
doctor.
This antenatal card has been a considerable help
in the process of improving the performance of
TBAs and represents a first step in making their
antenatal care more effective. Detailed studies of
this aspect of primary health care are urgently
needed to make the training of TBAs a more worthwhile activity.
REFERENCES
Chabot H T J & Savage F (1984) A Community Health Project
in Africa (slide set). Teaching Aids at Low Cost (TALC), PO
Box 49, St Albans, Herts, ALI 4AX, UK
Essex B J A Everett V J (1977) Tropical Doctor 7, 134-138
Sims P (1978) Tropical Doctor 8, 137-140
Watson D S (1984) Tropical Doctor 14, 133-135
Williams M (1980) The Training of TBAs: Guidelines for
Midwives Working in Developing Countries. Catholic Institute
for International Relations, 22 Coleman Fields, London Nl

7AF. (New edition in preparation)

Should vesicovaginal fistula
be treated only by
specialists?
J G Thornton mb dtm&h*
Chogoria Hospital, Chogoria. Meru, Kenya
TROPICAL DOCTOR, 1986, 16, 78-79

INTRODUCTION

Vesico-vaginal fistulas (VVFs) are still common in
many developing countries. The largest series ever
reported (Tahzib 1983) refers to the period
1969-80, during which no less than 1443 cases were
collected at one university hospital in Northern
Nigeria. This report is typical of most papers from
developing countries on VVFs which are usually
derived from large numbers referred to teaching
hospitals, including many difficult cases referred
after treatment elsewhere has failed. Such reports
give the impression that fistula repairs should only
be attempted by expert vaginal surgeons. Unfortunately experienced gynacologists are rare in the
rural areas of developing countries where most
fistulas occur, so the majority of repairs are probably
•Present address: Department of Obstetrics and Gynaecology,
Leeds General Infirmary, Leeds LS2 9NS, UK

0
■snERNi

CNM, MS

USE OF THE LABOR GRAPH IN MALAWI

Helen A. Burgess, CNM, MS

ABSTRACT
This article describes in detail the use of a one-sheet labor record available in every
maternity unit in Malawi. Africa. These graphs are provided by the government and
are a good method of ensuring that patients make appropriate progress in labor and
are referred to medical staff as necessary.
Malawi is a small, densely populated
country in southern Africa. The esti­
mated population of 6 million is 90%
rural. Sixty-five percent of the total
population consists of children under
14 years old and women 15 to 44
years of age. The annual growth rate
is 2.6% with infant mortality (up to
age 1) 50 per 1000 live births, and
child mortality (children under 5) just
over 50%. Total fertility is high, av­
eraging 7.7 births per woman of
completed fertility.1
There are 350 maternity units in
Malawi run by the central and local
government, religious groups, and
private individuals. Approximately
45% of deliveries nationwide occur
in these facilities. The majority of
women deliver with traditional birth
attendants. Staff in the maternity
units may be physicians (all of whom
are trained outside of the country),
clinical officers or medical assistants,
state registered nurse-midwives, en­
rolled nurse-midwives (about equiv­
alent to licensed practical nurses with

Address correspondence to: Helen A. Bur­
gess, CNM, Little Lane, Marshfield, MA 02050.
An article by Ms. Burgess entitled “Mid­
wifery at the University of Malawi” was fea­
tured in the International Exchange Column
of JNM 29(4): 1984.
Special thanks to Elizabeth Banda, rn, ms,
for advice on the manuscript

midwifery training), or nonprofes­
sional personnel. Skill and knowl­
edge vary widely among individuals
in differently staffed facilities. The use
of a nationwide labor chart is, there­
fore, an excellent way to ensure the
best possible care for the partunent
during labor and to maximize appro­
priate use of referrals.
In 1972. in Southern Rhodesia
(now Zimbabwe), R. H. Philpott de­
veloped a simple, visual labor graph
that was suitable for use in both cen­
tral hospitals and outlying units. The
graph is easy to use and is helpful for
identification of high-risk mothers
even by nonprofessional personnel
who may not understand the physi­
ologic reason for a problem, but who
are very capable of identifying it The
main feature of the graph is the use
of “alert” and “action” lines. The
alert line “. . . is a modification of the
mean cervicometric progress of the
slowest 10% of normal African primigravid patients admitted in the ac­
tive phase of labour, that is, with the
cervix dilated at least 3 cm and 100%
effaced.”2 The action line is arbi­
trarily “drawn parallel and four hours
to the right of the alert line. This al­
lows time to transfer the patient
without impairing the success of the
essential management and also al-

46
Copyright ? 1986 by the American College of Nurse-Midwives

lows many normal patients to deliver
vaginally without active interven­
tion.”3 For practical purposes, active
labor on the labor graph used in
southern Africa is defined as begin­
ning when the patient is 3 cm dilated
and 100% effaced in the primigravida, or 3 cm dilated regardless
of effacement in the multigravida. If
the patient is admitted before she has
reached 3 cm dilation, progress is
charted either on a blank graph
without an alert line or to the left of
the alert line. When she has dilated
to 3 cm, the time and dilatation mea­
surement are placed on the alert line.
If a multigravida comes in having
contractions and the head is deep in
the pelvis but is dilated <3 cm. she
also is started on the graph.
The labor chart currently used in
Malawi is based on Philpott’s work
(Table 1). All of the maternity units
in the country use the same chart,
provided by the Ministry of Health.
One side contains demographic and
admission data and a preprinted
graph for evaluating labor progress.
The reverse side has space for pelvic
assessment, information about the
second and third stages of labor, and
space for up to 6 days’ postpartum
observation of mother and infant.
This single sheet can serve as the pa-

Joumal of Nurse-Midwifery • Vol. 31, No. 1, January/February 1986
0091-2182/86/$03.50

I

II

bent’s entire chart during her mater­
is observed closely. Each horizontal
every hour, affording the care-giver
nity stay.
square represents 1 cm of dilatation.
time to assess potential problems be­
The demographic data section in­
The alert line is drawn to represent 1
fore much time has passed.
cludes the patient’s name, address,
cm per hour of progress, and the ma­
Other essential information about
a2€, gravidity, parity, last menstrual
jority of women will deliver within 7
labor is charted conveniently on top
period, estimated date of confine­
to 8 hours of initial marking on this
of or underneath the graph. Fetal
ment, by whom she was brought to
line. The action line is parallel to and
heart rate (FHR) monitoring is done
the health facility (most patients hos­
3 hours to the right for multigravidas
every 30 minutes or more frequently
pitalized in Malawi are accompanied
and 4 hours for primigravidas. The
if needed. Dips heard in the heart
by a guardian who stays with them),
heavy dotted vertical line drawn from
rate can be indicated as type 0 (FHR
and past/current pregnancy prob­
the top of the alert line represents the
irregular
all of the time), type 1 (de­
lems. Admission information in­
estimated time of delivery.
crease
with
a contraction but picks
cludes date and time of admission,
The other main feature of the
up before the end of the contrac­
date and time of onset of labor,
graphic part of the labor chart is a
tion), or type 2 (drop with a contrac­
whether labor was spontaneous or
record of fetal descent. Cephalotion and picks up after the contrac­
induced, status of membranes, ab­
pelvic disproportion is a major cause
tion has ended (see Table 3).9
normal symptoms, vaginal bleeding,
of perinatal morbidity and mortality
Patients who exhibit fetal distress
and presence or absence of pre­
in Africa.5 Descent of the head is
are turned on their left side and given
eclampsia. There also are spaces to
measured abdominally every hour
oxygen if it is available. They are
check whether or not the woman has
and plotted on the chart. Descent is
checked for cord prolapse and re­
had sleep, food, or homemade med­
measured in “fifths” of head pal­
ferred to the physician, if available,
icine (often given in the villages to
pable above the pelvic brim (Table
or to the senior-most health per­
induce labor). A section on the gen­
2). Landmarks used in describing
sonnel with midwifery or obstetric
eral physical examination follows,
“fifths” of fetal head descent are the
training. Mothers with fetal distress
and at the bottom of the chart there
fetal occiput and sinciput, and the
are given an immediate dose of 50
is a space for writing referrals, if nec­
maternal pelvic brim. According to
mL of 50% dextrose by intravenous
essary.
Myles,6 “When the head is engaged
push; then an infusion of D5W is
If the patient is in the latent phase
the biparietal diameter has passed
started. In a maternity center, if there
at the time of admission, the dilata­
through the pelvic brim, the occipital
is no improvement within 30 min­
tion is plotted on the line at 0 time
prominence can be felt only with dif­
utes, then the patient should be re­
and the actual time is charted below.
ficulty from above. The sinciput still
ferred to a hospital.4
When patients are in established
may be palpable above the brim be­
If the membranes are intact, they
labor (regular contractions at least
cause of the increasing flexion of the
are recorded as I in the appropriate
once every 10 minutes), vaginal ex­
head until the occiput reaches the
column. If the membranes are rup­
aminations are done every 4 hours
pelvic floor and rotates forwards.”
tured artificially, the time is recorded
in primigravidas and every 3 hours in
Studd7 argues that this method
on the graph. Artificial rupture of the
multiparas. If the patient has been in
is “. . . quantitative and reproduc­
membranes by the midwife is per­
the latent phase of labor for greater
ible and removes the problems of
mitted when the patient’s cervix is 3
than 8 hours from the time of ad­
variability that may be created by
cm dilated with the head at four-fifths
mission, the risk to the fetus in­
caput, moulding, or a different depth
above the pelvic brim.
creases; if the woman is in a health
of pelvis—factors that can be con­
Molding also is an important way
center, she should be transferred to
fusing in following abnormal and
to detect possible cephalopelvic dis­
a hospital.4 Once the patient’s cervix
prolonged labors.” Because the
proportion. Four degrees of molding
has dilated to 3 cm, the dilatation
head does not become engaged in
are used9: - means the bones are
and time are plotted on the active
African primigravidas until late in the
normally separated; + indicates that
phase section and the labor progress
first stage,8 accurate measurement of
the skull bones are touching each
descent is important in identifying
other; + + indicates that the skull
possible problems with cephalopelvic
bones are overlapping but on digital
Helen A. Burgess, ms. cnm. is an
disproportion.
pressure can be separated easily; and
instructor at Boston University School
With each abdominal examination
of Nursing. She is a graduate of the
+ 4-4- means that the bones are
for
descent, determination of the po­
midwifery program of New York
overlapping and on digital pressure
sition changes of the occiput and sin­
Medical School She spent the spring of
cannot be separated easily. When
ciput are noted. Determination of ab­
1983 teaching midwifery at the
molding is 2+ or 3 + , if the patient
University of Malawi.
dominal descent is less invasive than
is in a hospital, then the medical of­
descent noted vaginally, and is done
ficer should be informed. If the pa-

Journal of Nurse-Midwifery • Vol. 31, No. 1, January February 1986

47

»
i

4*
QO

Labor chart and information currently used in Malawi, originally prepared by R. H. Philpott

LABOUR CHART

PRIMIGRAVIDA/MULTIPARA

ANC:

180
160
FOETAL 140
HEART 120
100
80

LIQUOR
MOULDING

---------

10

ACTIVE

LATENT PHASE

A,

PHASE

L

9
8

7
c-

o
f=
3
B.

2,
Z

s
8

3?

5
<
2.
co

Pz
cv
□
c

I

Tj
O*

2
3
‘O

oo

O'

C
E
R
V
I
X
(cm)

/1

6

5

D
E
S
C
E
N
T

4
3
2

1
- 0
TIME
initials
5
4

3
2
1
PULSE

BP.
P1T0C1N units
drops/min
URINE vol
alb/acet
TEMPERATURE

1

2

3

4

5

6

7

8

9 10

1 1

12

1 3 1 4 1|5

1|6 1|7

1|8 19 20 2 1

22

pk

TABLE 1—Continued

o
EL

Z
•

Hospital

Show

Address

Soft tissues

Brought by

3

Age

>2

FIRST VAGINAL EXAMINATION

Name________________

£

I.

r.

No.

LMP

_

CERVIX: State
Gravida

EDD

Para

Effacement

Gestation

________________

Dilatation

Application ________________

Past pregnancy problems

MEMBRANES: Ruptured/intact
Liquor: colour

<
2w

Problems this pregnancy

p

Admitted

oT
a

I

n

CORD:

PRESENTING PART:
a.m./p.m

19___

Position: sutures and fontanelles:

Labour: spontaneous/induced
Onset------------------------ -- a.m./p.m.
Membranes ruptured

19

a.m./p.m.

19___

Level in relation to ischial spines:

Has she had food?Sleep?

CT-

3

Home-made medicines

3

Abnormal symptoms

vO
00
Oa

What is presenting:

Caput, nil/mild/moderate/severe
PELVIC ASSESSMENT

____

Vaginal bleeding yes/no amount

Shape of brim

Pre-eclampsia yes/no

Sacrum

__________________

Sacral promontory
EXAMINATION ON ADMISSION

Sacrospinous ligaments

General condition
Height
Fundus

Ischial spines

Anaemia

Oedema

weeks, and

Sub pubic arch

fingers below

Intertuberous diameter

xiphisternum. Lie
Presentation

Position

Bladder

REFERRAL

CONCLUSION ABOUT PELVIS:

F.H.:
a.m./p.m.

19

Anticipated course of labour and delivery

19

Admitted by:

Reason

Transport arrived

am. /pm

SECOND STAGE OF LABOUR: Fully dilated at
Delivery-------------------------- - a.m./p.m.

a.m./p.m.

19

Twin 2:

19____
a.m./p.m.

19___

TABLE 1—Continued
Cn

Twin 2: Method------------------------------- —

METHOD: SVD/breech/vac extr/forceps/C.S/other

o

Twin 2: Apgar 1 min.------ -------------- - ------

Apgar 1 min.----------------- —--------------- ------- m’n-

Twin 2: Live/full term/prem/SB/Mac/NND

Live/full term/prem/SB/Mac/NND

Twin 2: sex: female/male

Sex: Female/male

Twin 2: Abnormalities —

- --------------

Abnormalities

Weight:-------

Grammes (

oz)

Twin 2: Weight

Length-------

cm Head circumference--------------- cm
________ _________ First bath by

Twin 2: Length

Delivered by

lb.

Eye drops-------------------- -------Vitamin K given

Baby to nursery for observation yes/no

19

a.m./p.m.

THIRD STAGE OF LABOUR: Time of delivery of placenta------------Mode of delivery--------------------- ----------- - ------------------------------------------

ergometrine

PLACENTA: complete/incomplete. Membranes, complete/incomplete.

Weight

Cord length

Head circumference

Blood loss

cord insertion--------------------

Delivered by-------------------------- ——

AbnormalitiesChecked by-----------------

------

PERINEUM: intact/tear/cpisiotomy. Repaired by----------------------Pulse

POST-NATAL CHECKS: Immediately after delivery, B P----------

2

One hour after delivery: B.P. --------------------------------- -— Pulse

fl)

Lochia

s

2
&
5

oz.

lb.
No. of blood vessels

Condition--------------------------------------------------------- -----------------------------

oc
3

ml

19

a.m./p.m.

Urine passed

SUMMARY OF LABOUR: 1st stage

Uterus

Temp

min.

hours

min. Total-----hours

3rd stage

ml

hours

min. 2nd stage

hours

amount

PUERPERIUM
w

BABY

MOTHER

Date
c0>

Temp

Hi. of
fundus

Lochia

Perineum

NOTES

Date

Temp

Weight

Cord

Eyes

NOTES

-ri
o*

2
<2

5
00

O'

Mil

ilia

Reprinted with permission of the publisher of the British Medical Journal, Bntish Medical Association, London, England

t

TABLE 2

Level of fetal head (in fifths) above the brim
___________ 5/5
Completely above

4/5

3/5

Sinciput high
Occiput easily
felt

Sinciput easily
felt
Occiput felt

2/5
Sinciput felt
Occiput just

felt

Head floating

Nearly engaged

1/5

0/5

Sinciput felt
Occiput not
felt

None of head
palpable

Engaged

Deeply engaged

-S

— S-.

-s------

o

—-s

o

Brim

•S
0

0

S = sinciput: O = occiput.
• Adapted from: Philpott RH. Castle W* and "The Labour Graph" Procedure Manual. Kamuzu

tient is in a health center, then she
should be transferred to the hos­
pital.4
The frequency of contractions is

recorded every 30 minutes as the
number and duration in a 10-minute
period. Blocks are numbered from 1
to 5. If there are 3 contractions in 10

TABLE 3
Charting of fetal heart sounds
Normal fetal heart

180

Type 0 dip

180

160

160

140

140

120

120

100

100

0
Type 1 dip

Type 2 dip

180

180

160

160

140

140

120

120

♦

100

100
I

II

Journal of Nurse-Midwifery • Vol. 31, No. 1, January February 1986

Central Hospital. Lilongwe. Malawi, Africa?

minutes, 3 blocks are shaded. Du­
ration is represented by three types
of shading: dots represent a duration
of less than 20 seconds; diagonal
lines indicate contractions lasting 20
to 40 seconds; and very close vertical
lines represent contractions lasting
longer than 40 seconds. Spaces are
available for recording pulse, blood
pressure, and use of Pitocin every 30
minutes, and urine volume measure­
ment every hour.
The labor graph is used to help
make decisions about management
of patients. The alert line is a warning
system. Once a patient’s dilatation
moves to the right of the line, she
should be transferred so that if and
when the dilatation appears on the
action line, she will be at a facility
equipped for an operative delivery.
Patients with abnormally high fetal
heads should be transferred even if
the dilatation is progressing normally.
In a setting where the majority of
delivery facilities are in rural areas
and the most skilled people and best
equipment are located in major pop­
ulation centers, a clear, easy-to-use
method of recording labor progress
can be helpful in improving maternal

51

I

and fetal outcome. The labor record
used in Malawi is an excellent tool for
this purpose.

REFERENCES
1. Chiphangwi JD: Obstetric prob­
lems in a developing country—Experi­
ence in Malawi. J Med Assoc Malawi
13:5, 1983.
2. Philpott RH: Graphic records in la­
bour. Br Med J 4:164, 1972.
3. Philpott RH: Graphic records in la­
bour.2 p 165.
4. Kwast BE: For doctors: Guidelines
for the Management of a patient on the
labour graph, in Procedure manual. Li­

longwe, Malawi, Kamazu General Hos­
pital, 1974.
5. Philpott RH: Foetal quality pre­
served in cephalopelvic disproportion in
the primigravida. South Afr Med J
47:2021. 1973.
6. Myles MF: Textbook for midwives.
Edinburgh, Churchill Livingston, 1972. p
262.
7. Studd JW: A visual method of
charting labour: The partogram. Contemp Obstet Gynecol 18:25, 1981.

8. Philpott RH, Castle WM: Ceivicographs in the management of labour in
primagravidae. J Obstet Gynaecol Br
Commons 79:596, 1972.

9. The labour chart, in Procedure

manual. Lilongwe, Malawi, Kamazu
Central Hospital, 1974.
SUGGESTED SUPPLEMENTARY
READING
Philpott RH, et al: Obstetrics, family plan­
ning, and paediatrics. Pietermaritzburg,
South Africa, University of Natal Press.
1977.

