PLANNING REPORT FOR AN EPIDEMIOLOGICAL STUDY OF AGRICULTURAL WORKERS EXPOSED TO NOISE

Item

Title
PLANNING REPORT FOR AN
EPIDEMIOLOGICAL STUDY OF AGRICULTURAL
WORKERS EXPOSED TO NOISE
extracted text
r«ahs4sj^jssr-tf»*».-j»w

^=4i:,jw**•«.TnQfwawiJaffisnaKmc*-

‘ , r’>

r:

?

-•'“' ^*•

•j.'v-'' * •’ £ •.

U^*C***A-

r?i<
MM/-»..M ■ .--.x-M:’•- -..K..■■v.’K

M " .

’

j"-4 •••

■.:-

•

O| t

1(

':'■

■ '

■

.

■

•

-M M

■

.

■

..

/.

■

'.

•■

' '‘-

-

■;

'

■‘'A,MAA- a MAAS'^aAAAgaaaaM

.;J..

A a.MA- A : a. .-. a'. , . ..a M ...g. .

■

'■Mv

•’J-.

K;-AAAAM-.A'AMa.AK®A.
AAAAM; <MAAA;.?;; '‘"-A ...
AMB,;AAiAM?AAMMGA
AAAAM^AASMffaamaa^..,.. -.......

■G.KG'A;'AA-'.;' - vfcA a
■

• , .

-

'.fA^ASS

i\

’-' :. A -A. AgA;..,„
.WMaMk-MMa mm®

■ryf,'^/--'.

'T:.:^*■* * n n/| j j
■'■

;

■'■.

■

■'■'

;.

■

:

'

■■

'■< ■

:-.

,

.

|
.

;

aKWKW^.KWfiW!~.
L;s;fe— ■: <: —f' '■

....

MM

1T 'S
Y.

-'-KI

AMaa a; a

■■■

-

5.7'; -- mm *. ,- -G®-- 7^ ”f

-5 -,- ....

'

'

-AG. 'M‘G

.

. -

' ■- -

.'•.>-:A';' ■ :

-■

-'K-

<'G ■’■■■• -:

rn-

-w' ■"'A.; g ^^^^^ftB|KM'KA'AAiG3AAABAABM-®A
••'A< .A GG;G:;'■• ■■■

■•'.

4

t

,A

G ;; A- -GJ. A '-;A;<;4G A..;:?;A<

<®;ae»5s:'V©w«s-gK*jawK®s^

— •. <> '■ ■ i —,:Y... ■->-

W»war
tevV;.,:. "—.—
V.?".

B.-— -.

PLANNING REPORT FOR AN
EPIDEMIOLOGICAL STUDY OF
WORKERS EXPOSED TO NOISE

■

r^BglgOg^gg^

MM KK.A<V ^culwkal

mb
®
p
B
|W?\; .■ —G’. .’.;-' A
MfesMTO®

tati

-M M

’

'

•■

'

-

*«

•

.
:■

SGG—'

-

. ’•

. -

•

■ •

TR 176-002

26 June 1975

-m -• ,'M ;m®

•■

7

,

8 3-23A98 9

<

:

;
•
:^-!M
^■■'^
rs'B ■ ■. .•
M4
■

■.:.■

■■.-

<

’

-.

■-.

■

' .......
;-Y alWiWIOO
■

M-

'

•

-.:

.■/;'

■

:■■■■

■■

■

■■■

iftfiiiiiiiasww
•

-

-

.."

G

'
•
M«A ■■<•;, ;? . M:F
*>v? J
’■:' ■-s:
M'.r--. .-A- ' M
A-| ..v--;-;;''
gg»p|£Og|S iMhoww
•■'

wi:
-

■

'

.

■

•

•

.■

.

A -; v U-aVgA''---'
'■■ ;W' .:U>
ALA .;——A ''.
AMA'- ws®
-A''i& <A- -: •■■
YYfBr■AMKMA AA■"■ ''A M'aA.A .AAA; ..' A-, A A - GA
KAWMAA/AA-A.G.A -AA-f ;' ;;.- AAA aAA< A-.;..

-AAA-A

i! • ■•' ■-■

G A;

;H?;;

',.?! • y it*.

-■•-•.> ■:■ a‘a-M’KKGV^'MA',®.K -«.-■/
-.'■- •
- • -'•• v ■' /

V®r^- -j®?^ ’;M!m . •“’>■'
i'A'K•.;. /./GM; ■•' B

■/•;../-•-.-Gj,mb?", 'a ■ ■■■'B'’*

*

®

g:g

,c.::>.''

■■:

A'BAaaa,..ga MaaammakaMa-K -MAaaAAa^-®;.
M AA- . -mKGmAMKM LAMGmAA ;A”MKMA"AAaK^
S.«GaplsWe?»;
MKa-M^v-.---• m'MM M >.<C --AA- . " .
A..® ’'’/ •' ■ . /
■

1;'

W-. - -CY . '. Y
.-,.
‘ ’ '.gs* .< < / ■ ■
"' . ", »
■ / •' —
A"-< sSJ
— ’SftSKs® ■ ■'
■ ' a-Kaac
Bm: >■ B<a '.
. .. A.

■•■•

': '■-•
' ,M .

■

':

m:

' " <•.- ‘.'-T •. * r " ?>' f>* -j.

y-.k:'

'•'<->. .•-?. -' a •'' -.

-•‘Mg’-?--'Ag — A'M-g...

< ”• ■'' yk;'K^'WK'

;> <? Y^KKi?■■>:—/■""

ma<<• ■ t.a>Ma.,.■■

; ;■—■■.

:;. >’K‘.p ■

l. wrwwi

Illi I

lll■lll

This report was prepared under
Contract CDC-99-74-110
for the
National Institute for
Occupational Safety and Health

g

a
0
1
3
3

3'

a
n

by the
Utah Biomedical Test Laboratory

4

INTRODUCTION

This report was prepared and submitted in response to the contract
requirement to develop a detailed research protocol for an epidemiological

1.

study of farm workers exposed to noise, The technical content of the portions of this report entitled Statement of the Problem and Appendix A; More

Specific Recommendations for Research Procedure is based upon a paper pre­
pared by consultants to the contractor.

The consultants were Aram Glorig,

M.D. and W. Dixon Ward, Ph.D.
The report presents a Statement of the Problem, including current
understanding of epidemiological factors, relevant work, and needed research.

Study Plan; and Resource Requirements.

0
]

•)

1

STATEMENT OF THE PROBLEM

The most important hazard posed by noise is the danger of producing
damage to the hair cells of the inner ear, thereby causing a partial loss

of hearing.

In the presence of noise above about 70 dBA, it is true that

the autonomic nervous system does produce certain measurable changes:
dilation of the pupils of the eyes, changes in cardiac pattern, and vasoconstriction of the extremeties.

But these effects, produced by many

other changes in the environment, are temporary and reversible, according
to present knowledge. While there is suggestive evidence1 that steel
workers exposed to noise so severe as to produce hearing loss relatively
rapidly may also display an incidence of circulatory problems that is sig­
nificantly higher than in a control group of non-noise-exposed workers,

the noise workers concerned were also exposed to higher temperatures and a
greater concentration of fumes from the steel-making process.

The differ­

ence in circulatory problems is at the moment considered to be due primar­
ily to the stress induced by fumes and temperature; although there remains
the possibility of a synergistic interaction among the various stressors,

1

this is still speculative.

