THE STRATEGY OF RISK APPROACH IN ANTENATAL CARE: EVALUATION OF THE REFERRAL COMPLIANCE

Item

Title
THE STRATEGY OF RISK APPROACH IN ANTENATAL
CARE: EVALUATION OF THE REFERRAL COMPLIANCE
extracted text
0277-9536(94)E0095-A

Sot. Set. Med Vol. 40. No. 4, pp. 529-535. 1995
Copyright £ 1995 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
0277-9536/95 S9.50 4- 0.00

SDA-RF-CH-1.32

ck-1

I

THE STRATEGY OF RISK APPROACH IN ANTENATAL
CARE: EVALUATION OF THE REFERRAL COMPLIANCE
B. Dujardin,1 G. Clarysse,2 B. Criel,1 V. De Brouwere2 and N. Wangata4
'Public Health Unit, Institute of Tropical Medicine, Nationalestraat 155, B.2000 Antwerpen, Belgium,
!UNICEF Segou, Mali, 5WH0 Rabat, Morocco and “Projet same Pour Tous, BP 4832 Kinshasa Gombe,
Zaire
Abstract—The main goal of antenatal care in developing countries is to identify women whose pregnancy
or delivery is likely to raise problems and to refer them at the appropriate time to a hospital facility where
the necessary medical equipment and expertise (vacuum extractors, cesarian sections, human skill, etc.)
is available. This approach, which is known as the Risk Approach (RA) strategy, is expected to
significantly reduce maternal morbidity and mortality. However, the RA will function properly only if
the women identified at risk agree to give birth in a hospital on the one hand, and if they can indeed reach
this hospital on the other hand. In this article the authors assess to what extent women with a risk of
difficult labor (nulliparous or primiparous women under 150 cm, history of previous difficult delivery or
stillbirth, women with transverse lie) agreed to give birth in a hospital. This descriptive survey, which
covered 5060 pregnancies monitored in the Kasongo District, Maniema, in eastern Zaire, showed that the
referral success rate in this socioeconomically very disadvantaged region was only 33%, despite some
favorable conditions, such as a strong emphasis on community participation, a complementarity of health
centers and hospital, and the absence of financial barriers within the health services system. Of the various
hypotheses tested, the geographic accessibility of the hospital and the parturient’s perception of the risk
status were the two most important factors determining the compliance rate. A stratified analysis shows
that the intensity of the parturient’s perception has a different impact on compliance whether rural or
urban situations are considered. In their conclusions, the authors stress the importance of the problem
and the need for additional qualitative studies (open interviews, focus group discussions) to better
understand the reasons of this low compliance. The phenomenon observed in Kasongo is definitely not
a unique one, and unawareness of this problem is likely to be one of the reasons for the low success of
pregnancy monitoring programs when they are assessed in terms of reduction of maternal morbidity and
mortality.
Key words—antenatal care, risk approach strategy, referral compliance

INTRODUCTION

METHODS

1 ne Safe Motherhood Initiative advises to implement
the Risk Approach (RA) strategy as an effective
means to reduce maternal mortality [1], In this ap­
proach, women with the highest probabilities of
suffering from pathological conditions during preg­
nancy or delivery must be identified by a Risk Factor
(RF) and referred to a hospital facility for special
monitoring and delivery in a maternity ward. Many
publications have been focusing on the RA, the
identification of the RFs, the evaluation of their
predictive value, etc. [1-4]. Nevertheless, the success
of this method depends just as much on its compli­
ance, i.e., to what extent the women it targets will
accept and carry out the referral advised by the health
service.
In this article, which concerns the case of
women with a risk of difficult labor, we evaluate the
degree of compliance with the referral decision
and analyze some of the factors that may influence
it.