Philpott RH: Recognition of cephalopelvic disproportion, in Clinics in Obstet­
rics and Gynaecology: Obstetric Prob­
lems in the Developing World. 9(3):609624, 1982.
EDITOR’S NOTE

This paper is published with the ap­
proval of the Government of Malawi.

1
I

52

Journal of Nurse-Midwifery • Vol. 31, No. 1, JanuaryTebruary 1986

p
The Journal of Obstetrics and Gynaecology
of the British Commonwealth
July 1972. Vol. 79. pp. 592-598.

CERVICOGRAPHS IN THE MANAGEMENT OF LABOUR IN
PRIMIGRAVIDAE
I. The Alert Line for Detecting Abnormal Labour
BY

R. H. Philpott, Sims-Black Professor of Obstetrics and Gynaecology
AND

W. M. Castle, Lecturer
Department of Social and Preventive Medicine, University of Rhodesia
Summary
Adding a particular Alert Line to the cervicographs of Rhodesian African primi­
gravidae has simply and effectively screened normal patients from those with reduced
pelvic size and inefficient uterine action. The method can be applied by midwives
working in isolated areas. This paper describes the derivation of the line and the
results of a prospective study of 624 consecutive patients, whose management in
labour was based on its use. The Alert Line would have relevance to practice
elsewhere.

Rhodesia, in common with other developing
countries, has a serious shortage of doctors.
Their distribution is such that, although the
larger cities and towns are well covered, the
rural areas have a doctor to patient ratio of only
1 : 80 000. There is, therefore, a necessity for the
medical profession to utilize midwives and
maternity assistants extensively. In this situation,
one of the functions of a University Obstetrical
Department is to provide training for the
midwife and to establish simple, accurate guide
rules.
Cephalopelvic disproportion and inefficient
uterine action are two of the biggest obstetrical
problems in the African primigravidae. In 1970,
for example, 81 new cases of obstetrical vesico­
vaginal fistula, 12 with a concomitant recto­
vaginal fistula, were admitted to the gynaeco­
logical wards of the local teaching hospital. Our
difficulty is not only that of evaluating the degree
of disproportion in the individual but also that
of teaching the midwives how best to assess
these patients in the rural areas. All primi­
gravidae are managed with a trial of labour and
a simple efficient method of conducting such a

trial, we have found, is the analysis of a cervicographic record of progress, as first propounded
by Friedman (1967). This is based on the
recognition that cephalopelvic disproportion in
the primigravidae is accompanied by inefficient
uterine action and that this will be most evident
on the cervicograph.
We have gone further and from a clinical study
have established a set of guide rules based on the
cervicograph. For the midwife working in the
periphery or the junior doctor working in the
hospital, we have constructed an Alert Line for
primigravidae at an acceptable statistical limit of
normal cervicographic progress. Should a
patient's cervicographic progress cross this Alert
Line, then arrangements are made to transfer
her to the intensive care area of the Central Unit
so that, within four hours of crossing the Alert
Line, active management can be effectively
commenced.

Method
An Alert Line must satisfy two criteria. First,
it must be simple to use. It must also separate
efficiently the majority of the normal patients
592

CERVICOGRAPHS IN LABOUR

from the abnormal patients in sufficient time to
transfer the latter safely to the Central Unit for
treatment.
We tried to establish that the rates of cervical
dilatation of our normal African primigravidae
were so similar to the pattern described by
Friedman that we could apply his curve (Fig. 1)
as a yardstick against which to measure progress
of labour in our patients. This proved not to be
the case. This was firstly because we were unable
to define the commencement of labour in our
cases, and secondly because the rate of progress
during the “phase of maximum slope” of 100
consecutive normal African primigravidae was
half that of American patients (Table I). Although
we are still studying the details, we presume that
this is because of the higher prevalence of mild
cephalopelvic
disproportion among our
“normal” patients.
Table I
Comparison of maximum rates of cervical dilatation
in normal primigravidae

Mean (cm./hr.)
Median (cm./hr.)
Mode (cm./hr.)

Friedman

This series

300
2-75
1-50

1 60
1-25
100

593

Like Hendricks et al. (1970) we decided to
designate time of arrival at hospital, rather than
the problematical time of onset of labour, as zero
time. Figure 2 shows the average of the slowest
20 per cent of Hendricks’ patients compared with
that of our own series of 100 consecutive normal
primigravidae. Again, our patients took twice as
long in the “phase of maximum slope”, although
they were delivered more quickly than Hendricks
patients. Our patients tended to arrive at hospital
later in labour.
For simplicity we hoped to be able to construct
a single Alert Line for all patients regardless of
cervical dilatation on admission to hospital, even
though cervical dilatation on admission is
directly related to progress in the individual
case. Our series of Hendricks-type graphs
(Fig. 3) based on a study of groups of at least 50
normal primigravidae arranged according to
cervical dilatation on admission showed that this
was not possible. We have, however, shown that
we can use a single straight Alert Line for all
primigravid patients arriving at hospital with a
dilatation of 3 cm. or more. The Alert Line joins
points representing 1 cm. dilatation at zero time
(admission) and full dilatation (10 cm.) nine
hours later, a rate of 1 cm. per hour. It is a
modification of the mean rate of cervical

10 -

Deceleration phase

8 Cervical
Dilatation
(cms) 6 -

Phase of
maximum
slope

4

Acceleration phase
2

Latent phase
0

2

4
6
Time from

8

10

12

14

commencement of labour

Fig. 1
Friedman's curve showing phase of maximum slope.

16

Hours

594

PHILPOTT AND CASTLE

10

0

Cervical
Dilatation
(cms)

6 -

4 -

e

&

Hendricks
This series

2

i

0

2

6

4

Time from

8

10

Hours

admission

Fig. 2
Mean cervical dilatation of slowest 20 per cent of primigravidae compared to time from admission.

10

8
Cervical
Dilatation
(cms)

6

4

2

2

6

10

Time from

admission ( hours )

14

18

22

26

Fig. 3
Mean cervical dilatation of slowest 20 per w-nt primigravidae, with patients grouped according to cervical
dilatation on admission.

CERVICOGRAPHS IN LABOUR

dilatation of the slowest 10 per cent of primigravid patients in the active phase and is slower
than Friedman’s statistical limit for “the phase
of maximum slope” of 1 - 2 cm. per hour.
Our Alert Line is both simple and efficient. It
separates, when compared to other possible
lines, the highest proportion of normal from
abnormal patients, allowing time for transfer of
those who are abnormal to the Central Unit for
active management (Fig. 4). It is a statistically
better line than the mean 95th percentile (using
Hendricks’ method) of patients, either pooled
together or separated into groups according to
their cervical dilatation on admission. Our Alert
Line compares favourably with all lines parallel
to it. We also considered applying an Alert Line
based on an individual patient's cervical dila­
tation on arrival. The chosen Alert Line is
statistically better than these alternatives for
patients arriving with the cervix already 3 cm.
dilated, and it is now being studied prospectively
on patients arriving at less than 3 cm. dilatation.

Results
The use of the Alert Line in the management
of labour in primigravidae has alerted medical
staff to potential problems at an early stage, and
midwives in this hospital, in a local Mission
hospital and in Queen Elizabeth Hospital,
Blantyre, are employing the line correctly and
enthusiastically.
We present a prospective study of 624
consecutive primigravid patients who were found
to have a cervix that was already 3 cm. dilated
on admission. The only patients excluded from
this study were those with abnormal presenta­
tions (transverse lie, breech and brow), placenta
praevia, multiple pregnancy or eclampsia. For
the purpose of the study, patients were divided
into three main clinical groups (Fig. 4). Group 1
included all patients who were delivered before
their cervicograph reached the Alert Line,
Group 2 those whose cervicograph crossed the
Alert Line but were delivered before it crossed a
line parallel to the Alert Line, arbitrarily drawn

10

8 Cervical
Dilatation
(cms) 6

3

4

2

0

2

4

595

6
8
Time ( hrs )

10

Fig. 4

The Alert Line on the cervicograph, showing the clinical subgroups: Group 1:
Those delivered before cervicograph reached the Alert Line. Group 2: Those
whose cervicograph crossed the Alert Line but who were delivered before it
reached Action Line. Group 3: Those whose cervicograph crossed the Action Line.

596 PHILPOTT AND CASTLE
Table II
Average of various parameters for patients whose cervicograph did not cross (Croup /) or did cross
(Groups 2 and J) the Alert Line

Parameter

Group 1
(488 patients)

Groups 2 and 3
(136 patients)

Rate of cervical dilatation after 4 cm.
Area of brim
Transverse diameter of brim
True conjugate
Dilatation of cervix on admission
Head level on admission

1 -75 cm./hr.
100-47 sq. cm.*
11-85 cm.*
10-75 cm.*
5-20 cm.
3 30
(fifths of head
above brim)
18-29 years
1- 578 m.
2- 989 kg.

0 44 cm./hr.
87-33 sq. cm.+
10-90 cm.+
10- 17 cm. +
3 -84 cm.
4-08
(fifths of head
above brim)
18-34 years
1-571 m.
3 051 kg.

Age
Height
Birthweight

Statistical
conclusion

p<0 001
p<0 001
p<0 001
Not significant

p<0 001
p<0 001

Not significant
Not significant
Not significant

* 15 patients, f 60 patients.

four hours later and called the Action Line, and
Group 3 consisted of all those whose cervicograph crossed the Action Line.
Table II records the details of the differences
in various parameters between patients who
were delivered before their graph reached the
Alert Line (Group 1) and those whose graph
crossed the Alert Line. The scale which we have
used to measure the level of the fetal head was
that first described by Crichton (1962) and is
illustrated in Figure 5. To determine whether
there are any other factors at the time of crossing
the Alert Line which would guide the midwife
as to outcome, a comparison has been made
between patients in Group 2 and those in
Group 3 (Table III). Table IV shows the delivery
outcome in each of the subgroups. The patients
in Group 3 and their management have been
discussed elsewhere (Philpott and Castle, 1972).
Table III
Comparison offactors in cases separated by the
Alert Line between Group 2 and Croup 3
Significant Factors:
1 Rate of dilatation after 4 cm. (p< 0 001)
2 Dilatation at Alert Line (p<0 01)
3 Level of the head at the Alert Line (p< 0 05)
Nonsignificant Factors:
1 Age
2 Height
3 Dilatation on arrival
4 Level of head on admission
5 True conjugate
6 Transverse diameter of brim
7 Area of brim

Table IV
Delivery outcome in each of the three subgroups

Group 1 Group 2 Group 3
Spontaneous delivery
438
Spontaneous delivery after
oxytocic stimulation
0
Vacuum extraction
48
Caesarean section
2

54

0

0
14
0

35
14

Totals

68

68

488

19

Discussion
A study of the parameters in Table II shows
that our Alert Line separated efficient from
inefficient labour as reflected in the rate of
dilatation of the cervix. Furthermore, the Alert
Line separated patients with small pelvic size
from those of adequate size, as shown in the
smaller brim area and transverse brim diameter
of those whose graphs crossed the Alert Line.
This is of major importance to those without
X-ray facilities.
Other differentiating features that were
statistically significant, though of less practical
value, were the cervical dilatation on admission,
which in this series probably reflected the
efficiency of labour prior to coming to hospital,
and the level of the fetal head on admission.
Table JI shows that in the African primigravidae
the head is not engaged at the onset of labour,
and in fact does not become so until late in the
first stage.

1
CERV1COGRAPHS IN LABOUR

597

5/5 t/5 3/5 2/5 1/5 0/5
5

0

AO'og oo

COMPLETELY SINCIPUTHIGH
ABOVE
OCCIPUT

EASILY FEU

SINCIPUT
EASILY FELT

OCCIPUT

FELT

SINCIPUT

SINCIPUT

OCCIPUT

OCCIPUT

FELT

JUST FELT

FELT

NOT FELT

NONE OF
HEAD
PALPABLE

Fig. 5
The level of the fetal head measured by abdominal palpation and expressed in terms of fifths above the brim.
S = sinciput, O = occiput.

Recognition that the chosen Alert Line is
applicable from 3 cm. dilatation of the cervix
onwards fits in with Friedman’s differentiation
between the latent and active phases of the first
stage of labour. We are at present studying
those patients admitted in the latent phase by
applying the Alert Line to their cervicographs
immediately after cervical dilatation of 3 cm. or
more.
Although in patients seen in Aberdeen by
Bernard (1952) the outcome of labour was related
to maternal height, in our patients there was no
relationship between maternal height and
dysfunctional labour. The probable explanation
is that our patients had an average height of
under 1 - 6 m., and this means that we cannot use
maternal height as a screening method for
planning hospital deliveries for Rhodesian
Africans.
Analysis of the vacuum extractions carried out
in Group 1 showed that these were more simple
than those extractions performed in Group 2 and
could therefore be carried out by experienced
midwives. Of the two Caesarean sections in

Group 1, one was for a prolapsed cord and the
other was for severe intrauterine infection on
admission. Both these cases would have been
sent to a Central Unit in any event. Thus, in the
absence of other complications, patients whose
cervicographs keep to the left of the Alert Line
may be safely managed in the peripheral unit.
Table IV also reveals that the cervicograph of
22 per cent of patients crossed the Alert Line.
Outside the Central Unit these would be
considered for referral and, in fact, half of them
were delivered normally within the following
four hours, while the rest were the problem cases
requiring skilled attention.
Once the cervicograph of a patient has crossed
the Alert Line, the only statistically significant
factors guiding the midwife as to whether the
patient would still have a normal delivery
(Group 2) or would require intervention (Group
3), are, of course, the rate of cervical dilatation
and also the level of the head at that time
(Table III). The difference in head levels was,
however, minimal and of little or no practical
value.

598 PHILPOTT AND CASTLE
There would seem to be a need in other parts
of the world for a simple screening guide line
such as the Alert Line to be used by midwives,
doctors in general practitioner units and house
surgeons in larger hospitals. As our application
of the Alert Line after 3 cm. dilatation of the
cervix coincides with Friedman’s active phase,
and it is only slightly slower than his statistical
limit for that phase of labour, we believe that it
could have universal application in the management of primigravidae.

Acknowledgements
We thank the two Midwifery Research
Assistants, Miss I. Edmonstone and Mrs. D.

Ritchie, who carried out the clinical studies
reported in this paper; also the Secretary for
Health, Dr. Mark Webster, for clinical facilities.
References
Bernard, R. M. (1952): Edinburgh Medical Journal, 59, 1.
Crichton, E. D. (1962): Paper read at Congress of the
South African Society of Obstetricians and Gynae­
cologists.
Friedman, E. A. (1967): In Labor, Clinical Evaluation and
Management. Appleton-Century-Crofts, New York,
p. 27.
Hendricks, C. H., Brenner, W. E., and Kraus, G. (1970):
American Journal of Obstetrics and Gynecology, 106,
1065.
Philpott, R. H., and Castle, W. M. (1972): Journal of
Obstetrics and Gynaecology of the British Commonwealth, 79, 599.

Q

7

The Recognition of Cephalopelvic
Disproportion
R. H. PHILPOTT
The problem of definition
The definition of cephalopelvic disproportion (CPD) should be ‘the failure
of the fetus to pass safely through the birth canal for mechanical reasons .
The mechanical reasons referred to in this definition are the relative size of
the maternal pelvis and the fetal presenting part. The difficulty is that these
two components each vary from one labour to another in their threedimensional size and shape and also in the degree to which the fetal head
may be permitted to undergo compression without immediate or later
damage to the fetal brain. Each of these factors are difficult to quantify and
even more difficult to put together as one measurement.

The causes of CPD
The commonest cause of CPD in the developing world is a contracted pelvis
with an average-sized fetus. The pelvis may be contracted for genetic or
nutritional reasons or as a result of trauma. In some parts of the developing
world women are in a stage of transition. During their childhood and
formative vears poor nutrition led to poor pelvic growth. In their repro­
ductive years there may have been improvement in socioeconomic con­
ditions in their community and fetal growth has been optimal. This has led to
a considerable prevalence of CPD. In the Shona women of Zimbabwe the
mean pelvic brim area of primigravidas who at the end of a trial of labour
were found to have CPD. was 83.4cm-’ and the mean birthweight was 3118g
(Stewart. C owan and Philpott. 1979). The mean value for Shona women
who delivered spontaneously without CPD was 102.2cm-’ whereas in
English women studied by Ince and Young (1940) the mean brim area was
126.8 cm-, in the series presented by Stewart et al the main pelvic
abnormalitv was either a small gynaecoid or a platypelloid pelvis.
Sometimes the cause of the CPD is a large baby with an average-sized
pelvis. This may even occur in a multipara who has previously had normal
deliveries. There is also evidence that some women of high parity develop a
forward subluxation of the lumbar vertebrae thus reducing the antero­
posterior diameter of the pelvic brim compared with the diameter earlier in
their reproductive years. This can lead toCPD with a small baby after having
( littics in Obstetrics and Gynaecology—Vol. 9. No. 3. December 19X2
(l3O6-3356/X2/U9<)3-bO9$03.U0 Dlys- w B Saunders Company Ltd

609

610

R. H. PHILPOTT

delivered larger babies successfully in the past. Other fetal causes for CPD
are hydrocephaly or a malpresentation such as a brow or mentoposterior
position.
In parts of the world where absolute CPD is not prevalent, relative CPD
due to occipitoposterior positions is probably the commonest cause of CPD.
Here the fetal head is not too large to pass through the pelvis in an anterior
position, but with the larger presenting diameters of the fetal head in
occipitoposterior positions CPD can occur. It can cause delay in the first
stage of labour in the primigravida but this will usually respond well to
oxytocic augmentation of uterine activity as long as there is no absolute
CPD. The evidence of absolute CPD would be increasing moulding of the
fetal head with failure to descend. Where contracted pelvis is common,
occipitoposterior positions may be seen more frequently than occipito­
anterior. In Zulu primigravidas the commonest position in the first stage of
labour is left occipitoposterior with the occiput at a four o’clock position in
the pelvis. In these circumstances little attention is paid to the position of the
head during the first stage, and delay is managed according to whether or not
there is absolute CPD and not according to the position of the head.

The prevalence of CPD in different parts of the world

This is a condition that is seldom seen in the developed world and, even
when it is diagnosed, the problem is more often inefficient uterine action
than true disproportion. In the developing world the prevalence of CPD
differs greatly from country to country and even between tribes within a
country.
The complications and efTects of CPD
These are maternal and fetal and they tend to differ between primigravidas
and multigravidas. They are detailed in Chapters, on ‘Obstructed Labour’.