In the absence of conclusive data, such inter­

active influences shall be ignored in this document.

A similar conclusion

applies’ to the combination of noise and vibration, especially in operators

]

of tractors, which is the mo$t ubiquitous source of potentially hazardous
noise to American farmers:

although some speculation exists that vibration

may enhance the effects of a given noise exposure, the evidence is both

meager and inconclusive.

1

Noise may also occasionally be a contributing factor in accidents due
to failure to hear warning signals of one sort or another, In the agricul1

tural realm, however, such situations will only seldom arise, so the effect
of noise on the communication process may probably be safely ignored.

Finally, although the popular press and some anti-noise extremists have
4

contended that high-intensity noise may impair task performance, possibly
increasing the probability of error in judgment on the part of tractor oper­

]

ators, again the evidence is at best equivocal, despite dozens of experiments
designed to show such effects, if they existed.

By the same token, speculation

3

2
|

linking noise with-such things as insomnia, mental illness, sexual potency,
and the like are not supported by any firm evidence.
Thus, the primary, if not the only, result of noise on the farmer is
its effect on hearing.

Present evidence indicates that steady exposure

8 hr/day, 5 days/wk, 50 wk/year to industrial noise whose level is 85 dBA

and which has an ’’average” spectrum (not unlike that of a tractor) will just

produce, after 10 or more years, a measurable hearing loss -- i.e., a 10-dB
hearing loss at the frequency most sensitive to damage by noise, 4000 Hz.

If the level is 90 dBA, this loss will be about 20 dB.

'3

Inasmuch as noise

levels generated by tractors, combines, chain saws, grinding mills, elevators,

VJ

and most other powered farm implements exceed 90 dBA very often, it is clear

that a potential hazard exists.

On the other hand, the exposure of the typi­

cal farmer is anything but regular.

In the spring he may spend 16 or even

18 hours a day on his tractor, but in the winter his weekly exposure to levels
of 90 dBA or above may total less than an hour.

The eventual effects of such

variable exposure have to date not been adequately determined, but there are
two extreme points of view:

’’critical incident” theories.

the ’’integration of microtraumata” and the
According to the microtrauma theory, every bit

of acoustic energy entering the ear contributes its share to the gradual

breakdown of the sensory elements, a situation analogous to that generally

1

assumed to hold for radiation damage.

The critical-incident hypothesis, on

the other hand, contends that there is a tolerable daily acoustic dose, per­

haps peculiar to the ear concerned, below which no lasting effect whatever is

produced, so that the hearing losses that accrue over a long period of time

J

can be attributed to a few separate days in which an unusually severe exposure

was experienced.

Although the Environmental Protection Agency in its so-called

Levels Document2 has accepted the microtrauma theory, no persuasive evidence
for its validity has yet been uncovered.
It can be seen that the degree of ’’hazard”, particularly in terms of the
1

■ ?

number of farmers judged ’’exposed”, depends critically on which of these
hypotheses the truth is nearer. If the total-imniission (microtrauma) theory
is correct, then everyone nearly everywhere -- and of course each farmer --

is ”at risk”.

However, if (as seems more likely), the critical-incident

schema is the more accurate, then hazard must be estimated from empirical

hearing-loss data obtained on farmers directly.
I
3

1

r w—r-i

-e

The overall deterioration of auditory sensitivity with time can be

broken into four more or less distinct components:
(1) The loss due to occupational noise, which is what we want
(2)

to determine;
Sociacusis: hearing loss that can be ascribed to the noises
to which an individual voluntarily subjects himself, quite

apart from the working situation (gunfire, rock music, power

3:
(3)

tools, etc.);
Presbyacusis:

the hearing loss that is attributed to the

general process of aging as such; and
(4)

Nosoacusis:

the loss in auditory sensitivity caused by
i

diseases, drugs, blows to the head, industrial chemicals, and
the like.

The problem then is:

How much worse is the hearing of the hypothetical

’’typical farmer” than that of someone the same age, who is exposed to the

same amount of gunfire, recreational noise and otological hazards, but to no
occupational noise whatsoever?

It is clear that epidemiological studies of

farm-noise-induced hearing loss must use exceedingly meticulous controls if
a valid result is to be obtained.

.1

Merely showing that some farmers have

severe hearing losses is of course pointless.

Even demonstrating that- farmers

have worse hearing than city dwellers is not sufficient to establish occupa­

tional hazard unless one can also show that their exposure to recreational

noise sources is no greater.

Finally, fair surveys have been shown to be

often misleading, because of both the possibly non-random nature of that

3

group of farmers who are fair-goers and the self-selection processes that

influence the determination of who gets tested.
Unfortunately, no large random sample of farmers has ever been studied;
most extant data are either based on haphazard selection processes or involve

J

such a small number of farmers that once one breaks the sample into groups
with equal years of exposure, comparison with some other non-farm sample is

J '

of dubious meaning. There seems little doubt that some farmers have severe
hearing losses, and that there are occupational farm noises that could pro­

duce such losses.

Beyond these general statements little can currently be

said with confidence.

4

1

J- JUSSUb.

..-M

.....

Current Understanding of Epidemiological Factors

The current understanding of noise induced hearing loss in farmers in
the areas of prevalence, incidence, population at risk, relative rates, and

itiology is summarized in the following.
Hearing loss is not an all-or-none matter; nearly everyone has some hearing loss, in the sense of a deviation from the norm. There1.

Prevalance
Prevalance..

fore an estimate of "prevalence” must be based on some arbitrary degree of

One commonly used criterion is an average HL (Hearing Level -- dB of

loss.

loss relative to an internationally-approved standard ’’normal”) at 500, 1000

and 2000 Hz of 25 dB.

This is the point at which handicap is judged to begin,

according to the rule approved by the American Medical Association and hence

widely used in this country.

While the relative incidence of handicapping

loss, in a random sample of the population of the USA, broken down by age
decade and sex, has been determined by the Public Health Service3, no attempt
was made in that survey to determine occupation, so its incidence in farmers
is unknown. Even in the Fair surveys1**5,6,7 of hearing of farmers, categori­

zation in terms of this handicap was not attempted.
2.

]

Incidence.

Unknown, for essentially the same reasons as- the above.

Population at Risk. Probably everyone who works long hours with a
tractor; some unknown fraction of the total farm population of about 3 million
3.

male adults.

4.

Relative Rates*

Basically unknown.

Of the few references that bear

on the problem, three^’6,7 give audiometric data in a form that at least

I
1

allows comparison of

means HLs with those of the random sample of the USPHS

They all agree that after 10 to 30 years of work, farmers have 10 to
20 dB more hearing loss in the 4000-Hz range than the average USA male, This

survey.

substantiates the conclusion that if the samples in the three surveys are

representative (which, however, must not be taken for granted), farmers will
have a higher incidence "than "the general population.

OF course, nobody argues

that the general population is really the appropriate group with which to com­

pare the agricultural group.

1

All persons who have worked in noise over 80 dBA

should have been removed from the PHS survey if one is to deduce the effects
of noise on any particular noise-exposed group.

If this removal were possible,

the implied loss due to farm noise would doubtless be somewhat greater, but

’j

just how much greater is not known.

5
•3

JL. JPKSt.W.

5.

Severity... As indicated above, an average HL of 25 dB at 500, 1000

and 2000 Hz is taken as the dividing line between "no loss" and "some loss".