This descriptive survey was carried out in the
Kasongo District (population 200,000) located in the
Maniema Region in eastern Zaire. The analysis was
done in 1988, and relies upon the antenatal clinic
(ANC) records of 11 health centers (7 rural and 4
urban centers) out of the 20 existing ones. These
centers were chosen because of the quality of their
records. Three of these 11 centers also participated in
a similar research in 1986, 1987 and 1990. No notice­
able changes that might have caused variations in the
compliance rate occurred in the Kasongo District
between 1986 and 1990. The data pertaining to this
period were therefore added to the 1988 data as they
were of good quality (<10% noresponse rate). In
total, 5478 ANC records were studied. However, for
418 of them information was lacking indeed, the
place of delivery was not known. These records were
therefore excluded from the analysis. This was
deemed acceptable since the proportion of cases with
presence of a RF in that particular population (7.9%
529

530

B. Dujardin et al.

i.e. 33/418), was similar to the proportion of cases
with a RF in the remaining population (6.4% i.e.
324/5060). This difference is not significant, P = 0.24.
ANCs in the Kasongo District are organized and
carried out at the health center level. Women are
advised to give birth at home if no risk factor is
identified, or to deliver in the hospital if one is
identified. The development of community partici­
pation is considered a priority in this district. The two
levels of the health system, i.e. the Health Centers
(HCs) and the referral hospital (the only one in the
district), function in a way so as to complement each
other. The cases referred by the HCs are given
priority by the hospital team. Moreover, it is import­
ant to point out that the antenatal care attenders pay
a onetime fee which covers the cost of the referral
consultation and the cost of the admission when the
latter is justified. These factors are favorable con­
ditions expected to yield a high referral compliance
rate. However, travel costs and other indirect costs
(lost of working days ...) have to be supported by the
women and their families.
Two visits are scheduled: one during the first
trimester with the purpose to detect pre-existing risk
factors; the other at 9 months with the purpose to
provide a final assessment before the delivery. If
necessary, additional appointments are scheduled.
The effectiveness of the various risk factors, especially
the five risk factors of difficult labor mentioned
above, i.e. nulliparity or primiparity and small stature
(under 150 cm), difficult labor antecedents (cesarian
section or use of vacuum extractor), stillbirth at
previous delivery, and finally transverse lie of the
foetus, was identified and evaluated during an epi­
demiological survey conducted in 1975 [5].
The compliance with the referral decision (i.e. the
proportion of at risk women who accepted to go to
the hospital and who indeed delivered in the hospital)
was determined as follows: for each referral decided
upon identification of a given risk factor, a follow-up
was done so as to check where the woman had
actually given birth (at home, which meant ‘rejection’
of the referral; or in the hospital, which meant
‘acceptance’ of the referral) and whether the delivery
was difficult or not. A delivery was considered
‘difficult’ in the following cases: a ruptured uterus, a
cesarian section for difficult labor, the use of the
vacuum extractor, a stillbirth and a neonatal death
(the deaths of children with birth weights under
2500 g were excluded because the probability of a
difficult labor to be the cause of death is low in such
cases). Given the scarcity of resources, the doctors
tended to be very specific in their indications and
obstetric surgery of vacuum extraction were per­
formed only when absolutely necessary.
Epi info software (version 5.0, Centers for Disease
Control, Atlanta) was used for the statistical analysis.
For the two main compliance determinants, a
stratified analysis (Mantel-Haenszel /2) was done in
order to identify a possible interaction.

In the presentation of the results-we will (imit
ourselves to the analysis of the compliance with the
referral decision based on these different RFs and to
the identification of determinants that may have
influenced it.