The problems of over and underdiagnosis of CPD

This will be a particular problem in countries where CPD is common for, not
only will major CPD be prevalent (requiring caesarean section), but also
borderline CPD will be prevalent (which may or may not require ceasarean
section). An easy solution is to be liberal with caesarean sections, both
electively when CPD is suspected before the onset of labour, and as an
emergency during labour whenever there is delay in progress, or fetal
distress, with a suspicion of CPD. This approach can lead to maternal risks,
for caesarean section carries a higher mortality and morbidity rate than
vaginal delivery and there is always the risk of rupture in a subsequent
pregnancy.
The other solution is to practise active management in trial of labour for
all primigravidas. This is good management as long as the attendant is skilled
in observing fetal condition throughout the labour. If this cannot be ensured
then there will be a high price to pay in fetal cerebral birth trauma and
asphyxia.

r

F

I

THE RECOGNITION OF CEPHAl.OPI I VIC DISPROPORTION

611

The skill of obstetrics in these circumstances is to keep the caesarean
section rate as low as possible while ensuring good fetal quality.

An assessment of the different parameters used in diagnosing CPD
There is no one measurement that will diagnose CPD and each of the
following need to be taken into consideration.

History. Information regarding previous skeletal injury or disease would
suggest the possibility of contracted pelvis. In a multigravid patient a
detailed history of previous labours will be a valuable guide, particularly if
the labours were conducted under the supervision of an accurate observer
and recordings were detailed and accurate.

Height and external measurements. Bernard (1952) showed that, in a study of
labour in a group of Scottish women, the degree of mechanical difficulty in
labour was inversely proportional to the patient’s height. In the study of
Shona and Zulu women done by Stewart, Cowan and Philpott (1979) the
mean height of women with major CPD was less than that of those with no
CPD. but there was a wide range, and so the individual patient’s height
could not be used as a predictive factor. In Shona women with major CPD
the mean height was 151.32 cm and in the controls who delivered without
CPD the mean height was 156.8cm. Likewise, foot and external measure­
ments. though smaller on average, had a wide range and could not be used
for prediction or even screening purposes.
Abdominal and pelvic assessment. In the Caucasian primigravid patient the
feta! head is normally engaged in the pelvis from the 38th week of pregnancy
and a high head at term can be an indication of CPD. In the African patient
the head is often above the brim of the pelvis until late in the first stage of
labour and then descends quickly and easily. It is thought that this is due to a
high angle of inclination of the pelvic brim (Stewart, Cowan and Philpott.
1979) though others (Briggs. 1981) believe there is no difference in the angle
of inclination and believe that it is due mainly to increased head circum­
ference.
The head-fitting test of Munro Kerr can be used, pushing the head into the
pelvic brim with an abdominal hand while assessing descent with two fingers
of the right hand in the vagina and the thumb above the symphysis pubis. If
this test is positive, it is valuable. If it is negative, it may be due to many
factors other than CPD. including poor formation of the lower uterine
segment.
Digital pelvic measurement in late pregnancy is less accurate than most
realize or will admit. If all patients were to labour under supervision, with
immediate access to caesarean section if needed, it would be better to drop
pelvic assessment in late pregnancy and depend on a trial of labour.
However, such facilities are not always available and it is necessary to make
an attempt at screening in late pregnancy those primigravidas who have

R. H. PHILPOTT

612

small pelves and arrange for them to have a trial of labour in a maternity
institution.
The one measurement that is more reproducible and reasonably accurate
is the diagonal conjugate of the brim. Accuracy can be improved by touching
the sacral promontory, marking off the point on the proximal part of the
index finger that touches the under surface of the symphysis and then
measuring that distance on the fingers with a ruler. The true conjugate can
be estimated by subtracting 1.5 cm from the diagonal conjugate. If the true
conjugate is less than 10cm, then CPD must be suspected. To improve
clinical skill, whenever a patient is going for an X-ray pelvimetry, assess the
diagonal conjugate digitally first and compare the findings. Other clinically
estimated measurements of the pelvis, particularly the transverse measure­
ments. are of such poor accuracy as to be of little value.

X-ray pelvimetry
Suitable facilities and technical expertise to carry out x-ray pelvimetry are
often not available. Furthermore, in a community where CPD is prevalent,
pelvic measurements on their own tend to influence the obstetrician to do
more caesarean sections than are necessary. There is really no indication to
do an antepartum x-ray pelvimetry on a primigravida who has no other
problem than suspected CPD. for she is best managed by a careful trial of
labour. Even when there is delay in progress in labour in the primigravida it
is not difficult to recognize, at that stage, the patient with major CPD who
requires an immediate caesarean section. The rest are best managed by a
carefully controlled oxytocic augmentation of labour, which will reveal any
evidence of hitherto unrecognized true CPD before there is any harm to the
mother or fetus. This is a more accurate, yet safe, means of assessing CPD in
the primigravida than the use of intrapartum x-ray pelvimetry.
The situation is different with the multigravida. If there is delayed pro­
gress in the first stage of her labour ft is imperative that even minor degrees
of CPD are detected, for oxytocic augmentation of labour in those circum­
stances can lead to uterine rupture (this virtually never happens in the
primigravida). Here is a valuable place for intrapartum x-ray pelvimetry. If
the head is still high with an average-sized baby and x-ray pelvimetry
measurements show a true conjugate diameter of less than 10.5 cm, oxytocic
augmentation of labour would be unwise and caesarean section would be
indicated.
The indications for antepartum x-ray pelvimetry are in the patient who
has suspected CPD plus another obstetric problem that would
contraindicate a trial of labour, for example cardiac disease, a previous
caesarean section scar, a malpresentation (e.g.. a breech) or a multigravida
with a history of previous pregnancy loss from a long or difficult labour.
TRIAL OF LABOUR

For the reasons given above the ultimate method of assessing CPD in the

THE RECOGNITION OF CEPHAI.OPEI.VIC DISPROR>R I ION

613

otherwise uncomplicated primigravida will be by trial of labour. Comment
w ill be made about CPD in the multigravida later, and this section will be
devoted to the primigravida only. Having said this, in parts of the world
where CPD is prevalent (and possibly everywhere else too), every
primigravida should be regarded as having possible CPD until she is safely
delivered. For this reason every otherwise uncomplicated primigravid
labour is managed as a trial of labour. The only primigravidas who would not
have a trial of labour would be those with malpresentations such as breech
and transverse lie. those with evidence of placental insufficiency or severe
maternal disease, for example cardiac disease or severe hypertension.
Ideally, a trial of labour should be conducted in hospital where clinical
expertise and facilities for operative delivery are available. However, in
parts of the world where CPD is most common, hospital delivery is not
feasible for all primigravidas. In those circumstances, patients can be
observed in labour in a peripheral midwife-run clinic and with experience
and use of the partogram. problems can be detected early and patients
transferred to hospital in good time, while the rest deliver normally in the
clinic.
Definition

A trial of labour is conducted in a patient in whom CPD is suspected, in
order to determine whether it is safe for the patient to deliver vaginally or
not. The trial continues as long as there is no fetal or maternal distress (that
cannot be corrected) or delay in the progress of the labour. Each of these
determinants in a trail of labour need to be defined. They can each be
recorded on a partogram.

The use of the partogram
There are many variants of the partogram. but essentially they enable one to
record the three main determinants of a trial of labour, viz. fetal condition,
labour progress and maternal condition. Figure 1 illustrates an example of
one currently in use. The graphic recording of labour enables the midwife or
doctor to recognize deviations from normal labour, particularly when an
Alert Line is used as a reference. The partogram is valuable in any labour,
but particularly for the midwife managing a labour in a peripheral clinic.
When she finds it necessary to transfer a patient to a referral hospital the
partogram should go with the patient to facilitate continuity of management.
In addition to the recordings on the partogram there should be a separate
admission examination record and also a four-hourly, ot more frequent,
evaluation of the recordings of fetal condition, labour progress and maternal
condition. This evaluation should always be followed by a written decision
on the management indicated by the evaluation.

Assessment of fetal condition

This can be assessed with different degrees of sophistication, depending on
facilities available. There is little doubt that the cardiotocograph and fetal

614

R. H. PHILPOTT

aWMM !••••>»••»••••»»I

Ml«

atCXMD 0* I *MXJt

i

6 mo*xs

i

bA*.fc A B

X2.

M»«anaa«i ••«

««|MW

MW*UtlHlU»
V*h MV*LLIM.
I

n11 /u

■'"■

;^r a

■fj

Pq*T &A.TC S

■ u ■ H , * ,; i i i ■ - >-

!
*

:j

l—ipLU-i-en
hiss

=

’ f t < ‘ T1T11 t T T i

?<■■■■■—■—wra—■—■■■■■■■■■I

* «*«M

nzn ;J±1
-it
!

fl

*rt: •1 -TH-?

:

-T

■

‘: 1

- -------- ,

i-

f •

?

‘

H- 1

I

-)■

5

i

i: H ; 1-1 t1 i
“ tinr.
it

.JU

>t 4 /W
rnTTT

p:!::

p

■■ i
■ ■ 1.. 1,4 r..T
I—

Sfe

I 1 I; ' ! P~7ttP^77t7~~

i

|tHt

a

J

I

3

E

*

•--4t -+h4
Wi

i

1

==S

tote
ijfi? 5^

■p s

tin
r-w

? 5
I ;
! >
! =

r

5

HSt'4j1

3/oQ ~4 [jcrxs^L !£ j
|:= io |:=
I*- /o »X.
Figure I.

------ i------------ *--4-------

J—I—J-

: "Sfc-1 r--=z, itoo

|

■

22 j

<(J.

Example of a composite labour graph

I
II

THE RECOGNITION OF CEPHALOPELVIC DISPROPORTION

615

scalp blood sampling enables the observer to assess fetal asphyxia with
greater degrees of accuracy. It is also true that, in a trial of labour for
suspected CPD, where there has been no previous placental insufficiency, a
trained observer can assess fetal condition very well without this equipment.
Doctors and midwives working in busy referral hospitals in the developing
world where a few cardiotocographs. though not sufficient, are becoming
available, face a very particular problem. It is easy to ’become overdependent on the few monitors and lose one’s skills when one has to care for
the larger number of patients for whom monitois are not available.
In these circumstances a few principles are worth mentioning. Firstly, if
money does become available to purchase equipment (and other basic
health care needs have been met) the first item to be purchased should be
equipment for measuring the fetal scalp blood pH. This will help avoid the
tendency to over-diagnose fetal asphyxia if only the cardiotocograph is used.
Before purchasing the very expensive cardiotocographs there is greater cost
benefit in purchasing simple ultrasound fetal heart detectors which cost
about one-tenth as much as a cardiotocograph. When the first cardio­
tocograph is purchased it should be regarded as a teaching aid before it is
used as a service monitor. (In practice we use them firstly as antepartum
monitors, and only if there is unused time on the monitor do we use them in
labour.) In labour they should be used to teach doctors and midwives to
detect the different fetal heart rate patterns clinically, so that they do not
become totally dependent on the monitors. If the number of monitors
available is insufficient to monitor all high-risk labours continuously, then it
is a good idea to monitor a number of patients for 20 to 30-minute intervals in
rotation on the one monitor.
Having made these comments with regard to the use of monitors, the truth,
of the matter is that, for the foreseeable future, the assessment of Ictal
condition in labour in the developing world will be dependent entirely on the
clinical skills of the observer, without the availability of monitors. These
skills need to be refined and eventually challenged by the frequent review of
perinatal outcome. Postmortems need to be carried out on all perinatal
deaths to reveal the evidence of gross fetal asphyxia and cerebral birth
trauma and neonatal and infant follow-up needs to be done to reveal lesser
degrees of labour pathology.
The fetal heart rate should be listened to before, during and after a
contraction to elicit abnormalities in the base-line rate and to time any
decelerations in relation to the contractions. As head compression increases
and there is reduced placental blood flow with contractions there will be
early decelerations with a gradually increasing base-line rate. Then the
decelerations become more prolonged and an advanced feature of fetal
distress in CPD is a deceleration that starts early in a contraction and persists
long after the contraction is over. This is in effect a combination of an earls
head-compression deceleration and a late placental insufficiency deceleation (Stewart and Philpott. 1980). Only a cardiotocograph can detect the
other sinister fetal heart rate change of loss of beat-to-beat variation.
Seldom is there significant CPD in labour without the presence of
meconium in the liquor. It is interesting, and difficult to explain, that in the

616

R. H. PHILPOTT

presence of CPD. the meconium often appears in the liquor before the fetal
head has been driven sufficiently into the pelvis to be compressed. The
combination of meconium in the liquor plus an abnormal fetal heart rate
pattern is sinister not only because it is evidence of fetal asphyxia but it can
lead to meconium aspiration at delivery, which is a not uncommon cause of
fetal death in CPD.
A further sign of fetal distress in CPD is increasing moulding of the fetal
head. This needs to be measured accurately. It usually appears first at the
occipitoparietal junction to be followed later by overlap at the parietoparietal junction. We score the moulding at each junction out of three and
add the two together to give a score out of six. The bones will be apart when
there is no moulding (0). When they touch the score is one. When they
overlap but can be reduced by digital pressure the score is two. If the overlap
cannot be reduced the score is three. If there is an increasing moulding score
without head descent this is a sign of serious CPD and of developing head
compression.
Work done by Wilson et al (1979) has shown that fetal electroencephalo­
graphic changes that persist after delivery do not occur unless the fetal pH
drops markedly or drops below 7.25 during labour. Everything possible
needs to be done to detect early evidence of fetal distress to prevent cerebral
birth trauma.
Before making a final decision with regard to fetal condition, and whether
or not to intervene, it is important to take into account all the factors of that
labour and not to act on one parameter. Each finding must be evaluated
carefully. For example, an isolated late fetal heart rate deceleration is of no
importance whereas persistent late decelerations that are getting worse, plus
meconium staining of the liquor, warrant a fetal scalp blood sample and pH
measurement with a view to considering an early delivery of the fetus. The
method of delivery will then depend on the cervical dilatation and the degree
of CPD.
Assessment of maternal condition

This is not so difficult. Evidence of maternal distress is a pulse rate that is
rising above l(X)bpm. a rising temperature or blood pressure, or the
presence of protein or acetone in the urine. These signs are only meaningful
if the mother is being properly managed.
Assessment of labour progress

This is done mainly by recording and interpreting the rate of dilatation of the
cervix and the rate of descent of ti’.e presenting part. The increasing momen­
tum of the uterine contractions is also of importance but not as easy to
measure accurately by clinical means.
The first stage of labour needs to be studied separately in its latent and
active phases (Friedman. 1978). By definition the active phase commences
when the slower rate of cervical dilation in the latent phase changes to a
faster rate of dilatation. This can only be determined in retrospect and varies
slightly from one labour to another. In the majority of labours the active

THE RECOGNITION OF CEPHA1.OPELVIC DISPROPORTION

617

phase commences when the cervix is 3cm dilated and fully effaced and in
practice we use this as our guide for the commencement of the active phase.
However, if it is not recognized that this is an arbitrary division, mistakes can
be made by diagnosing a delayed active phase when the patient is still in the
latent phase. Too early augmentation of the latent phase may cause more
problems than it solves.
There is value in recording the changing effacement of the cervix on the
cervicograph. in addition to the cervical dilatation. We thicken the vertical
line on the cervicograph according to the length of the cervix in centimetres
to record effacement. In this way progress in cervical effacement can be seen
at a time when there is still minimal progress in cervical dilatation.
We draw guidelines on the labour graph to alert the attendant to delay in
labour (see Figure 1). These must only be regarded as guidelines and
intelligent flexibility must be allowed, particularly in the latent phase. The
normal latent phase has a very wide range of variation. Another problem is
the difficulty in determining the onset of labour in a patient who is admitted
after being in labour for some while. The pattern of admission in labour in
relation to cervical dilatation varies from community to community. In our
practice, if a patient has been more than eight hours in the latent phase since
admission, this warrants reassessment. Thus, if a patient is admitted at the
clinic or hospital and is found to be in the latent phase, we record the
dilatation and effacement at zero time on the left-hand side of the labour
graph. We re-examine the patient four hours later. If she is still in labour,
but there is no change in the cervix, we give 100 mg pethidine to test whether
this is false labour. She is then re-examined four hours later (eight hours
after admission). If she has gone out of labour we regard this as false labour
and await the onset of true labour. If she has progressed to the active phase
we transfer her recordings to the active phase section of the graph. Her
cervical dilatation recording is then recorded on the appropriate point on the
Alert Line and all subsequent recordings follow from there. If at the eighthour mark she is still in labour, but there has been no progress in cervical
dilatation, we regard this as delay in the latent phase.
Delay in the active phase of the first stage of labour is more serious. This is
diagnosed by evidence of delay in the progressive dilatation of the cervix
ind/or failure of descent of the presenting part. In the primigravida the
mean rate of cervical dilatation in the active phase is 3 cm per hour and the
slowest rate of normal progress can be regarded as 1cm per hour. It is
valuable to draw a reference line on the cervicograph representing this
lowest limit of normal progress. We draw a line at 1 cm an hour and call it the
Alert Line. If the patient is found to be in the active phase on arrival, or as
soon as she is found to have moved from the latent to the active phase, we
plot that dilatation on the Alert Line. Any subsequent deviation to the right
of (hat Line alerts the attendant to a potential problem. The Alert Line is of
particular help to midwives in peripheral clinics. Depending on the distance
from the clinic to a hospital, the midwife re-examines the patient and if two
hours later there is still delay, intervention and therefore transfer is
indicated. We thus have a second reference line on the cervicograph two
hours to the right and parallel to the Alert Line and call this the Action Line.

f

618

R. H. PHILPOTT

If the patient is being managed in hospital it is still safe and worth waiting
until delay reaches the Action Line as many pick up and progress satis­
factorily on their own after crossing the Alert Line and before reaching the
Action Line. The use of oxytocin has complications and there is no point in
giving the busy midwives more work than necessary.
The rate of descent of the fetal head is also an index of labour progress,
particularly in the late active phase. There is minimal descent in the latent
phase or even in the active phase until the cervix has reached about 7cm
dilatation. Thereafter, descent is more rapid and is completed in the second
stage of labour (Friedman, 1978). The most meaningful method of recording
the"level of the head is by the number of fifths of head above the pelvic brim
(Crichton. 1974), as this, with the degree of head moulding, is the ultimate
index of CPD.

Management of a trial of labour
■ Maternal position in labour does affect the efficiency of uterine activity. The
supine position in bed imposed for so long by Western medicine is un­
doubtedly the worst position. The patient should be encouraged to be
ambulant for as long as she is comfortable and. if she does lie down, to lie on
her side. She should be given psychological support and encouragement,
which really means companionship. This can be provided best by the
husband, though in some cultures this is an unfamiliar concept and the
grandmother takes his place.
Analgesia needs to be considered, as unrelieved pain can have a
deleterious effect on uterine action. This is discussed in Chapter 11. An
epidural block can be of particular value in a trial of labour. It should not be
given too early as it can cause delay in the latent phase. It is of greatest value
at the time of delay when oxytocic augmentation is introduced. Adequate
hydration needs to be maintained, particularly in hot climates, when the
labour lasts more than eight hours and if there is much vomiting. Then,
intravenous dextrose in water is indicated.
Management of delay in the latent phase

Friedman (1978) regards the prolonged latent phase as ‘common but
innocuous’ and that the prognosis for vaginal delivery of an unaffected
infant is good. Cardozo et al (1982) practised active management of delay in the latent phase. They ruptured the membranes and augmented labour with
oxytocin if progress in cervical dilatation strayed more than two hours to the
right of their nomogram. In the group of primigravidas who showed a
prolonged latent phase (in which the latent phase from examination on .
admission to 3cm dilatation exceeded six hours) they found a 16.7 per cent
caesarean section rate (compared with 1.6 per cent in those whose progress
was to the left of the nomogram) and the babies had the highest incidence of
low Apgar scores at one minute compared with outcome in all the other
types of dysfunctional labour. In reviewing their results they felt that they
would need to consider a more conservative approach or to monitor intra­

I

!

