Modern Health Services and Health Care Behavior: A Survey in Kathmandu, Nepal

Modern Health Services and Health Care Behavior:
A Survey in Kathmandu,
JANARDAN SUBEDI
Miami University
Joumal of Health and Social Behavior 1989, Vol. 30 (December):412-^20
The most important problem regarding health service utilization in Third World
countries is that established indigenous forms of health care are readily available
and compete with modern health care. Thus, in addition to understanding the
components of the decision to seek medical help, we must understand the
conditions that affect the choice of a specific health care system. This study
examines the impact of medical pluralism on the use of modem forms of health
care in Nepal. The findings show that the presence of medical pluralism is a
significant factor which delays use of modem health services. Policy implications
are discussed, and the need for more research in this area is stressed.
The health of a population is an important
element in its ability to progress and develop.
If health is to be improved in a population—
particularly in developing countries—health
services must be capable of delivering
effective health care and members of the
population must use these services.
Several social scientists (Bhardwaj 1980;
Djurfeldt and Lindberg 1976; Fabrega 1980;
Foster and Anderson 1978; Frankenberg and
Larson 1976; Jaspan 1976; Kleinman 1980;
Leslie 1976, 1977, 1980; Nichter 1978; Press
1978; Taylor 1976; Young 1983) have noted
that the medical system in many places,
especially in most of the developing countries,
is pluralistic: that is, both Westem and
non-Westem forms of health care are delivered
and can be used simultaneously by the
population. For most inhabitants of developing
countries, however, access to local folk or
traditional sources of health care is easier than
access to Westem or modem health care
services.
Yet the aim of the govemments in many of
these countries has been to promote modem
* Direct all correspondence to Janardan Subedi,
Department of Sociology, Miami University,
Oxford, OH 45056. Acknowledgment: I am
grateful to Eugene B. Gallagher and Mark B.
Tausig for commenting on an earlier version of this
paper.
health care and to increase its availability.
The main goal is to make basic modem health
care services available to the entire population.
This undertaking poses a problem
because it requires both delivering modem
health care and making it acceptable to a vast,
mostly mral population that has its own set of
local beliefs and practices regarding illness
and appropriate treatment.
In most developing countries, folk and
traditional forms of health care existed and
were used commonly by the population
before the introduction of modem medicine.
These altemative types of health care services
are still used widely today despite the
introduction of modem medicine. Further,
folk and traditional health care is both socially
and culturally closer to the people, whereas
modem health care has been criticized for
being unacceptable and unsatisfying to most
of the population.
Kroeger and Franken (1981) report that the
preference for folk or traditional forms of
health care over modem health care is due
partly to the lesser "social distance" of the
former system. Gesler (1984) states that
healing or treatment has two functions:
control of sickness and providing meaning for
a person's experience of sickness. Modem
health care performs only the first function,
but folk or traditional health care performs
both. In other words, modem health care can
treat a problem effectively, whereas folk or
412
HEALTH SURVEY IN KATHMANDU, NEPAL
traditional health care not only treats a
problem but also offers a satisfying and
culturally meaningful interpretation of the
illness. Thus, in spite of their effectiveness
and curative power, modem health care
services generally are not used widely.
The literature lacks a comprehensive analysis
of the factors that affect the use of
modem health services in a pluralistic health
care system. Specifically, few studies
account for the presence and availability of
indigenous health care and how they may be
related to the decision to seek modem health
services or to the time required to reach such
a decision.
Several studies regarding use of health care
(Aday and Andersen 1974; Andersen 1968;
Kosa and Robertson 1969; Mechanic 1962;
Rosenstock 1966; Suchman 1965a, 1965b)
point out that various demographic, sociocultural,
economic, and need factors—based on
an individual's perception of his or her
illness—are important determinants of health
service use. These studies, however, were
conducted in developed countries, where
modem health care is the only recognized and
available system. The present study suggests
that in societies with pluralistic health care
systems, these factors may be necessary but
not sufficient determinants of modem health
service utilization. In such societies the
presence of medical pluralism will have a
significant impact, and must be taken into
account to provide a better understanding. For
the purpose of testing this idea, I selected the
developing country of Nepal.