Complete handicap is assumed to occur when the index reaches 92 dB, with a
linear rate of increase from the "low fence" to "full loss" of 1^% per dB.

There is some agitation at this time to lower these limits and/or to use
1000, 2000 and 3000 Hz instead of 500, 1000 and 2000 Hz, but such changes
will probably not be effected for several years.

6.

Etiology.

The internal combustion engine (tractors, chain saws,

loaders, grinders, etc.) is the main source of noise on the farm.

However,

one must remember that farmers as a group are also exposed to more gunfire

1

than the average citizen; although gunfire-induced losses are not strictly

"occupational", they will of course result in a greater proportion of farmers
exceeding the low fence at any particular age.

Levels generated at the ear

of the operator of many farm implements and tractors are given in reference 8

and in a presentation by W. F. Splinter before the Environmental Protection

Agency Office of Noise Abatement and Control Hearings in Denver, Colorado, on

October 1, 1971.
Relevant Work

1

Bozung5 conducted a study of farmers in 1973 in cooperation with the
Department of Audiology and Speech Science of Michigan State University,

Of

farmers attending Farmers’ Week at Michigan State University, volunteers

(inducement not stated) were screened against non-farm-related loud-noise
exposure (including service in the armed forces), gunfire other than during
hunting season, excessive ("more than 2 hours, often") snowmobiling, and

tractor noise exposure more than 2 hr. in duration on the preceding day.
Air and bone-conduction audiograms were administered to those passing the
screening, and individuals with more than a 5-dB air-bone gap were discarded.

J •

This left 80 farmers aged 26-65 years with 10 to 39 years of full-time farm­
ing.

An attempt was made to analyze results for different types of farming

J

(dairy, swine, poultry; area in field crops), but with such small numbers,
this was fruitless. When broken down by age groups, HLs are clearly greater

i

than the presbycusis norms advocated by NIOSH g at frequencies above 1000 Hz:
about 10 dB worse (relatively independent of age) at 2000 Hz, and 25 dB worse

7

J
1

at 3000, 4000 and 6000 Hz. About 17 dB worse at 3000, 4000 and 6000 Hz than
the PHS medians, Thus hazard appears clearly established; only randomness of
the sample is the question.

6

----TOjiHrasraWL^Ji--

JBUUl'JUBSBaJlBSt. — JS. - -J-3LU

II I— ■» a*

Dennis6 reported in 1969 the results of a study where 484 farmers were

interviewed and examined at 8 agricultural farms and 2 "mexabitions" (agri­
cultural exhibitions) in Saskatchewan.

Subjects were rejected if:

exposed

to tractors or combines within 18 hours of test (33); over 65 or under 25 in

age (19); prior known hearing loss from early childhood (5); audiometric

error (10); ear operation (4); or some combination of the foregoing (2).
remaining audiograms (412) were analyzed in terms of:

The-

(A) Area of land

farmed (tendency for more loss in farmers with more land, but no statistical

test for significance; difference less than 5 dB in the aggregate, it appears);
(B) shooting habits (no effect); (C) military service (no effect). Breakdown

by age shows losses 10 to 15 dB worse than PHS medians at 3000, 4000 and 6000

1

Hz (no comparisons with controls were made by the author).
known, as is the method of inducement to participate.

is the lack of statistical analysis.

Sampling is un­

j

The only other weakness

However, results must be judged incon­

clusive or negative at best.
A study reported by Willsey7 in 1972 involved fifty-one farmers from
around Lafayette (Indiana) who participated in a special hearing study (method
of selection not mentioned) performed at Purdue University Hearing Clinic.

Breakdown by age (N in each decade not indicated) shows HLs indistinguishable

from the PHS medians at all frequencies except at 4000 Hz.

1

At 4000 Hz, about

10 dB more loss in the 30-39 and 40-49 age groups, but not in the 20-29, 50-59

and 60-69 groups.

This study therefore implies very little hazard from occu­

pational farm noises -- no more than exists in the average American occupation.

It should be clear at this point that the extant audiometric data,

though in some cases completely reliable and valid, do not unequivocally prove

J
1

anything except that some farmers have hearing losses, which was really known
all along.

Although, because farm noises are often greater than 90 dBA, these

hearing losses may be attributed to farm noises, the assumption of such a

causal relation is gratuitous.

J
T-

J
]
3

Needed Research
Answers to the questions as listed in the following should be pursued.

1. Do the noises of the modern farm produce a greater average hearing
loss in farmers than in the general population? By no means certain.

7

1

4

2.

If so, is this loss greater than in the average of those citizens

who have not been exposed to supra-90-dBA noises for extended periods of

The answer is probably affirmative, from the evidence in reference 5;

time?

it is difficult to imagine that self-selection processes alone could account
for the 25-dB implied difference at 3000, 4000 and 6000 Hz between the farm­

ers that were actually tested and all Michigan farmers.

However, the true

extent of the increased risk can be determined only if seif-seleciton is

eliminated.
3.

Assuming a positive answer to this question, the next one is:

what is the typical noise dose of a farmer?

Although the noise

levels generated by various farm machines are being measured at the Univer­

sity of Nebraska and elsewhere, not much is known about the typical exposure

3

3

of the average farmer (some sort of integral over time of the acoustic power

entering the ear) in daily, weekly, seasonal or yearly terms.

One of the

reasons for this is that there is at present considerable disagreement over

how the noise dose should be measured — whether to use a system whereby a
reduction of exposure time by one-half will allow an increase in level of 5

dBA, as the present OSHA regulation indicates, or of only 3 dBA, a relation
espoused by the Office of Noise Abatement and Control of the Environmental

Protection Agency; whether to include exposure to noise levels between 80 and

3
J

90 dBA and above 115 dBA in calculating the dose; how the effect of inter­

polated rest periods is to be treated; how steady noise and impulse noise act
jointly; and so on.

Until the definition of "noise dose" is settled (if ever),

however, it would still be possible to record (on tape for later multiple

analyses) sound-level histories over a few days of a group of randomly-selected

farmers; this would at least provide an indication of the range of exposures,
in a general sense.

■1

4.

Do hearing losses really develop gradually, or are particularly

severe single exposures responsible?

A longitudinal study for 3 to 5 years

jr«

of a group of randomly-selected farmers, involving audiograms every two weeks,

i
i—

might be able to provide an answer to this question (which, of course may well
be ’’both”).

Special attention should be paid to the effects of 14- to 18-hr.

exposures to tractor or machinery noises, because in such concentrated work

periods not only is the ear being exposed for unusually long times, but the
normal recovery period is also being shortened, so that the farmer begins the
next day’s plowing with a still-fatigued ear.

8

A A.

■ Wfirtr

4

As a summarization of the foregoing itemized questions there clearly

exists the need to determine the extent of hearing damage in farmers and the
causal relation between such losses and the cumulative noise dose (or

3-

critical-incident history). Someone must be willing to undertake the indi­
cated audiometric studies of an exhaustive or a random population of farmers,
both cross-sectional and longitudinal.

The brief outline of the beginnings

of such a study is included as Appendix A.

Studies of the type described

and needed require the long-term commitment (from three to six years) of
personnel, equipment, and funding.

0
1

These might best be carried out through

agreements with local, organizations who are concerned with the problem and

are conducting studies of only limited scope in terms of both resources and

geographical area.

The role of NIOSH would then be one of financial support

and coordination of study protocols, and data analysis and interpretation on

3

a national scale.