I HE RECOGNITION OF CEPHAI OPELVIC DISPROPORTION

619

uterine pressure and so determine oxytocic augmentation more accurately.
Friedman recommends therapeutic rest in the first instance, using mor­
phine. and only if the condition persists after the period of rest does he
advocate oxytocic stimulation.
Our advice to midwives in peripheral clinics is to be conservative in the
management of a prolonged latent phase. If the patient has made no
progress four hours after admission in the latent phase they give 100mg
pethidine to separate those who are in false labour. If the patient is still in the
latent phase at eight hours after admission with no progress in effacement or
dilatation of the cervix we advise transfer to hospital. On arrival in hospital,
if there is no evidence of CPD we rupture the membranes and give oxytocin.

Management of delay in the active phase
Friedman (1978) differentiates between protraction disorders and arrest
disorders. He defines protracted active dilatation as a cervical dilatation rate
of less than 1.2 cm per hour in the primigravida or less than 1.5 cm per hour
in the multigravida in the active phase of labour. He recognizes a number of
contributory factors including CPD. excessive sedation and fetal mal­
position. He finds that it is not possible to improve the dilatory pattern by
any of the modes of treatment known today. After excluding patients with
CPD (they have a caesarean section) he advises supportive care for the rest
and at the end of their prolonged labour tries to avoid instrumental delivery,
as this worsens the outcome.
Cardozo et al (1982) had a very similar group and labelled this group
primary dysfunctional labour. This was when the active phase ot labour
progressed at less than 1 cm per hour when a normal active phase had not
been established. Of these. 80 per cent responded to oxytocin but in the
second stage of labour the instrumental delivery rate was two times higher
than normal. Of those who did not respond to oxytocic augmentation, the
majority had CPD and 77.3 per cent of these required caesarean section.
Friedman’s second group of patients with delay in the active phase were
described as having arrest disorders. Secondary arrest of dilatation occurred
when cervical dilatation stopped in the active phase over a two-hour period.
( PD was the commonest aetiological factor in these patients and so made
the condition particularly ominous. Other factors that often occurred in
combination with each other or with CPD were excessive sedation and fetal
malpositions. Friedman advised a thorough search for CPD when this
abnormal labour pattern was recognized. Those with CPD had an
immediate caesarean section and the rest had oxytocic augmentation ot
labour. If the oxytocic augmentation was adequate the majority of the latter
group delivered vaginally without assistance.
Cardozo et al (1982) found that, in their series, secondary arrest of
dilatation was more often due to malposition or deflexion of the fetal head
wliich could be corrected by increased uterine efficiency.
We find that frequently it is not possible to differentiate between pro­
traction and arrest patterns of dysfunctional labour and. furthermore, the
aetiological factors are the same even if they are seen in different propor-

620

R. H. PHILPOTT

tions in the two conditions. We are supervising 26 000 deliveries per year,
with a large proportion looked after by midwives in satellite clinics, and only
the problems referred to the referral hospital. Guide rules need to be as
simple as possible, aiming to keep the normal deliveries in the peripheral
clinics and transferring the problems to the hospital in sufficient time to
ensure a satisfactory outcome for mother and baby. Protocols of manage­
ment for those with delay in labour need to be clear as the referral hospital in
the developing country tends to be extremely busy and short-staffed.
We diagnose delay in the active phase whenever progress in cervical
dilatation reaches the Action Line. The midwife in the clinic will prepare to
transfer the patient whenever delay has reached the Alert Line so that the
patient reaches expert attention in the hospital by the time cervical progress
has reached the Action Line. This is a good name for this point in the
progress of labour for, apart from fetal or maternal distress, intervention
before this time is unwarranted and at this time expert attention is man­
datory if there is to be no compromise in fetal and/or maternal condition.
If delay in the active phase is diagnosed, then, before deciding on specific
treatment, an accurate and thorough assessment by an experienced observer
is absolutely vital. This is one of the most decisive points in a woman’s
labour. If a mistake in judgement is made at this point then the
consequences are fetal damage or fetal death. The labour record thus far
must be assessed carefully and this underlines the fundamental importance
of good observations and rcordings. The fetal condition must be assessed
accurately and if there is clear evidence of fetal asphyxia or CPD at this
stage, then a caesarean section is indicated. Fetal distress will be measured
clinically by listening to the fetal heart rate in relation to contractions, noting
meconium and head moulding. If facilities for cardiotocographic heart rate
recording and fetal blood sampling are available these will help in the
decision.
The diagnosis of definite CPD at the time of delay is made when the head
is still high (three to four-fifths above the brim) with increasing moulding
and no progress in the descent over the past few hours. Additional
information can be obtained by doing an x-ray pelvimetry if this is available.
An erect lateral view is possibly the only view necessary as a good true
conjugate diameter is usually accompanied by a good transverse diameter of
the brim (Stewart. Cowan and Philpott, 1979). However, radiographic
measurements of the pelvis are of less value than the clinical judgement of
the relationship of head to pelvis as judged by head level, degree of
moulding and the fit of the head in the pelvis.
The maternal condition must be assessed taking note of her emotional
response to the delayed progress, hydration, pyrexia, tachycardia and
hypertension.
Those primigravidas who do not have evidence of fetal distress or CPD
requiring a caesarean section will now have augmentation of what is assessed
to be inefficient uterine action. For augmentation to be most effective
attention needs to be given to the mother’s emotional response to her labour
and her hydration and analgesia. Assessments and decisions should be
•conveyed to the mother in a way that she can understand and if epidural

THE RECOGNITION OF CEPHAI.OPELVIC DISPROPORTION

621

analgesia is available this should be offered. There is usually some de­
hydration at this stage of the labour and this should be corrected adequately
before giving an epidural block.
The next step is to augment the labour with oxytocin. The aim is to give
the least amount of oxytocin necessary to provide physiological uterine
action. Rupture of the uterus with oxytocin is very rare in the primigravida
and this is not the limiting factor. The real hazards are head compression
from contractions that drive the head into the pelvis when the fit is too tight,
and reduced placental blood flow from overstimulation of uterine action. It
is therefore obvious that oxytocic augmentation should not be embarked
upon unless each patient can depend on the sole attention of one
experienced midwife who can assess fetal condition and labour progress
(including uterine activity) with accuracy and precision.
The amount of oxytocin to be infused is critical. Cowan and Philpott
(1982) have measured uterine activity in normal labour, delayed labour and
augmented labour in the primigravida. Using a fluid-filled intrauterine
catheter, pressure transducer and integrator they found that the mean figure
for uterine activity was 1800 kPa per 15 minutes in primigravidas that
progressed at 1 cm or more per hour in the active phase of labour. The lower
limit of normal progress was 1200 kPa per 15 minutes. When there was delay
in labour progress the levels of uterine activity were considerably less. They
then gave oxytocin to augment delayed labour. Six units of oxytocin were
placed in 11 of intravenous fluid and an Ivac Drop Counter was used to
control the infusion rate. Initially, oxytocin was infused at 2 mu per minute
tor the first 15 minutes. The infusion rate was increased arithmeticallv every
15 minutes until there was either progress in labour or ‘normal’ uterine
activity was achieved. Normal uterine activity was regarded as a contraction
frequency of five contractions per 15 minutes, when the contractions were of
'good' intensity, and when they lasted 40 to 50 seconds. This judgement was
made clinically with a hand on the uterus.
They found that, in those patients who were eventually demonstrated to
have no CPD. their delayed progress due to inefficient uterine action quickly
increased to uterine activity levels of 1800 to 2000kPa per 15 minutes with
oxytocic augmentation, and cervical dilatation then progressed at 1cm or
more per hour to full cervical dilatation. The rest had "increased uterine
activity on oxytocin, but often did not achieve levels of 1800kPa per 15
minutes of uterine activity. They were regarded as having delayed progress
on oxytocic augmentation and proceeded to caesarean section. X-ray pelvi­
metry and measurement of head size showed that these patients had a tighter
fit than those who progressed quickly on oxytocic augmentation and
delivered vaginally.
In a further study they showed that, by using the uterine activity as
measured by the integrator and intrauterine pressure catheter, and using
IS(X)— 2(XX) kPa per 15 minutes as the objective of escalating oxytocic
augmentation, they achieved more successful vaginal deliveries in those who
had previously shown delay in labour. Furthermore, the ‘head fit’ expressed
as a ratio of head size to pelvic size was closer to 1 (X) per cent (a tighter fit)
than in those whose oxytocic augmentation was determined by clinical

I

622

R. H. PHILPOTT

judgement of the strength of the contractions. In all labours fetal condition
was carefully monitored and no labour was allowed to proceed in the
presence of significant fetal distress.
This study shows that, as long as fetal condition is carefully monitored, an
oxytocic augmentation for delayed progress in the primigravid trial of labour
will quickly reveal the patient who has no CPD and who will, within an hour,
change to more efficient uterine action and progress in cervical dilatation of
len/per hour or more. When oxytocic dosage is determined by clinical
palpation of uterine activity some will not progress and among these will be
those with CPD who require caesarean section. A few who do not progress
with clinically judged efficient uterine action (and no fetal distress) would
progress if the oxytocic dosage could be determined more accurately by
intrauterine pressure measurements. However, this technique is not avail­
able to many hospitals in those countries where CPD is common, and they
will need to depend on clinical judgement of uterine activity. At least this
can be safe, if skilfully applied.
Thus, the evidence of a failed trial of labour, and therefore need for
immediate delivery by caesarean section (or sometimes symphysiotomy),
after commencing oxytocic augmentation will be either significant fetal
distress, failure in the progressive dilatation of the cervix or increasing head
moulding with no descent of the head. The change from slow dilatation of
the cervix prior to delay, to normal progress in dilatation after commencing
oxytocin, will commence within two hours of starting the oxytocin in those
who are going to deliver safely vaginally.

The method of final delivery
Those patients who progress normally with or without the need for oxytocic
augmentation will deliver spontaneously in the second stage of labour or,
because of maternal tiredness, may require simple assistance with the
vacuum extractor or forceps. If the trial of labour has to be terminated for
reasons given above it is important that the final delivery method is easy and
non-traumatic for the fetus. If the cervix is not yet fully dilated, this will be
by caesarean section, unless symphysiotomy is feasible and part of the range
of skills of the doctor or midwife in attendance.
Symphysiotomy has a very important place in the delivery of a patient
undergoing a carefully planned and controlled trial of labour. Of critical
importance is the timing of the procedure - not too soon and not too late. If
done too soon, some unnecessary symphysiotomies will be done. If done too ;
late there will be a high incidence of cerebral birth trauma. We do not do
svmphvsiotomies if the cervix has not reached 7cm cervical dilatation, f
delivery is indicated before that time then a caesarean section is indicated. It
there is fetal distress or delayed progress on oxytocin from 7cm dilatation
onwards, svmphysiotomv may be considered. If the head is not more than
two-fifths above with moulding, symphysiotomy will be successful. If the
cervix is 7 cm or more dilated and if the head is three-fifths above but. with a
trial push the head can descend to two-fifths, then symphysiotomy will be
successful. If the symphysiotomy is done prior to full dilatation, let the

I Hl. RhCOCiNI I ION OF’ CHPHAl.OPEI.VIC DISPROPORTION

623

patient lie on her side until full dilatation, when she will usually deliver the
baby without assistance. Any fetal distress that was present prior to the
symphysiotomy will disappear with the relief of head compression, and
delivery need not be hurried.
Vacuum extraction or forceps delivery should only be used once the cervix
is fully dilated and there is fetal distress or delayed progress in the second
stage. Judgement of the presence of CPD is of extreme importance in the
second stage. If. at the time an assisted delivery is indicated, the fetal head is
level with or one-fifth above the brim, then an assisted delivery will be safe
for the fetus. If the head is more than one-fifth above, with moulding, then
delivery must be by either symphysiotomy or caesarean section. Attempts at
vacuum extraction or forceps delivery with the head two-fifths above with
moulding, at the end of a trial of labour, will bring an unforgivable price in
fetal damage.
Recognition of CPD in the multipara

This can be more difficult than in the primigravida. Sometimes there will be
normal progress in cervical dilatation (more than 1cm per hour) to full
dilatation in the multigravid in the presence of unexpected CPD. This
almost never happens in the primigravida. When that happens in the
multigravida the CPD must be diagnosed by the evidence of failure of head
descent with increasing moulding. Usually, however, the multigravida with
CPD also has delayed progress in cervical dilatation in the active phase. We
use the same criteria in diagnosing delay as in the primigravida. The problem
lies in what to do at the time of delay. In the primigravida. oxytocic
augmentation, even in the presence of unrecognized CPD. will virtually
never lead to uterine rupture. This is not necessarily true in the
multigravida.
In studies done by Kambaran and Philpott, not yet reported, the mean
uterine activity levels in multipara who progress normally is 1458 kPa per 15
minutes with a lower limit of normal of 750 kPa per minute. When there has
been delay, and in the absence of fetal distress, oxytocic augmentation to a
level of 1800-2000 kPa per 15 minutes has proven safe in the multigravida, as
•n the primigravida. Translated into clinical terms this means not more than
hve contractions in 15 minutes, lasting not more than 60 seconds, with good
relaxation between contractions. If a safe, easy delivery is to be expected,
there will be a change in the rate of cervical dilatation within an hour,
proceeding to full dilatation at more than 1 cm per hour, and good head
descent. As with the primigravida, it is imperative that there be meticulous
control of fetal condition throughout the management of a patient who has
delay in labour progress, particularly if oxytocin is used.
Recognition of CPD in the second stage

This will be evidenced by either fetal distress, or slow progress or increasing
moulding with failure of head descent. The first and the last have already
been defined. Progress in the second stage is determined by time and
progressive descent of the head. Among Zulu primigravidas. if there is no

624

R. H. PHILPOTT

CPD there will be rapid head descent in the first ten minutes of bearing down
and delivery within 30 minutes. If there is no head descent in the first ten
minutes then it can be expected that the second stage will take longer than 30
minutes. This may or may not be a problem. If the fetal heart rate picks up
well between contractions and there is progressive descent of the head, it is
all right to wait up to 45 to 60 minutes. If delivery is not achieved in that time
the mother will become distressed even if the fetus does not. and inter­
vention is indicated. When reviewing the literature it is noted that many
authors are prepared to let the second stage last much longer than an hour,
as long as there is progress and no fetal distress. Possibly the Zulu mother
knows more about bearing down for. in spite of our policy, our instrumental
delivery rate in the second stage has always been well below 10 per cent.
REFERENCES
Bernard. R. M. (1952) The shape and size of the female pelvis. Edinburgh Medical Journal. 52,

1-16.
Briggs. N. D. (1981) Engagement of the fetal head in the Negro primigravida. British Journal of
Obstetrics and Gynaecology, 82. 1086-1089.
Cardozo. L. D.. Gibb. D. M. F.. Studd, J. W. W. et al (1982) Predictive value of cervimetric
labour patterns in primigravidae. British Journal of Obstetrics and Gynaecology, 89,
33-38.
Cowan. D. B. & Philpott, R. H. (1982) Intrauterine-pressure studies in African nulliparae:
delay, deliverv and disproportion. British Journal of Obstetrics and Gynaecology. 89,
370-380.
Crichton. D. (1974) Measurement of the fetal head above the pelvic brim in fifths. South
African Medical Journal. 48. 784-785.
Friedman. E. A. (1978) Labour. Clinical Evaluation and Management. Second Edition. New
York. Appleton-Century-Crofts.
Ince. J. G. H. A Young. M. (1940) The bony pelvis and its influence on labour. Journal of
Obstetrics mid Gynaecology nf the British Empire. 47, 130-196.
Stewart. K. S.. Cowan. D. B. & Philpott. R. H. (1979) Pelvic dimensions and the outcome of
trial labour in Shona and Zulu primigravidas. South African Medical Journal. 55.847-851.
Stewart. K. S. & Philpott. R. H. (1980) Fetal response to cephalopelvic disproportion. South
African Medical Journal. 87. 641-649.
Wilson. P. C.. Philpott. R. H.. Spies. S. et al (1979) The effect of fetal head compression and i
fetal acidaemia during labour on human fetal cerebral function as measured by the fetal 7
electroencephalogram. British Journal of Obstetrics and Gynaecology. 86. 269-277.
-

■J

PAPUA NEW GUINEA MEDICAL JOURNAL

THE CERVICOGRAPH IN LABOUR MANAGEMENT
IN THE HIGHLAND OF PAPUA NEW GUINEA

Christopher E. Lennox*

From the Immanuel Lutheran Church, Wapenamanda, Enga Province.
• Present address:
Dr. C. E. Lennox,
19 North Crescent,
Durham, DH1 4NE
United Kingdom.

SUMMARY

The use of the cervicograph with the incor­
porated action and alert lines in labour manage­
ment in developing country obstetrics is now
well established and patients in labour in Port
Moresby have been thus managed since 1973.
In this study the cervicograph was introduced
in hospitals and health centres in Enga Province
and the results .analysed.
Despite intensive tuition in its use, a large
number of cervicographs were incorrectly con­
structed, especially in health centres, and its
introduction made little difference to opera­
tive delivery or perinatal mortality rates.
The proportion of patients crossing the action
line was much higher than in Post Moresby
but this difference was probably also largely
due to the inaccuracy of many cervicographs.

labour and/or operative delivery. Cervicographs have been in use in Port Moresby Gen­
eral Hospital since 1973 and their use is en­
couraged throughout Papua New Guinea3.
However, no assessment of their use and value
has been carried out outside of Port Moresby.
This paper records the results of an attempt
of establish the efficient use of cervicographs
(including alert and action lines) in labour pat­
ients in health centres and hospitals in a rural
Highland Province.

A detailed description of the principles and
practice of the use of the cervicograph and the
incorporated alert and action lines is not
included; the reader is referred to the reports
by Philpott4, Philpott and Castle2, Studd5,
Studdfl/aZ6 and Bird7.
METHODS

It is likely that these difficulties will not
be overcome until many more nurses with
intensive training in the use of the cervicograph are available and until more deliveries
take place in health institutions to maintain
practice. In the interim a simpler rule of thumb
for referral of labour patients may be helpful.

A simple cervicograph based on that used
in Port Moresby3 in which the most important
feature was the dilatation of the cervix was
designed for use in health centres and hos­
pitals in Enga Province. It was found that only
those very few nurses in the Province
who had undertaken the Midwifery course
at Port Moresby General Hospital were at
all familiar with the cervicograph, although
its use should be taught in all nursing schools.