UTILIZATION BEHAVIOR
A number of studies conducted in Nepal
show that persons seek different types of
health care practitioners, depending on their
perceptions and beliefs regarding the problem.
These perceptions and beliefs, in tum,
are influenced and defined by the person's
social surroundings and network relationships
(Durkin-Longley 1982; Justice 1981).
In her study, Durkin-Longley (1982) found
that upon recognizing an illness, a patient's
family first responds to the symptoms by
preparing home remedies, which are generally
folk or traditional medicines. If the
symptoms persist, the family then forms a
"therapy managing group" and selects from a
multitude of therapeutic options. In general
413
the author noted that most people reported
using multiple types of therapy, sometimes
during a single illness episode.
According to Streefiand (1985), "the most
widely prevailing medical system in Nepal is
faith-healing" (p. 1155). In Nepali society,
traditional belief and superstition play a
distinctive role. Faith healers such as jhankries
and dhamis enjoy wide public acceptance
and play a significant part in meeting the
villagers' health care needs (ESCAP 1980).
Achard (1983) reports that a patient in a hilly
region of Nepal is more likely to contact a
"jhankri/dhami" than any other health service
provider. Similarly, Shah, Shrestha, and
Parker (1978) report that more than threefourths
of all illnesses in Nepal are treated by
the traditional health care system.
Individuals seeking care for illness are
usually pragmatic; they tum to different
health care systems according to their subjective
assessment of the problem and of the
most suitable type of help (Leeson 1974). In
doing so, they test their decisions. If one
remedy does not work, they will tum to
another. The more severe and prolonged the
ailment, the more likely that individuals will
seek altemative health services when a
previously chosen option fails (Heggenhougen
1980).
Justice (1981) found that for most illnesses,
patients delayed seeking professional help and
used home remedies instead. These remedies
included herbal treatments and foods to eat or
avoid. If the problem continued, traditional
healers were the next step. According to
Justice, modem health care services were
sought only as a last resort, usually for
serious and persistent problems.
Justice also found that patients who sought
treatment at health care facilities used both
traditional and modem medicine according to
their own perceptions of effectiveness. When
modem health care failed, however, the
patients retumed frequently to the use of
home remedies and traditional practitioners.
These findings suggest that the presence of
altemative sources of health care can affect
significantly the decision to seek modem
health care services and the time required to
reach that decision. In the present study I
suggest that the effect of the pluralistic
context on the use of modem health services
will be observed as a delay in the amount of
time between initial recognition of illness
414 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
symptoms and use of modem health care
services.
SAMPLE AND METHODS
The sample consisted of 760 systematically ^
selected outpatients (389) and inpatients (371)
who were already using modem health care
services at two major hospitals (Bir Hospital
and Patan Hospital) in Kathmandu. A stmctured
survey questionnaire was administered
in face-to-face interviews. The questions were
retrospective: respondents were asked to
recall events and experience that led them to
seek modem health care services.
In order to identify the social and individual
determinants of health service utilization,
I employed Andersen's (1968) widely used
Health Behavioral Model (HBM). Certain
variables used to determine the predisposing,
enabling, and need factors that infiuenced
health service utilization were modified to fit
the characteristics of Nepali society. I also
used variables related to various aspects of
medical pluralism in order to understand their
impact on the use of modem health care
services.
The analytic strategy of this study was to
use the HBM as a baseline, to which I added
indicators of pluralistic context, and to
interpret increases in explained variance as
evidence that medical pluralism is important.