RESULTS

Of the 5060 pregnancies studied, 324 (6.4%)
women were identified at risk for difficult labor (one
or more of the RFs present) and were referred to a
hospital for delivery. Out of these 324 referred cases,
108 (33.3%) women gave birth in the hospital mater­
nity, whereas it was the case for only 505 (10.7%)
women out of the 4736 non-referred cases. This
difference,
which
is
statistically significant
(P < 0.001), illustrates the effectiveness of the referral
strategy.
Nevertheless, the overall success rate of referral
(percentage of referred women who actually gave
birth in the hospital) was poor, only 1 woman out of
3 at risk delivered in the hospital. We broke down the
overall rate by difficult labor RF in order to better
understand the reasons for this low figure. In Table
1, several explanatory hypotheses for this overall
poor result are presented.
Table 2 presents the referral rates (referral rates
refer to the proportion of women identified at risk
during ANC, to whom it was advised to deliver in the
hospital) and the compliance rates (compliance rates
refer to the proportion of women at risk and thus
referred, who indeed delivered in the hospital) for
each RF. Table 3 presents the proportions of difficult
deliveries, amongst women identified at risk during
their pregnancy. Figure 1 enables us to check the
validity of the three first hypotheses.
Our first hypothesis, i.e. “the overall rate may
conceal large factor-specific variations”, is partly
founded according to these results. Compliance rates
vary according to the risk factor considered. How­
ever, this overall significant variation (P < 0.05) is
not very wide, ranging from 24.8% and 26.4% for
‘primiparity and length under 150 cm’ and ‘previous

Table 1. Hypotheses to lest in order to identify the various factors
_____________ influencing the referral compliance
1.

The overall rate may conceal large factor-specific variations

2.

The RFs may be too frequent (i.e. too sensitive and not specific
enough)

3.

The positive predictive value (i.e. the probability of a difficult
labor when the RF is present) is loo low

4.

The patient’s behavior does affect the referral compliance

5.

The geographical accessibility is a major constraint

6.

Referral compliance will increase if antenatal care regularity
increases
5
7

7. The quality of staff will increase the referral compliance

Tabic 2. Referral and compliance rales for each RF

Risk factor

Referral rates
■V (%)

Compliance
rales
/V (%)

1. Nulliparous under 150 cm

89/815(10.9%)

33/89 (37.1%)

2. Pnmiparous under 150 cm

53/592 (9.0%)

14/53 (26.4%)

3. Transverse lie

31/5060(0 6%)

15/31 (48.4%)

4. Previous difficult delivery

35/4245 (0.8%)

15/35 (42.9%)

5. Previous still-bom child

137/4245 (3.2%)

34/137(24.8%)

P <0.001

P < 0.05

Significance level

Table 3. Difficult deliveries amongst women presenting a RF
Risk factor

Difficult
delivery (%)

Outcome
unknown (%)

Nulliparous under 150 cm

15,81 (18.5%)

8/89(9.0%)

Prmiparous under 150 c’*'

3/50(6.0%)

3/53(5.7%)

Transverse lie

2/30(6.6%)

1/31 (3.2%)

Previous difficult delivery

3/33(9.1%)

2/35(5.7%)

Previous still-born child

14/133(10.5%)

4/137(2.9%)

Significance level

Total number of women

P =0.17
35/303* (11.6%)

21/324(6.5%)

*Women may present more than one risk factor.

stillborn child’ to 42.9% and 48.4% for ‘previous
difficult delivery’ and ‘transverse lie’, respectively.
This hypothesis proves valid and in some situations,
it might be interesting to put more emphasis on these
RFs that are best accepted.
Our second hypothesis, i.e. “the risk factors may be
too frequent, i.e. too sensitive and not specific
enough”, implying that too many women are reft^d, thereby decreasing the acceptability of the
referral by the women, seems to be confirmed. In­
deed, the rarer factors (‘transverse lie’ and ‘previous
difficult delivery’) implying lower referral rates, yield
the highest compliance rates. When the RFs are
aggregated according to their frequency, i.e. corre­
sponding referral rates above or below 1%, then the
compliance rate is 45.5% (30/66) when the referral
rate is less than 1%, and 29.0% (81/279) when it is
>1%. This difference is statistically significant
(P <0.01).
Our third hypothesis, i.e. “the positive predictive
value (i.e. the probability of a difficult delivery when