INTRODUCTION
The use of the cervicograph as a visual
aid to the progress of labour has become
well established since the work of Friedman
in 1967’ and it has become of even more
value to developing country obstetrics since
the introduction of the alert and action lines
by Philpott and Castle2. These lines are used
to indicate those patients whose labour is not
progressing satisfactorily and who require
referral to a central unit for augmentation of

Consequently intensive in-service courses for
nurses and health extension officers were held
at each centre at which the cervicograph was
to be used in addition to two in-service courses
at the provincial hospital, and frequent visits
to each centre throughout the period of the
study. Each centre with a doctor was issued
with a management flow chart for use with

286

I

VOLUME 24, No. 4, DECEMBER 1981

the cervicograph, based on the notes produced
by Bird8.
Included in the study were two hospitals,
Wapenamanda and Sopas and four rural
health centres, Laiagam, Pumakos, Kompiam
and Kandep staffed by nurses and health
extension officers, although Laiagam was inter­
mittently staffed by a doctor, (see map, last
page). Labour, delivery and cervicograph re­
cords of all patients delivered in these centres
during the study period were sent to the author
for analysis. Patients included in the study
were those meeting the criteria used by Bird in
his 1974 Port Moresby study3, viz; labour
not induced, single pregnancy, vertex presenta­
tion, no previous Caesarean section, no ante­
partum haemorrhage, no eclampsia, and birth
weight of infant at least 1000g. The period of
the study varied from centre to centre because
of local difficulties, from a minimum period of
12 months to a maximum of 26 months, all be­
tween August 1977 and October 1979.

Analysis was made of the use of the cervico­
graph and delivery outcomes and the results
were compared with the period before the
introduction of the cervicograph and with the
1974 Port Moresby study,3.

had one or more major faults. A number of
the faulty cervicographs were permitted to
enter the study as long as the action line
appeared reasonably accurately placed. The
commonest faults were wrong placement of
the alert and action lines and an inadequate
number of vaginal examinations. The incidence
of faults was much higher in health centres
(56%) than in hospitals with medical officer
supervision (26%). No assessment could be
made of the accuracy of cervical dilatation
assessed by vaginal examination but there were
occasions where it was obviously inaccurate,
usually over-rather than under-estimated, as
was Birds'experience3.

Patient Characteristics

Of the 959 patients studied, (here were 390
primigravidae (41%) and 569 multigravidae.
The average interval from admission in labour
to delivery was 5.3 hours and the average cer­
vical dilatation on admission was 5.6 centime­
tres. Each patient received an average of 2
vaginal examinations.

Patients were divided into groups identical
to those in the Port Moresby study ie.

RESULTS
No important differences were noted be­
tween different centres (although the incidence
of operative deliveries was naturally higher
at centres with a doctor) and the results from
all centres are combined. Throughout, some
comparisons are made with the results obtained
in Port Moresby by Bird3. References to Port
Moresby figures all refer to this study.

Group 1: Cervix less than 4cm dilated on ad­
mission and at 8 hours, but in labour
(33 patients)

Group 2: Cervix less than 4cm dilated on ad­
mission but 4cm or more dilated
at 8 hours (221 patients)

Cervicograph Reliability

Details of 1154 deliveries were returned,
of which 959 (83%) filled the criteria for entry
into the study and had acceptable cervicographs. Excluding those patients fully dilated
on admission (though these were included in
the study) 1004 patients should have had cervicographs which would have entered the study
but of these, only 659 (66%) were correctly
completed. In 345 cases (34%) there was either
no cervicograph at all, or the cervicograph

287

Group 3: Cervix 4cm or more dilated on ad­
mission (705 patients)

Some

interesting

differences

in

propor­

tions of patients in the different groups be­
tween this study and that in Port Moresby
were noted and are shown in table 1. These
differences are probably mainly a reflection
on the later presentation to hospital in labour
of Enga women.

r

PAPUA NEW GUINEA MEDICAL JOURNAL

TABLE 1:

GROUPS BY PARITY: ENGA AND PORT MORESBY (POM) STUDIES

PATIENTS

Primigravidae

Multigravidae

Total

GROUP 1

Enga

POM

Enga

POM

Enga

POM

390
(41%)

1053
(35%)

22
(6%)

93
(9%)

104
(27%)

512
(49%)

264
(68%)

448
(42%)

569
(59%)

1959
(65%)

11
(2%)

81
(4%)

117
(21%)

732
(37%)

441
(78%)

1146
(59%)

959
(100%)

3012
(100%)

33
(3%)

174
221
(6%) I (23%)

1244
(41%)

705
(74%)

1594
(53%)

No patients failed to dilate to 4cm or
more and they could thus all be divided into

those who subsequently did or did not cross
the action line. Ninety eight patients (10%)
crossed the action line and their method of
delivery is shown in table 2.

METHODS OF DELIVERY, OXYTOCIN USAGE
______ AND PERINATAL MORTALITY

ACTION LINE CROSSED
NO.
(%)

Caesarean section

16

(16)

Symphysiotomy
Vacuum extraction

14
23
45

(14)
(23)
(46)

13

(13)

1

98

Spontaneous
Intravenous oxytocin
Perinatal mortality

Total deliveries

n

GROUPS

POM

Action Line and Method of Delivery

TABLE 2:

GROUP 2

Enga

ACTION LINE NOT CROSSED
NO.
(%)

(1)

5
13
62
781
10
16

(7.0)
(91)
(1-2)
(2.0)

(100)

861

(100)

These results are compared with the Port
Moresby figures in table 3. The differences
are discussed below.

(0.6)
d-5)

Proportions of primigravidae and multi­
gravidae and of patients in the three groups
crossing the action line were studied. This is
shown in table 4 where comparison is again
made with the Port Moresby figures.

288

i

I

VOLUME 24, No. 4, DECEMBER 1981

METHOD OF DELIVERY IN ENGA AND PORT MORESBY (POM)

TABLE 3:

ACTION LINE NOT CROSSED
POM
Enga

ACTION LINE CROSSED
Enga
POM

Caesarean section
Symphysiotomy
Vacuum extraction/
forceps
Spontaneous
Total

23 (23%)

7 (18%)
5 (13%)
19(49%)

5 (0.6%)
13(1.5%)
62 (7.0%)

3(0.1%)
1
259 (9%)

45 (46%)

8 (20%)

781 (91%)

2710(91%)

98(10%)*

39(1%)

861 (90%)

2973 (00%)

16(16%)

14 (14%)

Percentage crossing or not crossing action line.

TABLE 4:

ACTION LINE CROSSED BY PARITY AND GROUP:
ENGA AND PORT MORESBY STUDIES
Primigravidae
No. (%)♦

Multigravidae
No. (%)*

Enga

68 (17)

Port Moresby

20

(2)

No. (%)♦

Group 2
No. (%)*

30, (5)

5 (15)

28 (13)

65 (9)

19 (I)

11

(1)

i 17 (1)

Group 1

(16)

11

Group 3
No. (%)*

* Percentages are of total patients of each parity or group,

Referred Cases

referred from a health centre in the study
(referral rate of 7% for those centres), and of
those 15, only 7 had crossed the action line
at the time of referral; 13 were primigravidae.
The method of delivery of the 15 is shown
in table 5.

Thirty one (3.2%) of the 959 study cases
were referred from health centre to hospital
in labour (complications such as antepartum
haemorrhage or malpresentation were excluded
from the study). Only 15 of the 31 were
TABLE 5:

METHOD OF DELIVERY IN ENGA AND PORT MORESBY (POM)

Caesarean section
Symphysiotomy

ACTION LINE CROSSED
(IN HEALTH CENTRE)

ACTION LINE NOT CROSSED

1

1

2

2

Vacuum extraction
Spontaneous

2

2

2

3

Total

7

8

289

PAPUA NEW GUINEA MEDICAL JOURNAL

livery is shown in table 6 (which includes those
cases in table 5); 22 of these were primigravidae.

The 31 referred cases delivering in hospital
were studied together and their method of de-

TABLE 6:

METHOD OF DELIVERY OF PATIENTS REFERRED IN LABOUR
ACTION LINE CROSSED
(IN HOSPITAL)

ACTION LINE NOT CROSSED

Caesarean section
Symphysiotomy
Vacuum extraction
Spontaneous

2
2
4

2

1
5
4
11

Total

10

21

not used at any of the other centres until the
present study was commenced. Delivery records
varied in availability and accuracy, so that the
period studied differed in different centres,
but all the deliveries analysed took place be­
tween 1974 and 1977. Comparisons of methods
of delivery and perinatal mortality are shown
in table 7.

Deliveries before the Use of the Cervicograph
The delivery records of all the centres
included in the study were analysed for the
years immediately preceding the introduction
of the cervicograph. The cervicograph was used
intermittently but not with any enthusiasm
at the Provincial Hospital in 1976, but was

TABLE 7:

DELIVERIES BEFORE AND AFTER INTRODUCTION OF THE CERVICOGRAPH

Pre-Cervicograph
(1974-77)

net-

Post-Cervicograph
(1977-79)

w

(^y

(2-2)

(1.8)
(3.2)

(100)

(3.1)
(2.0)
(56)

Referrals in labour

60
39
107
1719
30
22

(11)

21
27
85
826
17
31

Total deliveries

1925

(100)

959

Caesarean section
Symphysiotomy
Vacuum extraction/forceps
Spontaneous
Perinatal motality

(89)

(1.6)

(2.8)
(8.9)

(86)

I
i

290

i

VOLUME 24, No. 4, DECEMBER 1981
DISCUSSION

i

j

The value of the cervicograph in the manage­
ment of labour and the significance of crossj1! 7thc action line has been well esublished
• , , and it is not the purpose of this paper to
confirm or refute this, but merely to review its
use in a rural area where it should be most
valuable and to compare the findings in this
population with the 1974 Port Moresby study3.
Crossing the action line was confirmed as
highly significant in terms of complicated
delivery. Caesarean section and symphysio­
tomy rates in groups crossing and not crossing
the action lines were similar in both Enga
and Port Moresby although there were more
spontaneous deliveries in the Enga group
crossing the action line than in Port Moresby
(Table 3). There were however, striking differ­
ence between the studies in the proportion of
patients crossing the action line; 10% in Enea
and only 1% in Port Moresby. There are three
possible explanations for this-

Table 8:

(1) Different characteristics of Highland and
coastal population.
(2) Higher proportion of at risk women in
Enga study.
0) Cervicographs in Enga were not accurate.

It is recognized that there is a higher incid*
ence of disproportion in Highlanders compared
to coastal Papua New Guineans and this was
confirmed in the Port Moresby study, where
the operative delivery rates for Highlanders
alone were closer to the rates in Enga, especially for Caesarean section (Table 8). Bird does
not indicate whether the proportion of High­
landers crossing the action line was also higher,
but Duignan et al9 found no difference in the
rate of cervical dilaution among different
racial groups. It is of note, however, that
the proportion of Rhodesian women crossing
the action line in Philpott and Castle’s original
work2 was similar (11%) to this study.

major operatbt delivery rates-port Moresby and enga
Total
Study
Group

PORT MORESBY
Highlanders
Only

enga

Caesarean section (rate)

03%)

1.8%

Symphysiotomy (rate)

2.2%

0.2%

0.5%

2.8%

There is also no doubt that the proportion
of at risk to “normal” women delivering in
a health institution in Enga is greater than
in Port Moresby where a very large number
of “normal” women now deliver in hospital.
However, because of the selection of cases for
this study, many of the ‘at risk’ women were
excluded (antepartum haemorrhage, malpresentation, previous Caesarean section, etc),
and the incidence of spontaneous delivery in
the Enga and Port Moresby groups was very
similar (86%) and 90% respectively).
The main reason for the different propor­
tion of patients crossing the action line lies

291

probably, then, in the accuracy of the ceryicograph. Over one third of the cases studied
in Enga had either no cervicograph (these
were excluded from the study) or had one or
more obvious faults in the cervicograph. This
was despite the intensive teaching and supervi­
sion in the technique throughout the study
period. The major faults were the inaccurate
placing of the alert and action lines, failure to
perform vaginal examinations at the conect
times ( particularly failure to perform 2 hourly
examinations if the alert line was crossed) and,
undoubtedly (although this was not necessarily
obvious from studying the cervicographs),
inaccurate estimation of the cervical dilaution

f

PAPUA NEW GUINEA MEDICAL JOURNAL

on vaginal examination. The early dilatation
was obviously overestimated in a number of
cases and this led to “false positive'’ crossing
of the action line. Bird also found this a pro­
blem and subsequently altered the placement
of the alert line for women 4cm dilated to 4
hours later rather than 2 hours, keeping the
2 hour interval for those 5cm or more dilated7.

Although this would probably have reduced
the number of patients in this study crossing
the action line, it would certainly also have,
increased the inaccuracy and unreliability of
the cervicographs by adding a further com­
plicating factor for nurses who were already,
having difficulty fully understanding the
method.

Crossing the action line did not seem a major
factor in the referral of patients in labour
(Table 5) although patients refened from
health centres to hospiuls did have a higher
incidence of operative delivery* than average
(58%).

Disappointingly, the introduction of the
cervicograph has made little difference to the
operative dehvery or perinatal mortality rates
in Enga Province, in contrast to the findings
in Pon Moresby, where both declined markedly.
Indeed the vacuum extraction rate in Enga
has increased although this may well be partly
a result of changing medical personnel. The
threefold increase, however, in the reterrai rate
of patients in labour may be a result of the
cervicograph^ if only because of an increased
TABLE 9:

awareness among health centre staff of the
necessity to refer patients as a result of the
attention paid to the matter at in-service
teaching. On the other hand, better transport
over the years may have been just asjmportant!

As Bird emphasises, it is only possible to use
the cervicograph properly if it is thoroughly
understood and this only comes with intensive
training in a centre with large numbers of de­
liveries where it is used frequently, and where
there are sufficient experienced staff to super­
vise the vaginal examinations of those in train­
ing7. This is not true of any centre in Enga
and during the period of this study there was
rarely more than one midwife in the entire
province who had undergone intensive mid­
wifery training at Port Moresby Hospital.
The Enga Provincial Hospital averages only
some 250 deliveries per year and at one health
centre in the study only 30 deliveries took
place in 26 months. Even those well trained
in the use of the cervicograph and in the ass­
essment of cervical dilatation can become
very “rusty ’’ with such infrequent practice.
Although it must be hoped that in time,
many of the difficulties will be overcome and
that the cervicograph can become a truly
useful and effective tool in labour management
through out the country, perhaps in the mean­
while some simpler “rule of thumb” may help
health cenues in their referral of patients.

The outcome of patients who were still
undelivered 8 hours after admission in labour

OUTCOME OF PATIENTS UNDELIVERED AT 8 HOURS
AND OF THOSE CROSSLNG ACTION LINE
Entire Study
Group
No.
(%)

Symphysiotomy

21
27

Vacuum extraction
Spontaneous
•Perinatal mortality

85
826
17

Total deliveries

959

Caesarean section

(2.2)

Action Line
Crossed
No.
(%)

(16)

16
14

(2.8)
(8.9)
(86)

23
45

(1.8)

(100)

(14)
(23)

I

Undelivered
8 hours after
admission
No.
(%)
18
15

<9.6)

36

(8-0)

118

1

(46)
(1.0)

(19)
(63)

3

(1.6)

98

(100)

187

(100)

292

<

■ t

»

VOLUME 24, No. 4. DECEMBER 1981

1
?

i

2. Philpott RH. Castle WM. Cervicographs in the man­
agement of labour in primigravidae. J Obs Gyn
Bm Cwlth 1972; 79:592, 599.

was studied. This comprised 187 patients (19%
of total study group), 125 of whom were primigravidae and 62 multigravidae. This group
was compared with those patients who crossed
the action line and with the entire study group
(Table 9).

3. Bird GC. Cervicographs in the management of
labour in primigravidae and multigravidae. Papua
New Guinea Med J 1974; 17:324.

4. Philpott RH. Graphic records in labour. Br Med J
1972;4:163.

Failure to deliver within 8 hours of admis­
sion meant a 37% chance of a complicated de­
livery compared to a 14% dunce in the total
group and a 54% chance if the action line
was crossed. These differences are highly
significant. Although this is much less specific
than crossing the action line, perhaps, in
the absence of a member of staff well-versed
in the use of the cervicograph, serious con­
sideration should be given in health centres
to the transfer of a patient who is not delivered
or nearly delivered 8 hours after admission in
labour.

5. Smdd J. Partograms «nd nomograms of cervical
dilatation in management of primigravid labour.
Br;Ved/ 1973;4:451.

6. Studd JWW, Clegg DR, Sanders RR, Hughes AO.
Identification of high risk labours by labour
nomogram. Br Med J 1975; 2:545/
7. Bird GC. Cervicograph management of labour.
Prop Doctor 1978; 8:78.
8. Bird GC. The cervicograph; mimeographed notes
from Port Moresby General Hospital, 1975.

RECERENCES
9. Duiznan NM. Studd JWW, Hughes AO. Character­
istics of normal labour in different racial groups.
BrJobsGynae 1975; 82:593.

1. Friedman EA. Labour, clinical evaluation and
mananement. Neu York: Apple-Century-Crofts:
1967.

r

293

s
British Journal of Obstetrics and Gynaecology
October 1977. Vol 84. pp 747-750

EPIDEMIOLOGICAL FEATURES OF PERINATAL DEATH
DUE TO OBSTRUCTED LABOUR IN ADDIS ABABA
BY

R. L. Naeye

A. Dozor
Department of Pathology, Pennsylvania State University College of Medicine
Hershey, Pennsylvania, USA 17033
N. Tafari
AND

S. M. Ross
Departments of Paediatrics, Obstetrics and Gynaecology, Addis Ababa University
Faculty of Medicine, Addis Ababa, Ethiopia
Summary
Obstructed labour was the second most common cause of perinatal death in
Addis Ababa, Ethiopia, being responsible for 9-1 perinatal deaths/1000 births.
Most obstructed labours were due to cephalopelvic disproportion. There was
a ninefold increase in the perinatal death rate when the patients were anaemic
but most perinatal deaths were due to delays in seeking available obstetrical
care. Formal education of the patients had little influence on the death rates
but the informal education that comes with prolonged urban residence had a
markedly favourable effect. Use of prenatal medical services and adequate income
also had a favourable influence.