1

is proposed in the remainder of this report.

A much less ambitious study plan than that implicit in the foregoing

It has several recognized

shortcomings which may cause the principal results to provide nothing more
than the motivation for a commitment to the all-encompassing program suggested

above.

On the other extreme if the remedies suggested for the shortcomings

are successful the requirements in terms of needed research in this area will

1

be fulfilled to a satisfactory degree and no further investigations need be

carried out.

I

I
1

-J

9

-'aW >..BWiPra

sa

. ."KB. ^JBBMK33SSa>. —

4

STUDY PLAN

1

Approach

3;

getting the agricultural worker to the audiometry test equipment, or con-

Logistically, the most difficult aspect of the proposed study lies in
verseley, the audiometric test equipment to the farmer.

Any plan which in­

volves taking test equipment to the work place will require commitment of
equipment and personnel over an extended period of time.

This results from

the low density of farmers in areas suitable for the study and the require­
ment for valid results that do not reflect temporary threshold shifts that

test candidates not be exposed to high levels of noise for at least sixteen

hours prior to the audiometry exam.
J

It seems a conservative estimate that

it will require one week to test twenty-five individuals under these con­
straints. To achieve a sample size of any significance would then require j

a time commitment of from six months to one year and the accompanying in­
vestment of funds

this represents.

NIOSH owned test equipment would be

totally committed during this period of time and/or the cost of a long-term

]

lease arrangement from a private source would have to be borne.
A more logical approach would be to bring the randomly selected agri­
cultural worker to a central equipment location.

Proper scheduling under

these circumstances would enable the relatively rapid accumulation of

results from a large sample size.

For example, a representative* of an

industrial audiometric testing firm estimates that it is conservatively

possible for one of their audiometric test vans to test thirty subjects per
hour.

Assuming a relatively constant supply, this could result in the

gathering of data from up to 250 individuals per day.
The logistics of providing a constant supply of willing subjects at a
central location then becomes the real world problem.

It is proposed that

a workable solution to this problem may exist in the concept of locating
audiometric test vans at state fairs in appropriate states, and conducting
hearing tests on volunteers and/or preselected and screened individuals.
1
1 |

*Mr. Bert Scott, Environmental Tech. Corp., Cleveland, Ohio.

10

n

Several arguements against the gathering of data at state or provincial
fairs have been presented in the proceeding.

These involve questions regard­

ing the validity of data taken from the population sampled because of possible
biases introduced by self-selection processes and inaccuracies in background

] ■

L

information.

Self-selection biases result both from the non-random nature of

the group of farmers who are fair-goers and from individual subject character­

istics that influence the determination of who gets tested.

Inaccuracies in

background information are suspected because of the social environment in

which interviews are conducted at the fair, i.e. there is probably a tendency
to sacrifice accuracy, of responses in order to be done with it.

It can be argued that a self-selection bias will result under even the

1

most favorable conditions in a study of this kind.

Assuming that a valid

random selection of farmers is made in a particular geographical region, it

is not at all certain that all of such a group will consent to cooperate in

1

a hearing study. Some will simply not want to take the time arid trouble in
completing an interview form and submitting to an audiology test, Thus, a

self-selection bias is introduced.

The extent of such a bias may or may not

be as great as that which results in the less motivated situation which

exists on the fair grounds.

However, it seems likely that sufficient incen­

tive could be provided in the latter case so as to cause the bias to be more

.1

equal in the two situations.
The study plan presented here is comprised of three phases.

'1

and extent of phases II and III will depend heavily upon the outcomes of

J

]
1
J

phase I, and phases I and II respectively.

The nature

It is intended that these decision

points be provided so that an all out commitment be avoided of the consider­
able time and funds implied by a truly random selection and on-site testing

The outcome of this study will, at a
minimum, provide a test of the hypothesis that farmers as a group have a

of a representative group of farmers.

greater degree of sensory-neural hearing loss than all other groups taken as

•)

a whole.
1. Phase
Phase I.
I. Phase I of the study consists of the random selection and
survey of a group of farmers within a one to several county area surrounding

the site of the state fair in each of four determined states.
■?

j

states are Nebraska, Iowa, Illinois, and Kansas.

These four

The number of counties for

which this is to be accomplished in each case will depend upon how many will
be required to bring the size of the random sample to approximately 1,000
i

01321
OH 100

n

» «U-W-MiiWr>.

. SWJVl'WSUFWL-

>El

individuals.

-^Jtw’™a»*!j',w’~>^M^’l****M~:,P,^i,rc^r«ini'W(ii'iWvninrr~i-iTi -i wtm ff in—wn-ii- < mi- ni«vjrww«rr.-—~.'y.. -

Assuming a retention/participation rate of 25% this will result

in a sample size of 1,000 individuals over the four state area.

The random sample to be surveyed could be identified from a variety

of sources. The best is probably the state office of Agricultural Stabilization and Conservation Service. This is a USDA office maintained in each
state. Records of the identity of farmers throughout the state are kept via
branch offices in each county.
The random sample once identified is to then be surveyed relative to
members willingness to cooperate in the study and the possible existence of

1
J

any disqualifying factors that can be determined directly over the telephone.

These may include such elements as hereditary deafness or service in the armed
forces.

Prior to initial contacts, however, extensive efforts will be made

to publicize the study via appropriate groups to which the farmers relate and

attempts will be made to gain the support of these organizations on at least
a conceptual level.

The organizations to be contacted will include the pre­

viously mentioned county offices of the Agricultural Stabilization and Conser­

vation Service, the local 4-H Club chapters, the county extension agents, and
the farm bureaus and federations.
It is proposed that the surveys in each state be conducted through

the use of college students.

Preferably these students will be drawn from

the agricultural colleges.

In two of the states (Kansas and Illinois) this

may not be feasible as the agricultural schools are approximately one hundred

1
J

miles from the site of the state fairs.

In these two cases, if appropriate

agricultural students are nol: identifiable, efforts will be made to find suit­
able college age students in the local area.

J

1
J

Once initial telephone contacts have been made and a group of suit­
able individuals who are willing to participate in the study have been iden­

tified a personal contact and interview will be conducted for each.

The

interview will attempt to develop information relative to age, work history

I
j

noise exposure, recreational noise exposure (snow mobiling, hunting), and
other possible sources of noise exposure (e.g. chain saws, etc.). It is

j

anticipated that five individuals should be able to accomplish the complete

?

The

contact and interview process in each state for 1,000 initial individuals.

initial contact should be accomplishable in one month and the follow-up

interview process should take no longer than two months.

, :i

12

-JtBEk

IT*.

4r

With interview data available, a final study group of 250 individuals

is to be selected with some effort made to adequately represent each decade

from 20 to 60 years of age, with an additional single group to include all

those over 60 years.

Further, if some grouping according to exposure levels

0.

to recreational noise sources can be accomplished, this should be done as well.

r

local state fair so that an audiology examination can be accomplished.

5

The final group of study individuals will be asked to attend the
appropriate inducement for attendance will be provided.

form of a free admission for the study individual

An

This could take the

and his family or the pro­

vision of a specific -amount of cash or ’’fair script” to be spent in any way
In any event, the greatest effort reasonable will be

the person pleases.

made to convince the study subjects to attend the fair so that the study can
be concluded in a useful manner.

The number who do agree to attend in each

state will determine the manner in which phase II of the study will be con­
ducted.
!

2.