RISK
ra Compliance Rate

the RF is present) is too low, and these factors are
thus poorly accepted by the pregnant women” does
not seem to be confirmed. In other words, even if the
positive predictive values (PPV) which are a measure
of the efficacy of the RF, range from 6.0 to 18.5% (see
Table 3), they do not vary significantly from one RF
to another (P = 0.17). Nor does, as shown in Fig. 1,
the PPV seem to be linked to the compliance rates.
For instance, the RF ‘transverse lie’ has a predictive
value (6.6%) grossly similar to the one of RF ‘pre­
vious stillborn child’ (10.5%), although the latter has
the lowest compliance rate (24.8%). Other biases may
nevertheless interfere. For example, only cases of
difficult labor requiring technical intervention (vac­
uum extraction, cesarian) were retained. This ex­
plains in part the low PPV. The exact ages of the
pregnancies were also difficult to assess. This may be
the cause of large errors in the estimation of foetal
position at the last contact. Indeed, the health
workers may then diagnose an eventual normal

FACTORS

SM Referral Rate

EL) Positive Pred. Value

Fig. 1. Decreasing compliance rates with the corresponding referral rate and positive predictive values for
each RF.

B. Dujardin et al.

532

Table 4. Variations in referral acceptability according to the parturi­
ent's risk perception

Intensity of perceived risk

Hospital delivery

Woman feels herself to
be in danger

30/66 15.5%)

Woman identified as being
at risk

47/142 (33.1%)

Woman has lost a
neonate

34/137 (24.8%)

Level of significance

P < 0.05

position to be a transverse lie and convince the
parturient to give birth in a hospital: “Your delivery
will be very difficult and you have to go to the
hospital ...”, thereby explaining a higher than aver­
age compliance rate but a low predictive value.
Our fourth hypothesis is linked to the patient’s
behavior. Although we did not specifically analyze
the women’s individual health behavior, the follow­
ing comments can be made. A woman who has
experienced a difficult delivery before or who is told
a few weeks before term that the delivery is very likely
to be complicated and that she may die because of a
transverse lie, would a priori accept the referral more
easily than a woman whose previous child was still­
born or a woman who is a small primipara. The
compliance rate for the RF ‘previous stillbirth’ will
depend on the cultural perception of this event, which
may vary from one population to another. It also
depends on the woman’s perception of the link
between the delivery and the cause of the child’s
death. It seems that women in Kasongo consider
perinatal mortality to be a ‘chance mishap’ during her
reproductive years; a ‘mishap’ that is not as emotion­
ally loaded as a neonatal death in industrialized

countries. Consequently, Kasongo women will per­
ceive it to be a RF to a lesser extent. The compliance
of ‘nulliparous or primiparous under 150 cm’ as a RF
will depend above all on the persuasiveness of the
health personnel in charge of the ANCs. If one
accepts this hypothesis, the five risk factors of difficult
labor may be assigned to three categories, depending
on the intensity of the risk as it is perceived by the
parturient herself. In the first category, the woman
herself feels to be in danger (previous difficult delivery
and transverse lie). In the second category, the
woman is identified as being at risk by the health
personnel (nulliparous or primiparous woman under
150 cm). And in the third category, the woman’s
previous delivery did not go smoothly, although her
own life was never endangered (in fact, her last child
was born dead). The results are summarized in
Table 4.
The results show that the compliance rate ranges
from 24.8% to 45.5%, depending on the patient’s
perception of the risk. This difference is significant
(P < 0.05). Nevertheless, it should be pointed out
that the two criteria of risk perception ‘woman who
identifies herself as being at risk’ and ‘woman who
has lost a neonate’ may vary greatly depending from
the prevailing sociocultural context.
Geographical accessibility is the fifth hypothesis we
considered, i.e. “geographical accessibility is a major
constraint”. When we compare the compliance rates
for the urban HCs located in Kasongo township with
those of the rural HCs (see Table 5), they are 55.7%
(54/97) and 21.1% (45/213), respectively. For 14
women (4.3%), the facility where they attended ANC
is unknown.
As could be expected, the difference is highly
significant (P < 0.001). However, when ‘rural’ and

Table 5. Referral compliance: urban vs rural HCs
Referral compliance

Urban health
centers
(%)

Rural health
centers
(%)