Perinatal deaths due to obstructed labour are
most common in Third World nations because
maternal pelves are often undersized and
obstetric intervention is frequently delayed or
unavailable (Sambhi, 1973; Mtimavalye and
Nfaathuis, 1974; Mphahlele and van der
Meulen, 1975). Our study was designed to
search for reasons why women with obstructed
labours in Addis Ababa, Ethiopia, failed to
obtain obstetric care in time to save their infants.

take place. There were 148 perinatal deaths
due to obstructed labour, and all the infants
had postmortem examinations. The fetal presen­
tations associated with the obstructions are
found in Table I. The pregnancies that ended
with fatal obstructions were compared with
568 successful pregnancies, selected to be
representative of the delivery population. These
controls were comprised of the first ten deliveries
after 0800 each day that resulted in the birth of
infants who survived the neonatal period.
Hospital and clinic records, an interview with
the mother and her physical examination on
the day after delivery provided 124 medical,
demographic and other items of information
for analysis in most patients. The information
collected included duration of pregnancy based

Patients
A study of perinatal mortality was undertaken
in Addis Ababa, Ethiopia, in 1974 to 1975
in the hospitals and clinics affiliated to the
Addis Ababa University Faculty of Medicine
where about 40 per cent of births in the city
747

748

NAEYE, DOZOR, TAFARI AND ROSS

Table I
Perinatal mortality rate due to obstructed labour by type
of presentation or delivery. Cephalopelvic disproportion
was the cause of the obstructions in those with vertex
presentations

Type of
presentation
or delivery

Frequency of
perinatal
deaths per
1000 births
of type stated

Normal vertex
Breech
Face
Brow
Forceps

21
27-7
26-1
126-3

Vacuum extractor
Caesarean section
Ruptured uterus

23-2

401

11-2

1000 0

P value
compared to
normal vertex
delivery

<0 001
<0 001
<0 001
<0 001
<0 001
<0 001
<0 001

on a church calendar well known to most
women, mother’s education, size and sources
of income, specific expenditures for food, water
and other items, place and duration of residence
housing density, tribe, religion, details of
employment, facilities for excreta disposal,
sources of water,' age of first marriage and
duration and status of marriages.
Data were also collected on maternal age,
previous obstetric history, wanted/unwanted
status of the pregnancy, efforts at contracep­
tion, abortifacients, gestational hypertention
peripheral oedema, vaginal bleeding, poly­
hydramnios, jaundice, parasites, venereal
diseases including serologic tests for syphilis,
other specific disorders during gestation, number
of visits for prenatal care, blood haemoglobin
values, blood groups, leucocyte counts and
medications taken during pregnancy. In­
formation was also collected on details of
labour and delivery, fetal distress, Apgar scores,
infant blood groups and the clinical course of
the newborn infant and mother after delivery.

Statistical Methods
Details have recently been published on the
analyses used to determine which of the many
variables were primarily related to the perinatal
deaths (Naeye et alf 1977). Primary variables
were judged to be those that had a significant
influence on mortality rates withnut
without being

dependent on other variables. First, the
frequency of each variable in the fatal cases was
compared with its frequency in the controls
using the chi-square test in two-way contingency
tables. This resulted in a short list that included
both primary and secondary variables. Then a
log-linear model analysis of contingency tables
was carried out on the short list to distinguish
the primary from the secondary variables.

Results
The frequency of perinatal deaths due to
obstructed labours was 9-1 per 1000 births.
Ninety per cent of mothers were at term.
Eighty-two per cent of the infants were stillborn,
54 per cent were male and 6 per cent were*
twins. Eighty-eight per cent of the labours
lasted over 20 hours, many for several days.
Thirty-eight per cent of the infants had resultant
congenital pneumonia due to bacteria common
in the vagina. Twenty-four of the patients had
a ruptured uterus, most on arrival at the delivery
centres and one mother died.
Duration of residence in the city, use of
prenatal medical services, cash income, maternal
blood haemoglobin levels and marital status
proved the primary factors in the perinatal
deaths (Table II). None had significant inter­
relationships with each other or with other
variables in terms of influencing perinatal death
rates due to obstruction. Long residence in the
city markedly reduced the frequency of the
deaths (Table II). A surprising number of
rural mothers obuined prenatal medical care
and those who obtained such care, whether
rural or urban residents, had only one-fifth
the perinatal death rate of those who had no
prenatal care. Surprisingly, formal education
did not significantly reduce the mortality rates.
The death rates were high when patients were
anaemic. Women who had been divorced and
remarried had a lower frequency of perinatal
deaths and a higher ratio of prior successful
to unsuccessful pregnancies than did women in
other marital categories (Table II). This ratio
of successful to unsuccessful pregnancies was
7 : 1 in women divorced who remarried and
3 : 1 in both those divorced and not remarried or
those married and never divorced (P<0-05).

OBSTRUCTED LABOUR AND PERINATAL DEATH IN ADDIS ABABA

749

Table II
The relationship ofprimary factors to the frequency ofperinatal death due to obstructed labour

Frequency of
perinatal deaths/
1000 births

P value and comparison

Mother's residence
Rural, outside Addis Ababa
Addis Ababa, 1-12 months
1-10 years
Over 10 years
All life in Addis Ababa

51 6
7-7
6 1
52
1-8

<0 001 compared with whole life in Addis
<0 001 compared with whole life in Addis
<0 001 compared with whole life in Addis
<0 001 compared with whole life in Addis

Prenatal medical care
Mother had care
No care

4-4
21-2

<0 001 compared with no care

831
150
3-7
5-8

<0 001 compared with > $40/month
<0 001 compared with > $40/month

Maternal haemoglobin level
Less than 10-1 g/dl
101-11-7
Over 11-7

63-3
13-3
6-9

<0 001 compared with >11-7 g/dl
<0 001 compared with >11-7 g/dl

Marital status
Never married
Married
Divorced
Divorced and remarried

11-2
8-7
180
3-7

<0 001 compared with divorced and remarried
<0 05 compared with divorced and remarried
<0 001 compared with divorced and remarried

Family income, Ethiopian §/month
None
1-40
41-200

Over 200

None of the following factors had any
significant influence on perinatal mortality rates
due to obstructed labours: maternal height,
maternal weight, specific illnesses during
pregnancy, use of oxytocic or anaesthetic agents,
vaginal bleeding, tribe, religion or type of work
during pregnancy.
Discussion
Obstructed labour due to cephalopelvic
disproportion or abnormal fetal presentation
w-as the second most frequent cause of perinatal
death in Addis Ababa,.Ethiopia, and the number
of such perinatal deaths was equivalent to
half to two-thirds of all perinatal deaths in
most European and
North American
communities (Pharoah, 1976). The pelvic
contractions responsible for most of the
obstructions were probably due to childhood
and adolescent undernutrition which is prevalent

in Ethiopia (Clegg et al, 1972); rickets may also
have played a role. Fetal head sizes in Addis
Ababa are not unusually large (Ross, 1975).
Most of the fetuses were dead when the
mothers arrived for delivery and were then
usually delivered by embryotomy. When the
fetus was alive there were often signs of intra­
uterine infection and severe fetal distress.
Caesarean section was seldom undertaken in
such cases because it posed a serious threat of
generalized peritonitis and septicaemia in an
undernourished patient.
A high proportion of the mothers who
arrived too late at delivery centres had received
no prenatal care. Surprisingly, the level of
formal education did not have much influence
on the death rates but the informal education
that comes with prolonged urban residence did
have a markedly favourable effect. Divorced
women who had remarried had both lower

f

750

NAEYE, DOZOR, TAFARI AND ROSS

perinatal death rates from obstruction and a
higher ratio of prior successful to unsuccessful
pregnancies than did women in other marital
categories. In Addis Ababa, as in many other
third world cities, successful pregnancies make
a woman attractive for remarriage.
One surprising finding was that the mother’s
height had no significant influence on perinatal
mortality rates due to obstructed labour. Baird
(1949) stated that reproductive performance
is best when the patient's height is more than
160 cm and worst when it is less than 152 cm,
partly because pelvic shape and size are related
much more to height than to ethnic origin.
In Addis Ababa, most women have^ adequate
sized and normal shaped pelves, even when
they are short in suture (Ross, 1975).
Our study suggests that a greater under­
standing of the advantages of prenatal medical
care and obstetrical services by mothers would
reduce the perinatal mortality rate from
obstructed labour in Addis Ababa. Some women
also need money so that they can obtain prompt
transportation to free or low cost delivery
services in the city. Maternal anaemia is another

factor that increases the risk of perinatal death
in obstructed labours. Many of the anaemias
would undoubtedly have been corrected if the
mothers had attended clinics earlier in pregnancy.
A CKNOWLEDGEMENT

The project was supported by US Public
Health Service Grant HD 08130-02.
References
Baird, D. (1949): Lancet, 1, 256.
Clegg, E. J., Pawson, E. G., Ashton, E. H.» and Flinn,
R. M. (1972): Philosophical Transactions of the
Royal Society of London, B 264, 403.
Mphahlele, M., and Van Der Meulen, A. J. (1975):
African Medical Journal, 49, 1204.
Mtimavalye, L. A. R., and Maathuis, J. B. (1974):
Journal of Obstetrics and Gynaecology of the British
Commonwealth, 81, 380.
Naeye, R. L., Tafari, N., Judge, D., Gilmour, D., and
Marboe, C. (1977): Journal of Pediatrics, 90, 965.
Pharoah, P. O. (1976): Proceedings of the Royal Society
of Medicine, 69, 335.
Ross, S. M. (1975): Ethiopian Medical Journal, 13, 153.
Sambhi, J. S. (1973): International Journal of Gynecology
and Obstetrics, 11, 51.

]n\. J. Gynaecol Obsiet.. 1987. 25: 93-98
International Federation of Gynaecology & Obstetrics

93

Clinical and Research Articles
MATERNAL MORTALITY - A TWELVE-YEAR SURVEY AT THE UNIVERSITY
OF ILORIN TEACHING HOSPITAL (U.I.T.H.) ILORIN, NIGERIA
O.O. ADETORO

J

Department of Obstetrics and Gynaecology. University of Borin. P.M.B. 1515. Borin (Nigeria)

(Received June 27th;*1986)
(Accepted September 10th. 1986)

Abstract

Adetoro OO (Department of Obstetrics and
Gynaecology, University of florin, P.M.B.
1515, florin, Nigeria). Maternal mortality —
a twelve-year survey at the University of
florin Teaching Hospital (U.f.T.H.) florin,
Nigeria.
Int J Gynaecol Obstet 25: 93-98. 1987
This paper concerns an analysis of mater­
nal death at the University of florin Teaching
Hospital (U.f.T.H.) florin over a 12-year
period (1972-1983). There were 138,577
births and 624 deaths making a maternal
mortality rate of 4.50 per 1000 births.
Hemorrhage, ruptured uterus and obstructed
labor were the major direct obstetric causes of
death. The most important indirect causes
were cerebrospinal meningitis, pulmonary in­
fections and fulminating hepatitis. The main
avoidable factors were ineffective and cum­
bersome blood transfusion services; poor
management of the third stage of labor; large
number of unbooked patients and poor del­
ivery room structure encouraging sepsis.
Suggestions are made for a .more integrated
type of maternity' services in our hospital,
health education programs for the public and
particularly the expectant women and
availability of an effective blood bank service
within the maternity hospital premises for
prompt treatment of patients requiring emer­
gency blood transfusion. The analysis under(MI2< 1-7292/87,51 >
£ 1987 International Federation of Gynaecology & Obstetrics
Published and Printed in Ireland

lines the great problem of maternal mortality
in the developing world.
Keywords: Maternal deaths; Hemorrhage;
Ruptured uterus; Obstructed labor; Mater­
nal infections.
Introduction

For most developing countries of the
world like Nigeria, the desire for marry
children amongst women is great. Con­
sequently, the birth rate is high. Per capita
income is alarmingly low and family plan­
ning is often reluctantly accepted because of
religious beliefs. Also, in a significant pro­
portion of families in Nigeria women are
working and are therefore expected by the
polygamous husbands to provide for the
children. In view of the multiple role of our
women, it is extremely important that we
look into the factors responsible for mater­
nal mortality.
Most developing countries do not have
accurate statistics on births and deaths. At
the present time, Nigeria does not produce
periodic statistics on the maternal mortality
rate. It also appears that for a long time to
come, the only figures or statistics on
maternal mortality will come from the few
Teaching

and

Specialist

Hospitals

2

which

keep data These figures can not truly
reflect the National mortality rate since
Int J Gynaecol Obsiet 25

k

I

^4

Adeioro

about 80% of deliveries still occur outside
the hospital under the supervision of the
traditional birth attendants.
Furthermore, a periodic review of the
causes of maternal deaths in an obstetric
unit is desirable in order to continuously
evaluate the effects of the maternal services
on the obstetric performances. This was the
over-riding reason for the review of the
maternal deaths in our hospital over a 12year period. The objectives of the study
were: (1) to determine the trend of maternal
mortality at the University of Ilorin Teach­
ing Hospital; (2) to identify the causes of
death; (3) to identify ways of minimising the
frequency of preventable maternal deaths.
In order to appreciate fully the factors
which may affect the results of this study,
the following peculiar circumstances of our
hospital are noteworthy. (1) The major
blood bank is located at the General Hospi­
tal Wing of this Teaching Hospital which is
4 km away. This General Hospital Wing
accommodates all the other University^
departments except the Obstetrics and
Gynaecologic Department. The Obstetrics
and Gynaecologic Department is served by
a small blood bank unit which operates from
0800 h to 1600 h daily and thereafter all the
emergency blood grouping and crossmatch­
ing are carried out at the General Hospital
Wing. (2) The delivery suite has a toilet
opening directly to the passage between
admission room and first stage room. Also
the windows of this delivery suite open
directly to the outside space. (3) The only
Gynaecologic and Obstetric operating
theatre leads directly to the outside through
a small corridor, and this theatre is served
bv an inappropriate single unit air con­
ditioning system, which rarely worked well
continuousiy for 24 h. (4) Nurses anes­
thetists provide most of the anesthetic ser­
vices.

Materials and methods
The University of Ilorin Teaching Hospi­
tal (U.I.T.H.) Maternity Wing is a referral
Int J Gvnaecol Obsiet 2.'

centre in Kwara State of Nigeria. This
U.I.T.H. was until 1981, the Kwara State
Specialist Hospital for Obstetrical and
Gynaecological cases located at the State
Capital. Referral to this hospital are from
State’s district hospitals, turai maternity
centers and private maternity homes.
The information on maternal deaths in
pregnancy, labor and puerperium was
obtained from the case notes. Relevant facts
collected from the record include the num­
ber of maternal deaths, maternal age, parity
and causes of death. The causes of death
were examined under two headings: deaths
directly related to obstetric complications
such as antepartum hemorrhage, eclampsia
and ruptured uterus and secondly deaths
indirectly related to the obstetric state of the
women, for example pulmonary diseases,
and diabetes mellitus in pregnancy.
}

Results

Six hundred twenty four maternal deaths
occurred at U.I.T.H. Ilorin, during the
period under review. In the same period
there were 138,577 births, giving a maternal
mortality rate of 4.5 per thousand (Table I).
The overall trend in the maternal mortality
rate per year suggests an improvement
(Table II and Fig. 1). Five hundred one
(80.3%) of the maternal deaths were un­
booked emergency admissions with 72.5%
Table I.
Nigeria.

Maternal mortality rate in different centres in

Reference

Year

Center

11

1973

6
8
5
7

1975
1974
1975
1976
1977
1977
1984

Eastern State of
Nigeria
Western Nigeria
U.C.H.. Ibadan
Enugu
Benin
A.B.U.. Zaria
L.U.T.H.. Lagos
U.I.T.H., Ilorin

4

12
Present series

Maternal
mortality
rate/1000

17.1
4.7
8.2
13.5
6.9
10.0
8.5
4.5

Maternal mortality at Ilorin

Table II. Maternal mortality per year (1972-1983) At the
University of Ilorin Teaching Hospital (U.I.T.H.)
'tears

Total
deliveries

Total deaths
per year

1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983

5752
6756
7718
9053
11228
12530
13120
13146
14844
15779
17278
11373

36
45
39
52
62

Maternal
death/1000

6.3
6.7
5.1
5.7
5.5
5.7
6.1
jl 5.9
4.7
5.8
4.3
3.4

72 *■

80
77
70
91
75
39

V5

causes of death (Table III), while cere­
brospinal meningitis (5.6%), pulmonary in­
fections (5.0%), and fulminating hepatitis
(3.4%) were the leading medical compli­
cations of the maternal deaths (Table IV).
The maternal deaths from hemorrhage
Table HI. Direct obstetric causes of maternal death
(excluding abortions and ectopic pregnancy).
Direct causes

of the total deaths in the age range of 20-34
years.
Hemorrhage (24.2%, consisting of ante­
partum 8.2%, and postpartum 16.0%), rup­
tured uterus (13.9%), obstructed labor
(11.9%) were the major direct obstetric

No. of
maternal
deaths

Percentage
deaths

Post panum hemorrhage
Ruptured uterus
Obstructed labor
Eclampsia
Septicemia
Antepanum hemorrhage
Severe anemia
Hemoglobinopathies
Anesthetic death

100
87
74
70
52
51
47
7
4

16.0
13.9
11.9
11.2
8.3
8.2
7.5
1.2
0.6

Total

492

78.8

(%)

7

6
I

5

4
6
c

i

r—

cZ

X

3

<
UJ
Q

<
z

2

cr

UJ

<
5

o
b
z

1

0
1972

Fig. 1.
1981.

1973

1974

1975

1S76

1977

1978

1979

1980

1981

1982

1983.

Graph showing trend in the maternal mortality rate from 1972 to 1983. Note dramatic decline in mortality rate from

I nt J Gynaecol Obstet 25

Mb

Adeioro

Table IV.

Maternal deaths from associated causes.