Phase II.

The objective of this study phase is to conduct audiology

examinations on each of the 250 individuals identified as the study group in
each state, plus as many randomly selected fair goers as possible.

As indi­

cated earlier, it is possible that up to 250 individuals could be tested per

1

day with the resultant potential of 2,500 total for a ten day fair.
The approach proposed will take advantage of the central focal point

naturally provided by the state fair.

]
J

1

It will attempt to reduce the self­

selection bias by independent, random selection of subjects from the fair
crowd and the offering of an appropriate inducement to convince them to have

their hearing tested.

Such an inducement could consist of cash, a particular

amount of fair script, or a donation in their name to a local charity; as,
for example, to the local childrens speech and hearing clinic.

The latter

approach may provide the greatest incentive, especially if the individual

doing the selection is accompanied by a member of 4-H or the Boy Scouts.
In an attempt to further minimize self-selection bias, and any bias

I •
I
i

1

that may be introduced as a result of the test subject providing inaccurate

data, the basic study design is proposed to be that of a case-control study.
In this situation, the case group will be all those individuals tested who

can be demonstrated to have sensory neural hearing loss.

The control group

will consist of all those individuals who have been tested and found to have

13

*

"normal" hearing and can be matched on a one-to-one basis with someone in the
study group.

Assuming this can be accomplished, the question is then asked;

is there a statistically significant difference in the number of farmers in

the case group than in the control group?

The ramifications of a "yes" answer

will depend quite heavily upon the extent to which members of the case and
control group can be matched.
Ideally, of course, the desired approach is to be able to identify

1

a farmer group of representative size (ranging from 20 to 50 depending upon

the sensitivity of the extent of hearing loss it is desired to detect) for
each cell of a three dimensional matrix involving age, annual noise exposure,

and recreational noise exposure.

If such selection and assignment is done

carefully, then the results should provide, as indicated in the Appendix A
text, evidence relative to the growth of hearing loss as a function of ex­
posure to the annual cycle of work related noise (probably tractors) and

1

recreational noise (probably gunfire).
Certainly with the sample sizes suggested, 5,000 to 10,000 indivi­
duals over four states, an adequate number of farmers should be available to
fill each slot of the estimated 100 element matrix with a sufficiently large

group.

The overriding

concern is that the population tested be representa­

tive of the farming population at large.

]
]

.1

It is to resolve this concern that

the 250 study individuals are to be randomly selected for testing in each
state.

These will provide a group for comparison with the farmer group which

results from the selection process on the fair grounds.

If essentially all

of the preselected group short up at the fair for testing, then an excellent
comparison will be possible at all levels.

If a significant percentage of

this group do not follow through and come to the fair for testing, then only

a limited comparison will be possible, but none the less, one that will be

J

useful.

Any differences detected between the study group and the case group

selected on the fair grounds can be used to introduce corrective factors in
data analysis and interpretation.
It is intended that the process of selecting individuals on the
fair grounds as candidates for testing will be accomplished with the students

J

]

employed in the previously described phase I.

It would seem that the most

appropriate inducement to consent to participation would be to provide a cash
contribution in the volunteers name to a relatable charity.

This inducement

would be enforced if a 4-11 person or Boy Scout were to accompany the selecting
14
j

safeifrar^f• •^■-•?». "'--i'~^tu'-•»'^

■“•••'Xi•/^•*?•-

4

<•'

individual.

Once the volunteer is brought to the audiology test site, the

same questionnaire will be administered that is to be used in phase I.

Again,

students employed in phase I can assist in this function.
Actual audiometric testing can be accomplished either utilizing the

IIO'
3
i
0

existing NIOSH capability or leasing from organizations which provide this

service on a commercial basis.

Use of the NIOSH equipment would be most de­

sirable from the points of view of cost and positive control over operator
expertise and consistency and appropriate equipment calibration.

A binaural

test will be administered requiring approximately 10 minutes.
It is suggested that a test also be administered to each volunteer

that measures the impedance of the middle ear.
order of 5 minutes and equipment is minimal.

The time required is on the
The test results provide an

indication of possible middle ear pathology and thus if any hearing loss is

sensory-neural in origin.

This will provide a clinically acceptable alter­

nate to an exam by an otologist and can be administered by a trained tech­
nician.

To insure the validity of results, equipment will be calibrated when

it first arrives on location at the fair grounds and at one week later.

3.

Phase III.

This study phase attempts to address the problem of

measurement of noise impinging on the agricultural worker and hopefully of
providing some degree of correlation with demonstrated existence or lack of

1

hearing loss.

It is proposed that this effort be one of a long-term nature,

utilizing the results of the previously described phases I and II.

It seems certain that the data provided in phases I and II will
J

enable an identification of ’Representative” farmers in each of the recog­

J

nized segments of the agricultural industry (e.g. dairy farm, feedlot opera­

tion, grain farm, confinement housing).

Cooperation of a small group of

these farmers (approximately six) will be obtained for each segment for both
hearing loss and no hearing loss categories.

It would then be a relatively

straightforward matter to instrument the selected individuals with appropri­
ate recording equipment as they proceed through a ’’typical” work day in an

J
■ 1

.

identifiable phase of a yearly work cycle.
A crucial element in this type of endeavor is the valid identifica­
tion of typical phases in the yearly work cycle of the particular agricul­

tural segment under consideration.

This will be accomplished in collabora­

tion with the appropriate staff of the agricultural college in the state of

1
J

•>

15

il ■ inri*'‘i I

■i

0
0
I.
i-

0;

concern and with the cooperating farmers.

Once phases have been validly

identified, then several days can be randomly selected during these periods
for instrumentation of the cooperating farmers.

obtainable over a one year study period.

Useful data should then be

Thus, a one year study involving

the measurement of noise impinging on the farmer is a logical follow on to
the physiological effects study.

Instrumentation is readily available to enable either of two ap­
proaches to actual measurement of noise exposure.

These are the recording

and later spectral analysis of the analog signal in the sound frequency
range of interest or the utilization of wearable noise dosimeters which simply

of total noise exposure (integration of inten­
sity over time) from some established point in time.

provide an on-demand readout

3
3
3

The recording of analog noise signals can be readily achieved using
commercially available personal tape recorders,

Such recorders are routinely

used in a number of medical applications such as continuous EKG monitoring.

In the event commercially available equipment is not suitable as a result of
extreme requirements for wide variation in signal amplitude or resistance to
impact, devices developed for NASA manned and unmanned space programs would

most certainly be appropriate and should be readily available to NIOSH for
this type of study.

"1

J

Several commercial organizations currently market personal dosimeters
for noise exposure*. These devices are relatively inexpensive but provide
only a reading of accumulated exposure referenced to permissible OSHA expo­

sure levels. They can operate over a 24-hour period and provide an output
either directly or via a readout device.

The answer to the question of which approach should be utilized is

1
j

influenced by the manner in which the data are to be interpreted and used.
The analog recording approach would allow the characterization of noise ex­

posure as a function of time over a ’'typical” work day. However, it would be
difficult to correlate this result with a currently existing hearing loss or

lack of hearing loss.

Any correlation would be strictly inferrential without

an extensive time history of noise exposure and hearing degradation.

I

*General Radio, Concord, MA
Model 1944 Noise Dosimeter

Tracor Medical Instruments Division, Austin, TX
Model SPL-104AB Personal Noise Dosimeter
J

I

16

*

Use of the noise dosimeter would provide a measure of total noise
load over the recording period.