Significance
level
urban/rural

Nulliparous under 150 cm

23/35
(65.7%)

10/54
(18.5%)

P <0.001

Primiparous under 150 cm

10/19
(52.6%)

4/34
(11.8%)

P <0.01

9/15(60.0%)

6/16(37.5%)

NS

Previous difficult delivery

10/21
(47.6%)

5/14
(35.7%)

NS

Previous stillbirth

13/30
(43.3%)

21/107
(19.6%)

P <0.01

NS

NS

65/120
(54.2%)

46/225
(20.4%)

Risk factor

Transverse lie

Significance level
(between risk factors)
All criteria together:

Total number of women*:

P <0.001

P <0.001
45/213
(21.1%)
•Women may present more than one risk factor: for 14 women nlace of ANC
is unknown.
54/97
(55.7%)

Strategy of risk approach in antenatal care

‘urban’ rates are broken down for each of the five
RFs for difficult labor, the difference of compliance
between the different RFs is no longer significant. On
the other hand, it is interesting to note that the
compliance rates in the rural population for two of
the RFs (‘antecedent of a difficult delivery’ and
‘transverse lie’, respectively), are not significantly
lower than the same rates in the urban population.
On the contrary, there is a significant difference for
the three other RFs considered. These findings
suggest that the patient’s proper perception of the
risk is indeed an important determinant. However,
accessibility may be a confounding factor for the
association between parturient’s perception and com­
pliance. A stratified analysis (Mantel-Haenszel y2)
was done in order to identify a possible interaction.
The three categories of Table 4 have been plotted in
two groups: the category of ‘woman feels herself to
be in danger’ on the one hand, and the two other
categories together on the other hand. For the rural
Vs the compliance rate is 39.3% (11/28) when the
woman herself feels to be endangered, and 18.5%
(34/184) for the other cases. The relative risk is: 2.1
(1.2-3.7). For the urban HCs, the compliance rate is
52.8% (19/36) for the first group and 57.9% (44/76)
for the second one. The relative risk is 0.9 (0.6-1.3).
This stratified analysis shows that the parturient’s
perception is influencing the compliance in the rural
areas but not so in the urban areas. This result is
consistent with the fact that hospital utilization is a
much more difficult process for patients living in
remote rural areas.
Two hypotheses remain still to be tested. The sixth
hypothesis states that the regularity of pregnancy
monitoring influences the compliance rate, and the
seventh that the performance of the staff delivering
the antenatal care plays a significant role in the
compliance with the referral decision.
<s mentioned above, two visits are scheduled for
a normal pregnancy. The referrals were divided into
two groups: women who attended both visits and
those who only attended the first one. As the diagno­
sis of ‘transverse lie’ is made only at the end of the
pregnancy, the 31 parturients presenting this RF were
excluded from this analysis. Among the 297 remain­
ing cases (4 women with transverse lie had also
another RF), the compliance rates were 37.0% (67/
181) and 22.4% (26/116) for ‘regular’ and ‘irregular’
women respectively. This sixth hypothesis proves
valid, for the difference is significant (P < 0.01). This
suggests that a more regular contact with the health
service increases the compliance of a decision that has
already been made at the first visit. Obviously, a
selection bias cannot be ruled out. Indeed, it may very
well be that women who are ‘regular’ attenders of the
ANC are also women more concerned with their
health status, so that they would anyhow comply to
a larger extent with the decision taken by the nurse.
In order to verify the seventh and last hypothesis,
we compared the following qualitative criteria for the

533

best and the worst performing rural HCs (no differ­
ence was identified in terms of quality of care for the
urban centers): quality of,the relationship between
the staff responsible for the ANC and the population,
time spent on the job, quality of preventive and
curative care, quality of interaction with the patients,
knowledge of the vernacular, cleanliness of the
premises, community involvement in educational
meetings and finally the HCs management perform­
ance. The results show that referral rate is signifi­
cantly lower-for the ‘best rural- HC:~5.6% instead of
11.2% for the ‘poorly’ performing center (P < 0.01).
On the other hand the compliance rate is higher for
the best rural HC (21.7% instead of 9.1), but
the difference is not significative (P = 0.11). The
data therefore do not allow us to confirm this last
hypothesis.
DISCUSSION