Associated causes

No. of
cases

Percentage
deaths

Fulminating hepatitis
Cerebro-spinal meningitis
Pulmonary infections
Native drug intoxication
Unexplained death
Tetanus
Uremia
Pulmonary embolism

21
35
31
18
9
7
7
4

3.4
5.6
5.0
2.9
1.5
1.1
1.1
0.6

Total

132

21.2

were caused by disseminated intravas­
cular coagulation (DIG) and irreversibly
hemmorrhage shock because of lack of
prompt adequate blood transfusion in al) the
cases. Seventy two of the 87 cases of rup­
tured uteri were in unbooked emergency
admissions. All of them were admitted in
shock and died in the process of active
resuscitation. The remaining cases of rup­
tured uteri were booked patients with pre­
vious cesarean section scars for cephalopelvic disproportion. They defaulted from
appointment given for elective cesarean
sections only to be seen in advance labor
with uterine rupture and collapse. All the
patients (74) who died from obstructed
labor came in after over 5 days in labor.
Forty one of them developed anuria and
died of acute renal failure, whilst the
remaining 33 died of sepsis. The eclamptic
patients (70) in this study arrived late in the
hospital, already in pulmonary failure and
comatose from overdose of native drugs.
Gram negative fulminating septicemia ac­
counted for 50 maternal deaths while severe
anemia, complicated by anemic heart failure
and cardiac arrest was the cause of the
maternal deaths in 47 cases. Hemoglobino­
pathies caused 7 maternal deaths. 5 of these
clinically died of embolism following bone
marrow infarction while the remaining two
died of acute sequestration of red cells. Two
of the four maternal deaths from anesthesia
lm J Gynaecol Obsiet 25

were due to Mendelson syndrome, whilst
the remaining two died of cardiac arrest
during intubation.
Of the indirect causes of maternal deaths,
cerebrospinal meningitis (6.2%), pulmonary
infections (5.8%) and fulminating hepatitis
(3.8%) ranked highest. They were seen with
severe overwhelming infections and failed
to respond to all medical treatment avail­
able. Native drug intoxication was associ­
ated with a 2.9% of the maternal deaths.
They all died of acute respiratory arrest.
Severe tetanus (1.1%), uremia (1.1%) and
pulmonary embolism (0.6%) were the other
indirect causes of maternal deaths.
Discussion

This review of maternal mortality rate is
based purely on hospital figures and this
does not necessarily reflect national figures.
However, irrespective of the limitations of
our survey it adds to the fact that the
maternal mortality rate is ven' high in
Nigeria. The United Kingdom maternal
mortality rate published in the report on
confidential inquiries into maternal deaths in
1973/75 was 0.11/1000. [10]. In Hongkong,
a maternal mortality rate of 0.14/1000. was
quoted for 1971 [8]. The present maternal
mortality rate of 4.5/1000, in our hospital is
unacceptably high and compares with what
existed in the United States of America and
United Kingdom around the 1930s when
their rate was between 6-10/1000 [1].
Hemorrhage was the commonest cause of
maternal death in Ilorin. and this finding
agrees with Waboso [11] and Oduntan and
Odunlami [6]. The high percentage of
maternal deaths from postpartum hemorr­
hage was attributed to unorthodox ways of
conducting the third stage of labor, whereby
a violent fundal pressure combined with
uncontrolled cord traction was employed
usually by the midwives. Also ergometrine
was administered when available after the
delivery' of the placenta by the midwives
who conducted most of the deliveries. Lack

i

Maternal mortality at Ilorin

t

lI
I

i

i

of adequate quantity of blood for prompt
transfusion, with inefficient blood bank
facilities in our hospital, in addition to the
poor response of the husbands or relations
of the patients to donate blood willingly
when so requested compounded the prob­
lems posed by hemorrhage.
The high incidence of maternal death
from ruptured uterus and obstructed labor is
in keeping with the findings of Rao [9] and
Caffrey [2]. This is attributed to a poor
communication ^system as seen in most
developing countries, illiteracy, and book­
ing at wrong places as these cases were
referred from far distant rural health cen­
tres, and arrived late in moribund condition.
It was obvious that most of these deaths
could have been prevented if delivery' had
taken place in a well equipped nearby hos­
pital, where surgical facilities were avail­
able.
Eclampsia, as a fourth leading cause of
maternal death is attributable to a large
number of unbooked cases and this finding
agrees with that of Waboso [11] and Odun­
tan and Odunlami [6] in Nigerian com­
munities. Septicemia at 8.3% of maternal
deaths, is in contrast to the findings in Zaria,
where it was the commonest cause of
maternal death [4]. Severe anemia as a
cause of maternal death was seen in 47
(7.5%) out of our 624 cases. This is in
contrast to the reports of Fullerton and
Watson-William [3], Waboso [11] and Ojo
and Savage [8] that anemia was a major
significant cause of maternal death in
Nigeria. Our finding is probably due to a
change in the pattern of anemia as a result
of widespread use of antimalarial prophyl­
axis and a higher proportion of patients
receiving hematinics during the antenatal
period.
Hemoglobinopathies as a cause of
maternal death w’as seen in 1.2% of our
cases. This agrees with the finding of 1.1%
by Oduntan and Odunlami [6] but is in
contrast to the findings of Ojo and Savage
[8] of 4.4% of their cases of maternal deaths

97

in earlier years. Improved medical care in
this group of patients over the years have
contributed to the reduced fatality. Death
from anesthesia accounted for 0.6% of our
cases. This figure is perhaps not too bad as
most of the anesthetic services were pro­
vided by nurses anesthetists. The high in­
cidence (15.8%) of infectious conditions as
a cause of maternal death in our study is in
contrast to the reports of Ojo and Savage
[8] and Oduntan and Odunlami [6] and this
is probably due to higher degree of poor
environmental sanitation and illiteracy.
The overall maternal mortality rate in the
unit continues to improve in recent years
from 5.8 per 1000 in 1981, 4.3 per 1000 in
1982 to 3.4 per 1000 in 1983. This im­
provement is not only due to the recent
increase in the medical and midwifery staff
strength but also to the quality of the staff in
the unit.
Our analysis suggests that the majority of
the maternal deaths could have been
avoided if the expectant women attended
antenatal clinics, accepted medical advice,
declined unorthodox interference and
reported in hospital in early labor. There is
no doubt that we urgently need to reduce
drastically our maternal mortality rate. This
can be achieved by an improvement in the
medical care facilities, health education of
the expectant mothers on the value of
modern antenatal care and the need for
hospital delivery. Also correct management
of the third stage of labor and better
arrangements to make blood available on a
24-h basis will reduce the maternal mor­
tality rate in the unit. Furthermore, an im­
provement in the supply of pipeborn water,
environmental sanitation, communication
and transportation systems as well as im­
provement of the nutrition and the general
standard of living of the population would
result in marked decline in our maternal
mortality rate.
It is hoped that this analysis shall further
stress the immensity of the problem of
maternal deaths in third world countries and
lm J Gynaecol Obsiet 25

98

Adeioro

stress the need for better planning of the
national health services.

Acknowledgment
The author is grateful to Professor O.
Ogunbode and Dr. F. Komolafe both of
University of Borin for their helpful advice.
1 wish to express my gratitude to Mr. Yusuf
who assisted in the analysis of the data, and
to Mrs. F.B. Ojongbede and Mr. J.S. Okoji
for their secretarial assistance.

References
1 Ademowore AS: Review of maternal mortality at Wesley
Guild Hospital, Ilesa, Nigeria. In Obstetrics and
Gynaecology in Developing Countries: The proceedings
of an International Conference Organised by the Society
of Gynaecology and Obstetrics of Nigeria (eds Ojo,
Aimakhu, Akinla, Emmanuel and Chukwudebelu) pp
287-294. 1977.
2 Caffrey KT: Maternal mortality - a continuing challenge
in tropica) medicine - A report from Kaduna, Northern
Nigeria. Afr Med J 56: 274. 1979.
3 Fullerton WT. Watson-Williams EJ: Haemoglobin SC
disease and megaloblastic anaemia of pregnancy. J
Obstet Gynaecol Br Commonw 69: 729, 1962.
4 Harrison KA: Child bearing in Zaria. A public lecture in
Ahmadu Bello University. Zaria. p 4. 1978.

I

I

Int J Gynaecol Obstet 25

u

5 Megafu U: Causes of maternal deaths at the University
of Nigeria Teaching Hospital Enugu. West Afr Med
XXIIK2): 87, 1975.
6 Oduntan SO. Odunlami UB: Maternal mortality in west­
ern Nigeria. Trop Georg Med 27: 313, 1975.
7 Okojie SE: Maternal mortality at the University of Benin
Teaching Hospital Benin City. Nigeria. In Obstetrics and
Gynaecology in Developing Countries; The Proceedings
of an International Conference Organised by the Society
of Gynaecology and Obstetrics of Nigeria (eds Ojo,
Aimakhu, Akinla. Emmanuel and Chukwudebelu) pp
280-286. 1977.
8 Ojo OA, Savage VY: A ten year review of Maternal
Mortality rates in the University College Hospital
Ibadan, Nigeria. Am J Obstet Gynaecol 118(4): 517,
1973.
9 Rao KB: Maternal mortality in a Teaching Hospital in
Southern India. A 13 year study. Obstet Gynaecol 46:
397, 1975.
10 Reports of Confidential Enquiries into Maternal Deaths
in England and Wales 1973-1975. Reports on Health
and Social Subjects London H.M.S.O.
11 Waboso FM: The causes of maternal mortality in the
Eastern States of Nigeria. Nig Med J 3(2): 99. 1973.
12 Wakile EP: Maternal mortality in Lagos. University
Teaching Hospital; A three year review. Nig Med Pract
7(5): 147. 1984.
Address for reprints:
Dr. 0.0. Adetoro
Department of Obstetrics and Gynaecology
University of Dorin
P.M.B. 1515
t
Dorin, Nigeria

u
WHO Chronicle, 40 (5): 175—185 (1986)

Maternal mortality: helping
women off the road to death
One of the widest health disparities between rich and poor is in maternal mortality.
There are, for example, more maternal deaths in India in the space of a week than
there are in all of Europe in a whole year. The estimated half million such deaths that
occur every year are all the more intolerable in that they are theoretically preventable
with current technology. As part of its long-term programme in this field, with the
support of the United Nations Fund for Population Activities, WHO convened an
Interregional Meeting on the Prevention of Maternal Mortality in November 1985. The
findings and conclusions of the 41 participants—health professionals, researchers
and policy-makers from 26 countries and agencies—are already guiding the Or­
ganization’s joint efforts with countries to help women off the road to maternal death.

Every four hours, day in, day out, a jumbo
jet crashes and all on board are killed. The
250 passengers are all women, most in the
prime of life, some still in their teens.
They are all either pregnant or recently
delivered of a baby. Most of them have
growing children at home, and families
that depend on them.
This shocking scenario, presented by Dr
Malcolm Potts at the WHO Interregional Meeting
on the Prevention of Maternal Mortality, high­
lights both the enormity of the problem and the
extent to which it has been overlooked. If the
500 000 maternal deaths that are estimated to occur
each year took place in such a concentrated and
visible way, there would be an international outcry.
But maternal deaths take place a few at a time, in
poor countries, among poor women, and often in
small villages. These deaths do not make headlines,
they just leave behind motherless children,
bereaved families, and health workers frustrated by
their inability to prevent such deaths from happen­
ing again and again.

The magnitude of maternal mortality
At the Interregional Meeting, numerous par­
ticipants presented information from their studies

of maternal deaths, defined as deaths among
women who are or have been pregnant during the
previous 42 days. Maternal mortality rates (MMRs)
at the national or local level are shown in Table 1.
Overall, Table 1 makes the point that maternal
mortality in developing countries is quite high.
With the exception of Cuba, Portugal and China
(Shanghai), all the studies found MMRs above 50,
and rates over 500 are not uncommon. This means
that, each time they become pregnant, women in rural
Bangladesh, for example, face a risk of dying that
is at least 5 5 times higher than that fiaced by women
in Portugal and 400 times higher than women in
Scandinavia.
An obvious feature of this table is that national
studies of maternal mortality in developing coun­
tries are rare. Few such countries have records of
all births and deaths, and special studies of a whole
country are difficult and expensive. On the other
hand, special studies at the local level provide a
great deal of important information. In some cases.
Based on the report of the Interregional Meeting on the
Prevention of Maternal Mortality, November 1985, prepared by
the Chief Rapporteur, Ms Deborah Maine, Senior Staff Asso­
ciate, Center for Population and Family Health, Faculty of
Medicine of Columbia University, New York, NY, USA. A
limited number of copies of the full report and of the papers
presented are available to professionally interested persons;
please write to the Division of Family Health, World Health
Organization, 1211 Geneva 27, Switzerland.

175

I

i

r
I

Table 1. Maternal mortality rates*: results of studies
presented at the meeting

Contincnt/country Maternal mortality rate Location
national
AFRICA
Egypt

local

566

northern Egypt
southern Egypt
Addis Ababa

370

four regions

566
833
’3

rural area
rural area
urban Shanghai
rural Shanghai
urban Anantapur
rural Anantapur
Bali
Selangor State
two rural areas

190
300

Ethiopia
United Republic of
Tanzania
ASIA
Bangladesh

China

22

India

I

545
874
718
70

Indonesia
Malaysia
Turkey

EUROPE
Portugal

119

national

16

LATIN AMERICA
Colombia

110

Cuba

3>

Jamaica

106

Peru

73

Cali
national
national
Callao Province

* Maternal deaths per 100000 live births.

these studies supply the only data available on
maternal deaths in a country, other than official
estimates (which are notoriously low).
As Table 1 show’s, there is considerable variation
in reported maternal mortality rates. While some of
this variation may be due to differences in study
design, in general the patterns are those that one
might expect. Countries with very high crude mor­
tality rates (such as Bangladesh, Ethiopia and India)
have higher MMRs than do those with lower crude
mortality rates (e.g., China, Colombia, Cuba,
Malaysia, Portugal and Turkey). Furthermore,
within-country differences conform to other mor­
tality patterns. For example, MMRs in China and
India were shown to be lower in urban areas, where
health services are more accessible, than in rural
areas. Similarly, the MMR reported for a northern
region of Egypt is lower than that for a region in
the southern, less developed part of the country.
A hospital MMR is the number of maternal
deaths taking place in the hospital divided by the

176

number of live births taking place in the same
institution during the same period of time. Such
rates are not good indicators of the general risk of
maternal death in developing countries. One rea­
son is that most births do not take place in hos­
pitals. On the other hand, because women wrho
experience serious complications during delivery
are more likely to try to reach a hospital, hospital
MMRs are sometimes much higher than the rate in
the population at large.
Nevertheless, there is valuable information to be
gained from hospital studies. First of all, as de­
scribed below, they are a major source of informa­
tion on medical causes of death. Secondly, they tell
us something about the functioning of the medical
system as a whole. For example, among the hospital
MMRs reported at the meeting were the following:
Nepal, 598 per 100000 live births; Nigeria, 1050;
Pakistan, 170; Sudan, 305; and Vietnam, 576. In
each of these studies, for every thousand women
who delivered a live baby in hospital, at least one
woman died, and in Nigeria the ratio was more
than one for every 100. These rates tell us that
something is wrong, because most women can be
saved with prompt and adequate medical care.
As distressing as the rates in Table 1 are, some
of them are probably still underestimates. Ex­
perience in both developed and developing coun­
tries has shown that maternal deaths are virtually
always underreported.
In industrialized countries, almost all deaths
come to the attention of medical and civil authori­
ties. Even so, there is considerable underreporting
of maternal deaths because the death certificate may
not mention the fact that the woman had recently
been pregnant. A recent study in the state of Wash­
ington, in the USA, found that maternal deaths
were underreported by 100%.
Several of the studies presented at the meeting
demonstrated the inadequacy of official statistics.
In Jamaica, the official MMR was 48 per 100 000
live births, but a national study uncovered a rate of
102. In Egypt, two separate studies found maternal
mortality rates of at least double the official rate of
90. Investigators in Colombia, India, Jamaica and
Sudan all discovered substantial underreporting
when death certificates were checked against hos­
pital records.
In developing countries, another major reason
for underreporting is that many deaths occur out­
side hospital. In a hospital study in Sudan, for
example, the number of cases collected was certain­
ly less than the actual number, as some cases were
not reported and some women arrived at the out­
patient department either dead or moribund and
were immediately taken back by the relatives. Data

r

from Egypt, India, Indonesia, Malaysia and Turkey
showed that large proportions of maternal deaths
took place either at home or on the way to the
hospital. These proportions ranged from 24% of
deaths in Turkey to 82% in rural India. In Bang­
ladesh, hospital staff were aware of
< only 4% of the
maternal deaths discovered by researchers.
In general, the studies presented demonstrated
that the larger the number of sources of data em­
ployed, the more maternal deaths are discovered.
In India it was learned that even schoolchildren can
be a valuable source of information on deaths that
might otherwise be overlooked.

The causes of maternal mortality
Dr Fathalla,1 the meeting’s Chairman, em­
phasized in his opening address that the causes of
maternal deaths are complex. To do this, he de­
scribed the case of Mrs X:
Mrs X died in hospital during labour. The at­
tending physician certified that the death was
from haemorrhage due to placenta praevia. The
consulting obstetrician said that the haemor­
rhage might not have been fatal if Mrs X had not
been anaemic owing to parasitic infection and
malnutrition. There was also concern because
Mrs X had only received 500 ml of whole blood,
and because she died on the operating table while
a caesarean section was being performed by a
physician undergoing specialist training. The
hospital administrator noted that Mrs X had not
arrived at the hospital until four hours after the
onset of severe bleeding, and that she had had
several episodes of bleeding during the last
month for which she did not seek medical atten­
tion. The sociologist observed that Mrs X was
39 years old, with seven previous pregnancies
and five living children. She had never used
contraceptives and the last pregnancy was un­
wanted. In addition, she was poor, illiterate and
lived in a rural area.
Why did Mrs X die, and how could her death
have been prevented? Dr Fathalla pointed out that
there were a number of points at which Mrs X
could have been helped off the road to death. In
order to identify these, and to design and imple­
ment effective programmes, the various kinds of
causes need to be understood.

Medical factors

There is considerable variation in ways of classi­
fying medical causes of death. For example, a wom­

an who bleeds to death when her uterus ruptures
may be listed as dying from either haemorrhage or
ruptured uterus. Nevertheless, the final “causes” of
maternal deaths—those diagnosed and recorded by
medical personnel—are remarkably consistent
throughout the developing world.
Maternal deaths are usually divided into three
categories: “direct” obstetric deaths; “indirect”
obstetric deaths; and unrelated deaths. Direct
obstetric deaths are those resulting from complica­
tions of pregnancy, delivery or their management.
Indirect obstetric deaths are the result of the ag­
gravation of some existing condition (such as hep­
atitis or heart disease) by pregnancy or delivery.
In developing countries, as the studies presented
at the meeting confirmed, direct deaths constitute
50—98% of all maternal deaths, and haemorrhage,
infection and toxaemia together make up at least
half of all maternal deaths in 11 of the 13 countries
for which this information was provided. In a few
studies, some other condition was listed as one of
the three leading causes of death. Most often, this
other condition was illegal induced abortion but in
two cases it was embolism. Ruptured uterus, hepa­
titis, anaemia and obstructed labour were each cited
once as one of the three leading causes of maternal
deaths.
The major medical causes of maternal deaths in
developing countries are thus already known, but
these diagnoses are usually just the last stretch of
the road to death.
Health service factors

The fact that medical causes of death are not the
whole story emerged clearly from the meeting’s
discussions of avoidable maternal deaths. The
medical records of women who had died had been
analysed in nine countries in order to identify fac­
tors that contributed to their deaths. The inves­
tigators found that 6^-8o%o of direct maternal deaths,
and 88—98% of all maternal deaths, could probably have
been avoided with proper handling. In a number of
cases, the researchers specifically stated that they
had evaluated the avoidability of deaths not by
standards of care under the best of circumstances,
but by standards realistic under the circumstances
prevailing in that country at the time. For example,
in Turkey, 51% of maternal deaths were judged to
be avoidable within the existing health system, and
another 24% avoidable with an improved health
’ At the time Dr M.F. Fathalla was Dean, School of Medi­
cine, Assiut University, Egypt. He is now Responsible Officer
for Research and Development, Special Programme of
Research, Development and Research Training -in Human
Reproduction, WHO, Geneva, Switzerland.

i 77

>

Imagine a jumbo jet crashing
every four hours somewhere in
the world. All 250 passengers on
board are women in their prime.
There are no survivors. This is
the magnitude of maternal mor­
tality.