This could then be reliably correlated with

the extensive existing data relative to hearing loss and exposure levels in
the industrial setting.

However, given a knowledge only of total noise load

over one day, it would be very difficult to determine which noise sourced pro­

vided the excessive levels.

a-

The best approach would appear to be to utilize

an analog recording system in conjunction with a personal noise dosimeter.*
Ideally, the cooperating farmer would wear only the analog system, and play­

r

back would then be monitorable by the dosimeter for an indication to total

3

were recorded.

1

of time and total noise load over the recording period.

noise load.

A more sophisticated, and faster, approach would be to play the

analog recording back into a computer

at a much faster speed than signals

Appropriate computer analysis and scaling could then be read­

ily accomplished to provide an output of both noise intensity as a function

Data? Analysis and Interpretation

Data will be returned on each test subject in the form of responses to
requested information on the interview form and results of the audiometric
examination on standard audiogram forms.

These latter forms will have a

visual indication of the test subjects hearing threshold at either six or

3

ten pure tone frequencies.

It is anticipated that thresholds will be deter­

mined for both ears at ten frequencies (.25, .5, .75, 1.0, 1.5, 2.0, 3.0,
4.0, 6.0, 8.0 K Hertz).

However, the minimum of six which are a subset of

the listed ten may be used depending upon the actual equipment employed.

The interview and audiometric data will be converted to a keypunch form
so that they can be conveniently input to a computer system for tabulation
and statistical processing.

The kinds of analyses it is expected will be

accomplished on the data will utilize relatively standard software capabili­

30•
I

ties such as are available in the SPSS and BMD packages.

Beyond this any

sorting or categorization capabilities that are required can be rather
easily and quickly generated in some higher level language (e.g. Fortran IV).

For the purposes of this study protocol it is assumed that the NIOSH computer
system will be utilized.
As indicated previously, the most fundamental analysis that will be per­

formed is that of a case-control study.

It is assumed that all individuals

1
s

17

A

ja.Ajgajsi»

inappropriate to the study will have been eliminated (e.g. those who were

extensively exposed to gunfire in the armed forces, or those with demonstrated

middle ear pathology).

All those with identified hearing loss will be

specified to be in the case group.

An attempt will then be made to accom­

plish a defensible matching of individuals in the hearing loss group with
those having no detectable hearing loss.

This matching will be accomplished*^

on the basis of a number of characteristics including age, race, and length
of exposure to recreational noise.

The matched subgroups of cases and con­

trols will then be compared to determine if there is a larger percentage of

1

farmers in the former than the latter.

If there is, the difference can only

be attributable to the areas where a matching was not accomplished.

One

area where this will be intentionally true is in occupation.

?

Hence, assuming no other complicating factors, it can be determined
with some reasonable assurance, that farmers do, or do not have greater
hearing loss than their non-farmer counterparts.

One potentially signifi­

cant complicating factor is the characteristic of the farmer who will go to

fairs.

As described in the earlier section on Approach, an attempt will be

made to control for this by means of the study group to be identified a priori

in each state.

It can be argued that the self-selection process will be at

work in the a priori study group as well.

come any such tendency.

All efforts will be made to over­

-In all probability it will at least be detectable

and the interview data will be available for the entire group, providing some

insight and control of the problem.

The ideal approach as presented in Appendix A text is to define a three
dimensional matrix with each cell representing a particular vector point de­

fined by age, annual noise exposure in the work environment, and cumulative
exposure to recreational noise.

Each cell would then be filled with a group

large enough to provide statistical validity (20 to 50 individuals).

Once

the matrix is completely filled, requiring some unique kinds of individuals

J ■

at the end points, audiometric tests are given to all individuals in each
cell. The overall study proposed here falls somewhat short of this exhaus­
tive process.

However, because of the large sample size anticipated it is

possible that a similar kind of analysis can be accomplished, albeit on a

somewhat limited basis, and results which can be of considerable value ob­
tained.

J
18

*

1
The data analysis will proceed by breaking the data down in varying
degrees of fineness.

The first analysis to be performed (for each fair)

will be a simple chi square analysis to determine if there is a relationship

between farmers and non-farmers and whether an individual has a hearing loss

1
1-

or not.

The contingency table used to perform' this analysis is displayed

below.
HEARING LOSS
Yes

No

Farmers

A

OCCUPATION
Non-Farmers

3

1

The next stage will use a Mantel Haenszel chi square 10 to determine if
there is a relationship between occupation and degree of hearing loss, A

sample table to be used in this analysis follows:
!

None

HEARING LOSS
Mild
Moderate

Severe

Farmers

OCCUPATION
Non-Farmers

3 '

1
J

Since age plays such an important role in hearing loss, it will be necessary to perform this analysis for different age groups (e.g., for individuals
in each decade from 20 to 60, and those over 60) and then use a summary MantelHaenszel chi square to see if there is over-all significance.

By using this

particular technique on either of the above tables, it is possible to determine
if there is a relationship at any particular age and then if there is an over­

all summary effect controlled for age.
Since more than one fair is involved, the above analysis can be performed
for each and then a summary Mantel-Haenszel chi square can be accomplished in­

1 ■

cluding

the results of all fairs.

Hence, if control is provided for both age

and fair, the summary chi square will be summed over both age groups and dif­

I

ferent fairs.
Since there will be different types of farmers involved it should be

]

possible to obtain an idea of whether there is a difference in the amount of

J

1

19

j4,^r

4

hearing loss among different types of farmers.

In order to determine this,

a regular chi-square analysis is performed using tables as displayed below.
OCCUPATION
Non­
Farmer

Farmer

Farmer

Farmer

nA’»

».Bh

”Cn

• YES
HEARING LOSS
NO

r

Utilizing the data collected from the questionnaire, further appropriate

analyses can be performed.

Some examples of these follow.

A Mantel-Haenszel

i

chi square can be performed with one variable as degree of hearing loss and
another variable as some categorical breakdown of certain kinds of farming or

recreational activities, ordered for possible degree of noise severity.

The

sample table to do this would look as follows:

HEARING LOSS
None

Mild

Moderate

Severe

MILD

1
J

1

I
' i

]
J
n

NOISE
SEVERITY

MODERATE
SEVERE

If it is supposed that degree of hearing loss is related in some rather

regular fashion to noise insult (e.g., polynomial), regressions can be run to
determine if there is a significant relationship between degree of hearing

loss and degree of noise in the occupation or recreational activity.

In order to perform the initial analysis more carefully, certain other
aspects (e.g., recreational hearing insult) can be introduced as controlling
variables in the Mantel-Haenszel analysis.

The way this is accomplished

is to group the individuals involved by the amount of time they spend hunting
The analysis is then run for each of
(i.e., none, moderate and extensive)',
these groups of individuals and a summary chi square test performed over all

20

J!SffiU, vOTSUU.IK'JaL •.'*. JSWfc. ••

--,* -■■■£■?- -.u'..

.JS1WU

-4B- URL,

4

classes of gun noise exposure.

As analysis of the large quantity of data

anticipated proceeds, other significant questions will be developed and use­
ful approaches to interpretation will become obvious.

These will be pur­

sued and employed as appropriate to provide the most comprehensive and

1 ■

realistic view of the physiological effects of noise on the farming popula­
tion.

Problem Areas
The most serious and potentially invalidating problem areas in this
study have been presented and discussed at some length in the proceeding

text.