A number of features in the Kasongo district
health system are expected to enhance the compliance
with the referral decision of patients at risk of difficult
labor. These are:

—a strong emphasis on community involvement
in the management of the health services,
—a complementarity between the two levels of
care within the district health system, fostered
by regular supervision, by a functioning refer­
ral and counter-referral system,
—an identification and quantification of RFs of
difficult labor based on local epidemiological
data [5],
—a fee schedule favoring continuity of care: each
pregnant woman registered for antenatal care
pays a lump sum covering all medical costs of
hospital admission and obstetric surgery (ce­
sarian section, etc.) when these are necessary.

Despite these favorable conditions, the compliance
rate is poor: 33.3% for all pregnancies (urban and
rural communities together) and not more than
66.6% in the best of cases in urban settings. It should
be pointed out that the real compliance rate is
probably higher. Indeed, as it is difficult for the HC
nurse to assess exactly the current age of the preg­
nancy, some women at risk who would have complied
with the decision taken and who would have given
birth in the hospital delivered at home a few weeks
before the ‘expected’ delivery rate. As a matter of
fact, 15% of all deliveries occurred more than two
weeks before the expected due date. If we assume that
15% of the referred women were unable to reach the
hospital for that particular reason, the new compli­
ance rate would be 39.3% (108/275).
Among the seven hypotheses that were tested,
geographic accessibility (urban vs rural HCs) appears
to have the greatest impact on the compliance rates.
The intensity of the patient’s perception of the risk
also seems an important determinant. The compli­

534

B. Dujardin et al.

ance rate is higher in case of antecedents of difficult
delivery and transverse lie (i.e. when women may fear
for their own life), than when other risk factors (i.e.
antecedents of stillbirths or women of short stature)
not picturing any direct threat to the parturients’ lives
are present. A regular attendance of antenatal clinics
enhances the probability of compliance, as does a low
referral rate. Nevertheless, these two hypotheses do
not explain the RF related variations in compliance
rates.
We have not found any comparable studies of the
compliance of RFs identified at ANCs in the litera­
ture. A study conducted in Kenya in 1975-1979,
however allows for some interesting comparisons [6],
In this study, 4716 pregnant women were asked
whether they intended to deliver at the hospital or at
the local village maternity, and were asked as well
what the determinants of their choice would be. After
the delivery, the same women were seen again and
asked where they actually delivered and what the
outcome of the delivery was. Amongst the 87%
(4103/4716) women who did express a choice at the
time of the first interview, 54% (2216/4103) wished to
deliver in a hospital environment. In fact, only 36%
(1477/4103) of them did so.
The findings of this survey allow for a comparison
between the ‘subjective’ degree of compliance (i.e.
what the woman planned to do) and the ‘objective’
degree (i.e. what she actually did). The following
results are of interest to the present discussion. The
factors significantly related to the desire to deliver in
a hospital setting were antecedents of a cesarian
section, previous use of vacuum extraction or forceps,
previous delivery in a hospital, age under 30 and
primigravidity. The factors related to the desire to
deliver at home were distance to the hospital, multi­
parity (>5 deliveries) and women of short stature
(<150cm). All these factors relate to the subjective
compliance. For each risk factor, the ‘objective’
compliance is much lower than the ‘subjective’ one (a
decrease rate of 35-50%), with two exceptions: a
decrease rate of only 15% for women with an­
tecedents of a cesarian section, and even an increase
rate of 20% for women with small stature. For all the
factors considered, there is an association between
the presence of a RF and the intention to deliver in
a hospital (‘subjective’ compliance). This association
is even stronger for the ‘objective’ compliance, except
for the RF ‘antecedents of a perinatal death’. The
findings of this study corroborate our own findings.
Indeed, like in Kasongo, antecedents of perinatal
death do not seem to be perceived sufficiently strong
to overrule the cost of a referral to the hospital. On
the other hand, it is interesting to note that the risk
factor ‘short stature’, which did not appear to motiv­
ate women to deliver in the hospital (i.e. the ‘subjec­
tive’ compliance), is in fact a very real RF since many
women who initially intended to deliver at home were
‘forced’ to deliver in the hospital (i.e. the ‘objective’
compliance).