I

I

Photo Camera Presst/Len Sirman Press.

r

I

I

system. In most cases, investigators identified more
than one avoidable factor that contributed to each
death.
Deficient medical treatment of complications was of­
ten an important factor. Mistaken or inadequate
action by medical personnel was judged to be a
contributing factor in between 11% and 47% of
maternal deaths in the developing countries stud­
ied.
hack, of essential supplies and trained personnel in
medical facilities was also mentioned frequently as
a contributing factor. In United Republic of Tan­
zania, lack of blood for transfusions, drugs and
equipment was a factor in more than half of the
deaths studied. In Jamaica, only 6 of the 18 hospital
deaths from haemorrhage took place in hospitals
that had a blood bank.
hack of access to maternity services is another crucial
step on the road to death. The studies in Cuba,
Egypt, Indonesia, Jamaica, Turkey and United
Republic of Tanzania demonstrated that maternal
mortality rates are increased in areas where access
to a hospital is difficult, and where women are
likely to arrive at the hospital, if at all, in a serious
condition. In Nepal, for example, 32% of women
who died in the hospital arrived in very poor con­
dition, and another 17% arrived unconscious.
hack ofprenatal care was frequently mentioned as
a contributing factor. For example, in Portugal,
more than half of the women who died had not
received prenatal care, compared with one-third of
women in the country as a whole. In Nigeria, in all
age-parity groups, MMRs were drastically lower
among women who had had prenatal care than
178

among those who had not, although “the risks of
teenage pregnancy and high parity were still very
evident”. Some data, however, indicated that more
research is needed on the role of prenatal care. For
example, in Viet Nam “very few adverse events
were found at antenatal visits”. Furthermore, the
community-based study of maternal mortality in
Addis Ababa illustrates the point that “antenatal
care and selection of high-risk women are not an
end in themselves”. All three women in the Ethio­
pian study who died of haemorrhage had had
prenatal care, but had delivered at home. This
shows that women must be convinced of the
benefit of referral and that, above all, services must
be accessible.
Another problem in interpreting data on prena­
tal care is the difficulty of distinguishing the wellknown effects of poverty on maternal health from
the effects of lack of prenatal care. In Nepal, for
example, only 34% of illiterate women had prenatal
care, compared with 91% of women with a college
education.
Reproductive factors

For decades it has been known that certain
groups of women—very young women, those aged
35 or older, and women who have already borne
four or more children—are at especially high risk
of dying during pregnancy and delivery. Many of
the studies presented at the meeting confirmed this.
Maternal age. Data showing higher MM Rs
among women aged 35 or older were presented for
eight developing countries. In six of the studies

i

that provided the data to make this comparison,
women aged 35-39 were from 85% to 461% more
likely to die from a given pregnancy than women
aged 20-24 (relative risks, 1.83-5.61). One study,
a case/control study in United Republic of Tan­
zania, did not show this expected relationship.
The same studies that showed an excess of deaths
among older women showed an excess among
women younger than 20, with the exception of
Cuba. Increased risks of death were especially pro­
nounced in Ethiopia, Indonesia and Portugal.
Again, the Tanzanian study did not show any dif­
ferences by age between women who died and
those who did not.
Parity. Although information on parity is more
difficult to obtain than information on age, several
studies also confirmed the increased risk of death
associated with having many children. In Jamaica,
compared with women having their second child,
those having their fifth through ninth births were
43% more likely to die. In Portugal, women having
their fifth birth were three times as likely to die as
women having their second, while women having
their sixth or later birth were at even greater risk.
The importance of these data is that the practice
of family planning could prevent a great many
deaths of women of unfavourable age or parity.
Unwanted pregnanty. Of course, given the high
overall rates of maternal death in poor countries,
the impact of family planning would be important
if unwanted pregnancies were averted at any age or
parity. This point is vividly illustrated by data from
the governorate of Menoufia in Egypt and the
island of Bali in Indonesia. When similar studies
were done in both places, a striking difference was
found in maternal mortality rates. In Bali there
were 718 deaths per 100 000 live births, compared
with 190 in Menoufia: 278% higher. However,
when the risk of childbearing was expressed in
another way—as maternal deaths per 100 000 mar­
ried women aged 15-49—the difference was greatly
reduced. In Bali, there were 69 deaths per too 000
women, compared with 45 in Menoufia: an excess
of only 53%. The reason for this seeming paradox
is that fertility rates are much lower in Bali than in
Menoufia, largely owing to the use of family planning‘

Illegal induced abortion is a major killer of
women, as the studies presented at the meeting
amply demonstrated. It was responsible for 7-50%
of maternal deaths, the median being 15%. As high
as these percentages are, many of them are under­
estimates because women who have illegal abor­
tions are reluctant to seek formal medical help. In
Ethiopia, for example, four of the six women who
died on the way to hospital had had an illegal

induced abortion. Reluctance or inability to get
medical care results in a selective underreporting of
abortion deaths. In India, 11 % of hospital deaths
were due to abortion, compared with 17% of
deaths at home in rural areas. Clearly, since induced
abortions occur in cases of unwanted pregnancy,
family planning could substantially reduce the
number of deaths from this cause.
Finally, unwanted pregnancy contributes to ma­
ternal deaths in ways which are not yet understood.
The Ethiopian study found that women who had
an unwanted pregnancy were less likely than other
women to seek prenatal care. In addition, two
deaths of pregnant women by poisoning were at­
tributed to unwanted pregnancy.

Socioeconomic factors
Socioeconomic factors undoubtedly play a large
role in maternal deaths, but how and why are still
obscure. What is known is that poverty is clearly
a high-risk factor. It is also known that poor
women are less likely to have formal education than
wealthy women, and are less likely to be in good
health and to seek (or receive) medical care. Which
of these factors are causes and which are effects, and
how can this vicious circle be broken? Much more
research needs to be done to answer these ques­
tions.
The kinds of questions raised above are also
relevant to health problems such as infant mortal­
ity. But another (and even less well studied) aspect
of socioeconomic status has special importance in
maternal deaths, and that is the status of women.
As papers presented by the Egyptian and Nigerian
participants emphasized, “in almost all societies in
the past, and in many societies in the present,
women are a socially disadvantaged group . . . The
status of women affects their nutrition, reproduc­
tive behaviour, utilization of health care services
and vulnerability to harmful traditional practices”.
The ramifications of the status of women are so
far-reaching that it may be that “nothing will really
change in so far as maternal mortality is concerned
until attitudes towards women change and people
are sufficiently motivated to improve their living
conditions”.

The papers presented and the plenary sessions
strongly indicated that a major new initiative to
prevent maternal deaths should be mounted—and
was in fact overdue. Furthermore, there was agreeI79

f

place outside hospitals. Furthermore, a sizeable
proportion of serious complications cannot be
?
predicted beforehand. Therefore, while efforts
■
must be made to upgrade hospital care and to refer q
high-risk women as early as possible, services need 4
to be designed to reduce the distance between
pregnant women and the care they require.
A variety of approaches are possible. Women
who are likely to have complications can be sent to
Policy initiatives
maternity waiting homes. These are facilities where
pregnant women can come in the last week of
In order for there to be a concerted and effective
pregnancy, stay while they await delivery, and have
effort to reduce maternal deaths in developing
either a supervised normal delivery or prompt
countries, maternal mortality must be given high
transfer to a medical facility if complications arise.
priority. As with all areas of action, initiatives need
Experience with waiting homes in Colombia,
to be taken at a number of levels—starting at the
Chile, Cuba, Uganda and Malawi has shown that
global level, with WHO helping to set policy and
they can be successful and need not be expensive,
coordinate actions and resources.
as the community can provide much of the labour
It was strongly recommended that the Member
and supplies.
States of WHO should designate maternal mortal­
However, in the large proportion of cases in
ity as one of the global indicators of “health for all
which complications cannot be predicted, more
by the year 2000”. Furthermore, WHO should help
effective means of treating complications must be
draw the attention of Member States to the greatly
made available at the first referral level, including
elevated risk of death faced by women in high-risk
the establishment of more basic obstetric facilities.
groups if they become pregnant.
These need not be new facilities. Health centres
While WHO can lead the global effort to reduce
could be upgraded to provide essential maternal
maternal deaths, the effectiveness of this effort
health services: vacuum extraction deliveries;
depends mostly on national governments. To be­
blood transfusions; simple general and/or local
gin with, governments must make maternal mor­
anaesthesia; caesarean section; suction curettage for
tality a priority public health issue, and should
incomplete abortion; insertion of intrauterine de­
review their policies and programmes with an eye
vices; and tubal ligation and vasectomy.
to preventing maternal deaths. Policy reviews
Promising approaches were suggested for each
should cover such issues as removing obstacles to
of the major causes of death.
family planning, e.g., taxes on and other barriers to
Haemorrhage. Postpartum haemorrhage is dif­
using or importing contraceptives.
ficult to predict and there is often little time or
Professional societies too have a role to play. In
opportunity to transport the woman to a hospital
order to lower maternal mortality in poor coun­
for blood transfusion. Therefore, any trained per­
tries, services must be spread more widely and in­
son who is considered capable of doing a delivery
novative programmes must be tried and assessed.
should be trained to handle this life-threatening
This will not be possible without the strong leader­
complication through the use of oxytocic drugs
ship of professional societies such as medical asso­
(which contract the uterus and its blood vessels),
ciations, both internationally and nationally.
manual removal of the placenta, and then ad­
ministration of broad-spectrum antibiotics. In
addition, the use of plasma expanders at health
Programme initiatives
centres that cannot provide transfusions should be
It is clear from the persistence of high rates of explored.
Antepartum haemorrhage can be predicted when
maternal mortality and morbidity that current pro­
there is third-trimester bleeding with placenta praegrammes are not adequate. Progress will require
via. In these cases, early referral to a facility where
bold and determined new thinking and effort. Pro­
blood transfusion and caesarean section are avail­
grammes should rest on the axiom that all services
should be provided at the most peripheral level of able is crucial. However, in many cases antepartum
haemorrhage cannot be predicted. Therefore, there
the health care system consistent with efficacy.
The design of services should be guided by what
is an urgent need to shorten the distance between
the place of delivery and a facility where emergency
has been learned from studies such as those present­
care can be provided. In addition to upgrading
ed at the meeting. For example, in many countries
peripheral health facilities, attention must be paid
most deliveries and many maternal deaths take

mcnt that much could be accomplished. The re­
maining question, then, was how best to begin.
Recommendations for action at a number of levels
—policy, programme, training and research—had
been prepared during two intensive days of work­
ing group sessions, and were discussed in the plenary
session.

i8o

I

I

I

I

I

to the key role of transportation. An effort should
be made to make all kinds of government vehicles
available in emergencies, rather than relying on
scarce (or non-existent) health department vehicles
alone.
Infection. Deaths from infection can be greatly
reduced (as they have been in China) through clean­
liness during delivery. Providing traditional birth
attendants (TBAs) with delivery kits is one way to
encourage asepsis. Adding antibiotics to these kits,
for use in cases of prolonged labour or premature
rupture of the membranes, could prevent many
maternal deaths in areas where physicians are
scarce.
Toxaemia. Only good prenatal and medical care
can prevent the majority of deaths from this cause.
However, sedatives for treatment of severe tox­
aemia should be made available at the primary care
level.
Unwanted pregnanq. As the studies presented at
the meeting showed, unwanted pregnancy contri­
butes to maternal mortality in a number of ways
—e.g., in the number of births to women in highrisk groups and the number of pregnancies per
woman. The most dramatic way in which unwant­
ed pregnancies contribute to maternal deaths is
through illegal induced abortion. Because these
pregnancies are, by definition, unwanted, this is an
area in which primary prevention holds great pro­
mise.
Family planning is the first line of defence
against illegal abortion, and education about avoid­
ing unwanted pregnancies should be provided in
schools, at all levels of the health care system, and
during all contacts with pregnant and recently de­
livered women. Special attention should be paid to
counselling women who are being treated for com­
plications of abortion, in order to help them avoid
repeated unwanted pregnancies and abortions. Fur­
thermore, whatever the accepted indications for
legal abortion in a country (and there are usually
some), this service should be made widely avail­
able, rather than being available only to wealthy
women in urban areas.
Obstructed labour. While there are certain groups
of women who are at especially high risk of ob­
structed labour (e.g., women of small stature,
women having their first birth, and women having
their sixth or later birth), in many cases this com­
plication is not predictable. So, again, access to
emergency services is essential. In the case of ob­
structed labour, much could be accomplished by
educating TBAs to be prompt in sending women
who are not making satisfactory progress in labour
to a facility where they can get medical care, such
as a caesarean section.

Anaemia. Depending on the cause of anaemia in
a particular region, iron and folate supplements,
malaria prophylaxis and/or treatment, and treat­
ment of hookworm disease and schistosomiasis
should be provided to pregnant women at the
primary care level.
Tetanus. In addition to being a major killer of
newborns, tetanus is a common cause of maternal
deaths in some areas (Bangladesh, India, In­
donesia). The administration of tetanus toxoid to
all women, especially pregnant women, should be
a high priority.
Training initiatives

To implement programmes successfully, train­
ing is crucial. Some of the needs for training are the
following.
Traditional birth attendants. TBAs are often the
first (if not the only) health care workers with
whom pregnant women in poor countries have
contact. Therefore, it is essential that they be made
as effective as possible through training, super­
vision and support.
A major role of TBAs should be referral—as­
suming, of course, that there are health care facilities
to which women can be referred. Topics suggested
for TBA training in referral include: recognition of
risk factors (e.g., age, parity, poor obstetric history,
bleeding during pregnancy); detection of anaemia;
recognition of infection, prolonged labour and ex­
cessive blood loss; and referral to a source of legal
abortion.
TBAs should also be given the training and
supplies to prevent or treat complications when­
ever possible. Preventive measures include use of
antiseptic techniques in delivery, administration of
drugs to reduce anaemia, and provision of con­
traceptives. Treatment skills could include first aid
for treatment of haemorrhage (application of pres­
sure, elevation of limbs, use of oxytocic drugs) and
safe removal of retained placenta.
Health centres. If health centres are to fulfil their
potential in preventing maternal deaths, centre
personnel need the training and supplies to be
effective. Suggested areas for training include:
recognition of blood pressure abnormalities and
anaemia; use of antibiotics, intramuscular iron sup­
plements, oxytocic drugs and plasma expanders;
and repair of lacerations. In areas where there is no
physician available to perform life-saving caesarean
sections, the feasibility of teaching trained mid­
wives to do this operation should be explored.
Referral hospitals. As the studies of avoidable
deaths demonstrated, hospital personnel need addi­
tional training in treatment of serious complica181

i'

J
J

K3

I
A traditional birth attendant
examines a mother-to-be. Xo
matter how good the care at pri­
mary health level, a certain
proportion of women will die
of unforeseeable complications
during delivery if the health cen­
tre or hospital is too far away.

tions. For example, it was suggested that special
teams of health care personnel be established for
coping with haemorrhage and eclampsia. Personnel
in these facilities need to have banked blood avail­
able, and to be able to manage such catastrophic
events as uterine rupture.

Research initiatives

I

Three broad types of research were discussed:
research on appropriate technology for preventing
maternal deaths; health systems research on inno­
vative programmes; and epidemiological research
on the incidence and causes of maternal deaths.
Appropriate technology research. A wide variety of
appropriate technology issues were suggested for
future research. These included such important
topics as: simple, inexpensive methods for detect­
ing and measuring anaemia; durable tubing for
vacuum extractors; appropriate plasma expanders
for use at health centres; and the content of delivery
kits for TBAs.
Health services research. Evaluating service deliv­
ery systems, especially innovative ones, is crucial if
scarce resources are to be used effectively. Promis­
ing topics for health services research include:
appropriate therapy for anaemia, such as new iron
preparations; the use of prophylactic antibiotics in
cases of prolonged labour; the role of maternity
villages; and the delegation of basic obstetric func­
tions such as caesarean section and suction curet­
tage to a more peripheral level.
182

-

-

Epidemiological research. Both for shaping policy
and for designing programmes, it is important that 1
more research be done on maternal mortality and
morbidity rates, and on their causes. It was recom­ i
mended that all Member States of W HO should, no
later than 1995, be able to provide reliable estimates
of their MMRs. Also by 1995, Member States «
should have begun research on the underlying b
causes of maternal deaths.
Four types of information on maternal mortality
should be sought. First is the absolute number of ;
deaths. As the studies presented at the meeting
showed, obtaining this information is not easy. ’
Nevertheless, even
incomplete counts can
sometimes be useful for policy purposes—e.g., j
when a small developing Country is found to have j
more maternal deaths a year than a very large |
developed country.
Secondly, countries should collect information I
on the rate of maternal mortality. As was shown •
above, hospital studies are not a good method for ;
determining MMRs in developing countries.
i
' The third type of information countries should i
gather is data on the characteristics of women who ]
die. These are especially valuable when compared ;
with information on women who do not die. Case­
control studies are a relatively inexpensive way to
accomplish this.
Lastly, data are needed on the causes of maternal I
deaths: clinical, health services, reproductive, and I
socioeconomic factors. Priority should be given to ’
research on the risk factors that have the greatest j

I

Table 2. The suitability of data sources and methods for obtaining selected information on maternal mortality
(Key: i=best, 2=satisfactory, 3 = poor, 4=not appropriate, UNK=unknown)

Data source/method

Information sought

No. of
deaths

Vital statistics
Routinely classified maternal deaths
Birth-death record linkage
Investigate all deaths of women aged 15—49

2

Characteristics
of the deceased
women

Medical
causes

5

5

Other causes
(social, etc.)

3
2

«+

Hospital records
Maternal deaths in obstetrics/gynaecology service
Maternal deaths in all services
Case-control studies
Obstetrician/gynaecologist peer review

4
4

3
3
4
4

Health worker interviews
Obstetricians/gynaecologists
All health workers (MCH/family planning)

3
3

3
3

Community studies
Identify/investigate deaths of women aged 15-49
Prospective monitoring of pregnant women up to
6 weeks after end of pregnancy
Household survey
Discussions with groups of knowledgeable/
local people
Case-control studies

!

Maternal
mortality
rates

2

2

3
3

2

2

2

2

2

2+

2

2

1

1-2
2

2

UNK

4
4

4

Confidential inquiries
Vital statistics
Multiple data sources

3

4
4

Indirect demographic methods

UNK

UNK

relevance for improving the provision and use of
health services.
The participants considered that it would be
useful for WHO to draw up a document describing
research methods that can provide these kinds of
data, and the circumstances under which each
method is most and least useful (see Table 2). If
necessary, new, low-cost methods for doing re­
search on maternal health should be developed.
WHO should also consider providing training
courses on study design.
Finally, it was recommended that countries
should begin collecting data on maternal morbid­
ity. Most countries have no idea of the magnitude
of this problem, although it can be assumed to be
large. For example, a study in India found that for
each woman who died a maternal death, there were
18 who survived with severe (and sometimes per-

2

UNK

UNK

4

4

4

manent) complications. An effort should be made
to take advantage of existing opportunities to
gather morbidity data, e.g., during contraceptive
prevalence surveys.

*
In her closing address. Dr Angele PetrosBarvazian, Director of WHO’s Division of Family
Health, said that the meeting should be seen as “the
end of the beginning”. The actions recommended
would now have to be carried out by WHO in its
broadest sense—not just headquarters but the re­
gional offices, Member States, and professional and
nongovernmental organizations. Only with com­
mitment on the part of all involved could women
be helped off the road to maternal death.

183

Position: 596 (7 views)