These are dominated by the possibility that the sample obtained will

not be truely representative of the agricultural work force.

Possible pre­

cautionary measures have been proposed which can correct for these biases
and represent an integral part of the study design.
A further significant problem area concerns the willingness, or lack

thereof, of the farmer to cooperate in the study.

This is a very real con­

sideration that must be dealt with effectively if a meaningful study is to

result.

It is proposed that the investigators interact extensively with

farmer organizations, such as cooperatives and federations, the local re­
presentatives of the USDA offices, the extension agents, and the 4^H Clubs.

1
3
1

Further, action will be taken through these organizations/individuals rather

than independently whenever possible.

The questionnaire to be administered

to each of the study participants represents a vital source of data which is

basic to the success of the study.

Appropriate care will be taken in its

design to insure that the intended meaning is conveyed by the wording, and

provisions are made to enable an accurate and useful answer to be given.
Additionally, the questionnaire will be kept as short as possible to en­
courage its completion in as accurate a manner as is reasonable under the

real world circumstances.
Accuracy and consistency in the actual performance of the audiometry
tests must be a primary consideration.

1

This will be obtained by using the

same technicians for testing in all locations and by frequent calibration
of the test equipment.

Technicians are to be trained in a well qualified

school for audiometry technicians and frequent contact will be maintained
J

with a professional audiologist throughout the testing period.

In general.

'i

21

JWiSiMH!., ^SSIUEL. ■.

a

4

data recording, handling, and transcribing must be planned and conducted so
as to minimize the probability of introducing errors.
There are a myriad of problems that will arise and must be considered
simply in the logistics involved in conducting a study of the magnitude pro­

posed.

As an example; waiting time must be minimized and/or made more

pleasant if the greatest participation is to be achieved and maintained.

These can only be recognized and dealt with as effectively as possible as
more detailed plans for the study are developed and implemented.

Schedule
The schedule of events for the study proposed here is shown on the

3

following page. It assumes that phases I and II will be the only portions
to be addressed and will be concluded in one year. The final products will
be an extensive data base and a final report. The report will contain a

description of the study as it was actually implemented and carried out and
will deal with the data base itself in terms of its form, content, possible

limitations, and interpretation along with any underlying assumptions.

It is not known at this point in time when the proposed study will be

initiated.

]

The most obvious constraining factor is the actual timing of the .

state fairs. In the four-state area involved, these fairs are scheduled each
year so as to follow each other in the fall in an essential non-overlapping
sequence over about a seven-week period.

J

This is to allow common attractions
such as livestock shows, travel time to appear at all four events. The
schedule shown in the following is keyed to the fair schedule for 1975 if it

is assumed that month 1 correlates with January of 1975.

Actual fair dates

and attendance records for the past two years are as follows:

STATE:

•I

FAIR DATES:

PAST ATTENDANCE FIGURES:
1973
1974

Illinois
Iowa

August 8 - 17, 1975
August 15 - 24, 1975

701,319

794,000

644,000

646,000

Nebraska

August 29 - September 7, 3975

543,000

577,500

Kansas

September 13 - 21, 1975

518,000

512,200

22

__ d

*__ 1

E3

i

—U

E53 ■ fga
i

■aon

UTAH BIOMEDICAL TEST LABORATORY

SCHEDULE

University of Utah
Revision:

0

Date:

6/26/75 •

TASKS

*

Epidemiological Study of Agricultural Workers Exposed to Noise
Months

TASK OR EVENT

1

1

2

4

3

5

7

6

8

10

9

12

11

I

Identify A Priori Study Groups
Develop Questionnaire

3

Establish Contacts With Ag Groups

4

Recruit Student Field Teams

5

Train Student Field Teams

6

Initial Screening of Study Groups

r

1
i

4

Field Interviews of Study Groups

8

Fair Attendance Follow-Up

9

Identify and Train Audiology Techs.

10

Training Follow-Up

11

Data Gathering:

Fair #1

12

Data Gathering:

Fair #2

13

Data Gathering:

Fair #3

14

Data Gathering:

Fair

Data Conversion and Validation

16

Data Analysis and Interpretation

17

Study Report

1
F

i

I-

4

1

r
r

IS
Notes:

i
i-

K)
V4

UBTL 003

■ iMiiat A

-a-

4

RESOURCE.REQUIREMENTS
Manpower loading and budget estimated to be required to carry out the

proposed study are summarized in the following three pages.

i-

The important

underlying assumptions were made that:
- The study will be conducted from Salt Lake City, Utah.

- The NIOSH Audiology Testing Trailer will be used.
- Audiology Techs will be trained in Salt Lake City and sent to

the four states involved to conduct testing.
- All required expertise is in-house; i.e.> no consultants are

1

used.

- Computer support will be provided by NIOSH.
- Student Field Teams will be trained on-site in the state.con­
cerned by the Field Study Coordinator.

It should also be noted that no overhead is included in the personnel
cost estimate.

This was not done as it is not known at this time who would

actually be conducting the study.

] ■

J
]
J

i

1

i

24

f

.

..

_ .

Date:

' Ena

•;

____

UTAH BIOMEDICAL TEST LABORATORY
University of Utah
Revision: 0

ESI

EZ3

LZj

SCHEDULE

MANPOWER

*

Epidemiological Study of Agricultural Workers Exposed to Noise

6/26/75 •

Months
1

TASK OR EVENT

1

2

3

4

5

6

7

8

9

10

11

1.2

Principal Investigator

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.5

1.0

1.0

1.0

Field Study Coordinator

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

1.0

.2

.2

.2

.2

.1

.1

.1

.1

.1

2.0 12.0 20.0 20.0 20.0 L7.5

5.0

3

Epidemiologist

4

Field Interview Teams

5

Audiologist

6

Audiology Techs

Data Techs
8

Biostatistician

.1

.1

.1

.1

.1

.1

.5

.5

.5

3.0

3.0

3.0

2.0

2.0

II
I

.1

.1

.1

2.0

1.0

1.0

.2

.2

.1

G

iE

10

11
12
13
14

15

I

16

17
IS
Notes:
N)
Ln

UBTL 003
h

4

Personnel Costs

a-

22,216.00

Fringe Benefits @ 19%

$139,145.00

3

s

$116,929.00

Direct Labor:

Equipment
Middle Ear Impendance Test Equipment

3
i

6,000.00

3 @ $2,000

Supplies
Forms, Pencils, etc.

3,000

Equipment Maintenance

500
3,500.00
Services
20,000

Telephone - 4,000 calls @ $5 ea.

■]

Reimbursement of fair attendees
5,000

1,000 @ $5 ea.

Incentive Cost

J

10,000

10,000 @ $1 ea.

35,000.00

Travel
j

1
j

1

Air Fare

11,744

Per Diem

7,890

Car/Truck Rental

3,685

Personal Mileage

11,250

34,569.00

TO'l’AL:

$218,214.00

26

4*

i
0-

1
S'
i

1
1

Travel Back Up
16 r/t to Springfield, Illinois

$ 3,360

16 r/t to DesMoines, Iowa

2,752

16 r/t to Lincoln, Nebraska

2,432

16 r/t to Hutchinson, Kansas

3,200

$11,744
Per Diem

263 days @ $30

7,890

Car Rental on 64 trips

2,760

Student Personal Auto Mileage Allowance:
Assumes 150 mi. r/t for 2 contacts
Total of 1,000 contacts =

75,000 mi. @

11,250

Tractor Charges for Moving Trailer

Approx. 1,500 mi. over 50 days
Assumes $14/day + $.15/mile for GSA Tractor

■J

925
$34,569

1

J
1

^7

4

1

REFERENCES
J

3'
!