The overall results of this descriptive study show
the importance of assessing compliance w.th the
referral decision, when promoting a nsk approach
strategy. Some hypotheses were tested to explain the
poor compliance results. However, such a broad
assessment is too general to identify the real expla­
nations for a lack of compliance at the individual
level. A more qualitative approach (open inter­
views, focus group discussions) are needed [6, 7] and
should be the next step in the process analysis of
the-.problems related to compliance with referral
decision.
CONCLUSIONS

The main result of this research on the compliance
with referral decisions is that it shows that, despite
favorable conditions, the overall compliance rate of
pregnant women at risk for difficult labor is poor:
33.3%, or only one out of three women complies with
the decision taken. Both accessibility and parturient s
risk perception influence this compliance. It is also
important to underline the fact that the medical staff
was unaware of this low compliance rate until this
research was carried out. This unawareness is
doubtlessly due to the fact that, as the overall referral
rate (i.e. the proportion of women attending ante­
natal clinics identified at risk) is rather low (6.5%),
the number of referred pregnancies remains quite
small. Such a phenomenon does not become apparent
until the data from different HC are pooled.
Maternal health problems in developing countries
were recently recognized as a priority by the World
Health Organization, the United Nations Fund for
Population Activities and UNICEF. Improving the
coverage of antenatal care (both qualitatively and
quantitatively), as well as appropriate hospital care
for high-risk pregnancies (obstetric surgery, etc.) are
two of the main activities that, if carried out together,
are expected to reduce maternal morbidity and mor­
tality. Unfortunately, such activities will have little
impact if the majority of the women in whom risk
factors of difficult labor are detected do not deliver in
the hospital.
This article was an attempt to highlight the import­
ance of this problem and to identify some of its
determinants. The interpretation and discussion of
the data were done with a specific context in mind.
We suggest that similar research be conducted in
other contexts in order to assess the validity of our
findings and to identify ways to improve the individ­
ual’s compliance with referral decision.
REFERENCES

Backett E. M„ Davies A. M. and Petros-Barvasian A.
Risk approach in health care with special reference to
won"?1 ua"d .Child health including family planning.
WHO Public health papers, No 76, 1984.
2. Omran A., Martin J. and Hamza b’. High risk mothers
and newborns, detection management and prevention.

1.

Strategy of risk approach in antenatal care

Proceedings: Second International Congress for
Maternal and Neonatal Health, 1984. Monastir Tunisia.
lamaneh-seminar Berlin/FRG 1985 and Asia Pacific
Echo Conference Manila Philippines 1986. Ott Publish­
ers, Thun (Switzerland) pp 1-488. 1987.
3. James D. K. and Stirrat G. M. (Eds) Pregnancy and
Risk: the Basis for Rational Management. John Wiley &
Sons Ltd, Chichester, 1988.
4. Wallace H. M. and Giri K. (Eds). Health Care of
Women and Children in Developing Countries. Third
Party Publishing Co. Oakland, CA, 1990.

535

The Kasongo Project Team: Antenatal screening
for
fetopelvic
dystocias.
A
cost-effectiveness
approach to the choice of simple indicators for use
by auxiliary personnel. J. Trap. Med. Hvg. 87, 173.
1984.
6. Van Ginneken and Muller (Eds) Maternal and Health in
Rural Kenya: an Epidemiological Study. Croom Helm,
London, 1984.
7. Janz N. K. and Becker M. H. The health belief model:
A decade later. Hlth Educ. Q. 11, 1. 1984.
5.
Media
SDA-RF-CH-1.32.pdf

Position: 6099 (1 views)