1
1
3

3

n'
-]

i
i

1.

Jansen, G., Larmbekampfung, 6>, 126-128, 1962.

2.

EPA Report No.. 550/9-74-004, 1974.

3.

Glorig, A. and J. Roberts, ’’Hearing Levels of Adults by Age and Sex”,
U.S. Dept, of Health, Education and Welfare, Public Health Service
Publication No. 100-Series Il-No. 11, October 1965.

4.

Baker, L. D. and R. H. Wilkinson, Occupational Health Survey of Michigan
Farmers. Report, Department of Agricultural Engineering, Michigan
State University, 1974.

5.

Bozung, J. B., Noise Induced Hearing Losses of Michigan Farmers. Un­
published Technical Report, Department of Agricultural Engineering,
Michigan State University, 1973.

6.

Dennis, C. A. R., A Survey of Hearing Levels of Saskatchewan Farmers.
Saskatchewan Occupational Health Report No. 1, Saskatchewan Depart­
ment of Health, 1969.

7.

Willsey, D., Hearing Test Results -- Indiana Farmers. Report presented
at the National Institute of Farm Safety Meeting, October 31, 1972.

8.

Simpson, E. W. and I. L. Deshayes, ’’Tractors Produce Ear Damaging Noise”,
J. Environ. Health, 31, 1969.

9.

Criteria for a Recommended Standard. . . Occupational Exposure to Noise,
National Institute for Occupational Safety and Health, 1972.

10.

Mantel, N., ”Chi-Square Tests With One Degree of Freedom; Extensions of
the Mantel-Haenszel Procedure”, JASA, 58, 1963.

-t

a
O'

ri-

o!

0
D
APPENDIX A

g
3

•1

! •

i
1

I

-J

<
t

E

to

>

-0

APPENDIX A
MORE SPECIFIC RECOMMENDATICNS FOR RESEARCH PROCEDURE

"W

9'-

a

Ekrmers are

exposed to presbyacusic, nosoacusic and sociacusic influences in

addition to the noises (chiefly those associated with tractors and tractor-driven
machines) of their profession.

These influences will be designated P, N, S and T;

HL.3 and HL^
the hearing losses associated with.these factors will then be HLp, HI^,

Assuming these are all additive, we thus have

1

HTL-- HLp 4- HI^ + HLs *
individual’s Hearing Threshold Level (ANSI 1969).

Also assumed in

a
a

where RTL is the

3

(use of chain saws, for example, and perhaps even snowmobiles, which would be a

o

Thus

this step is that each individual is assumed to have begun with normal hearing.

HI^ - RTL - HLp -

- Hlfc

Jh the case of farmers, we may assume that the major source of sociacusis is gunfire

sociacusic noise for most people, would be regarded as an element of a farmer’s
-occupational noise)*

Suppose, then, we knew the hearing status of all farmers in America.

To

determine HI^. (our ultdmate goal), we would compare the HTLs of different subgroups*

J

Consider two groups, 1 and 2.

if HLp! and HLp2, and also HI^j

HTI^ -

would be equal to HL^ - HL^!

^n2> were either equal or both zero.

Since

HLp and HLq cannot be set equal to zero , they must be equated within the groups
Presbycusis is therefore eliminated as a factor if cnly equal-age groups are

compared.

Such a practice will also get rid of differential nosoacusic influences

if the groups’ are large enough, so it will be assumed to be followed below.

i
1

this case.

HLfc « HTL - Hlfc + C

tjo that

- HL.
IILtl - HTLg - HTI^ -HLsl + hls2 .

With our assumption that the major source of sociacusic differences among

i

farmers is gunfire, so that other sources may be ignored,

j

In

we must proceed to get

farmers with various degrees of exposure to both tractor noise and to gunfire*

Each farmer will then bo characterized in terms of ago, dogreo of exposure to
A

1

4

Appendix A—Page 2
tractor noise, and degree of exposure to gunfire*

Breaking age down into $ or 6

decades, annual noise exposure into 5 dategories, and cumulative exposure to

gunfire into perhaps h would give one the order of 100 to 120 cells in the threedimensional jnatrix»

In order to be able to infer the significance of a 5-dB

r

difference in median RTIs or inferred HI^s, at least 20 individuals should be
assigned to each cell; if information on the most sensitive 10% of ears is to

be reliable, then probably $0 would be required*

3

Thus the population examined

audiometric ally should be at least 2000, with some 6000 needed if the major concern
is not with only the average farmer*
Of course, if one were content to show differences only at a single age,

1

assuming that differences among noise and gunfire groups by say age UO-fJO will

have reached an asymptote, then a population of only h.00 to 1000 would be needed*

•1

Obviously, however, more than l|00 h0-^0-year-old farmers would have to be

contacted before one ended up with 20 in each of the 20 categories (5 degrees of

tractor noise, h of gunfire), aid this applies to the situation with f>0 in each
group as well*

]

The procedure should be, therefore, to administer the interview

(on the basis of which assignment to category is to be made) to a random or

1

exhaustive sample of farmers in the appropriate age range, using a predetermined
order; audiometry would then be performed only if the category to which he is

assigned has not yet been filled*

Upon completion of the 20 or fJO in all cate-

gories, the analysis would be performed*
The questionnaire in this case should ask for both age and birth date, as a

1
i
i

J

check against errors that people sometimes make in stating their age, information
that will allow assignment to the appropriate noise-exposure group (some integral

of level over time), and questions relating to gunfire exposure for categorization

in that regard.

There is little to be gained by applying various exclusion prin-

ciplos that attempt to eliminate various sources of nosoacusis (head blows with
!

unconsciousness, various diseases, earaches, etc*), so the individual need not be
questioned in regard to such things. as long as the main emphasis is on the median,
A-?

4i

Appendix A— Page 3

possible exception, however, is military service, which confuses the issue
in regard to both noise and gunfire, so one might exclude all veterans, or all

J

¥ those with service-connected noise exposure and/or gunfire beyond basic training.
At any rate, if the rules of assignment to category are made sufficiently explicit.

1

’ there need be no delay between interview and test.
The least-exposed group (minimum tractor noise, minimuin gun-fire) would probably
be ths Amish farmers, so the study should include an area that contains a group

of these individuals, or.it might take forever bo fill that category properly.

In

a sense, then, the Amish will serve as a control group for the other 19 categories,

If the data are carefully gathered, we would have unequivocal evidence showing

the growth of both hearing loss due to tractor noise and that ascribable to hunting

in farmers, and also an indication of possible synergistic effects.

The main problem, of course, tha^s.Ull r.emto^s.Xh9-definition-nr_the-category limits for tractor noise and gui^^_.J^pi2ot_study_yroula.j3robab]yLbe
necessary to allow an estimate of the range of each so that cne didngt gnd up
’i

by needing to question 1000 people to get that last individual jto_fill -the
maximum-noise maximum-gun fire category.

The greater specificity as to exposure

that can-be gathered, the .better^t^lly, if one is to correlate some dose
measure with the extent of

and HLS associated with exposure to farm noises.

j

J r
<

!

A-3
Media
1381.pdf

Position: 586 (11 views)