The medical pluralism component was
divided into two categories. The first category—
contained a number of variables indicating
interpersonal relationships that may
infiuence pluralistic medical behavior. I
included this category among the medical
pluralism indicators because the social network
literature (Chrisman and Kleinman
1983; Horwitz 1978; McKinlay 1973) and the
research findings discussed earlier (e.g.,
Durkin-Longley 1982), suggest that interpersonal
relationships and the "lay referral
system" (Freidson 1970) can have a significant
influence on the choice of a particular
type of health care service and on the decision
to seek that service. Thus I included interpersonal
variables in the model in order to
separate interpersonal influence from the
effects of pluralistic institutional context
where otherwise they might be confounded. If
interpersonal processes are not considered,
changes attributed to pluralism might in fact
be due to interpersonal influence processes
that are reflected in the pluralism measures
but are hidden. The second category—
pluralistic therapy—included a number of
variables that accounted for actual behavior
and use of pluralistic medicine.
I used multiple regression in order to
determine independent effects and significance
of the HBM, as well as the interpersonal
and pluralistic therapy components of
modem health service utilization.
MEASURES OF UTILIZATION
Data for this study were not collected from
household surveys, as they are in almost all
studies of health service utilization, but from
individuals who were already using services
in two major modem hospitals. Thus, in
explaining variations in modem health service
utilization, it is essential to understand what
factors or determinants led these individuals
either to use modem health care or to delay
seeking such care. Accordingly, the time
from the recognition of the specific problem
to entry into the modem hospital (for both
outpatients and inpatients) was an important
indicator: it would reflect the effect of
choosing a particular health care system, and
would explain variations in the decision to
seek modem health care services and in the
time taken to reach that decision. Time taken
for entry was our dependent variable.^
The independent variables were grouped
under five categories: predisposing, enabling,
need (as in the HBM), interpersonal, and
pluralistic therapy (pluralism indicators outside
the HBM).
Predisposing
This category was divided into three groups
of variables, as in the HBM: 1) demographic,
including age in years, sex (1 if male, 2 if
female), and marital status (1 if married, 2 if
not married^; 2) social stmctural, including
education (1 if respondent had been to school,
0 if not), religion (1 for Hindus, 2 for
non-Hindus), employment status (1 for those
in the labor force, 0 for those outside'*), and
size of family (those who eat in one kitchen);
and 3) health beliefs and attitudes.
Three variables were designed to measure
beliefs and attitudes. In the first two,
respondents were asked to state their reasons
HEALTH SURVEY IN KATHMANDU, NEPAL 415
for seeking modem health care services. The
third was used to assess in general whether
individuals believed in the efficacy of the
indigenous health care system. The first
variable indicated whether belief in the
efficacy of modem health services made
individuals choose the modem health care
system. Reasons included better care, quick
relief, faith, and convenience; these were
grouped together to represent belief. If any of
these four reasons were cited, "belief in
modem" was coded as 1; otherwise it was
coded as 0. The second variable indicated
whether the particular nature of the problem
was the reason for seeking modem hospital
services; this variable was coded as 1 for
"yes" and 0 for "no." A "yes" indicated
belief in the efficacy of modem health care, at
least for this specific type of illness. The third
indicator was used to assess whether individuals
believed in the efficacy of the indigenous
health care system for certain ailments.
Responses indicating "yes" for belief in
indigenous medicine were coded as 1; responses
indicating "no" for belief in indigenous
medicine for any ailment were
coded as 0.
Enabling
Variables enabling the use of modem
health services included questions regarding
the individual's family resources, such as
total household income,^ and community
resources such as access to modem health
care facilities and distance from modem
health care facilities (in kilometers). Access
was measured by asking respondents on a
five-point scale (on which 1 was "very
difficult" and 5 was "very easy") how
difficult or how easy it was to gain admission
to the hospital (for inpatients) or to see the
doctor (for outpatients).
Need
Three elements of need were included. The
first indicated whether the need for modem
health services was based on the severity of
the problem. This variable was measured on a
four-point scale: 1 was "so-so" and 4 was
"extremely severe." For the second measure,
respondents were asked whether the length of
the problem—i.e., the time over which the
problem had persisted—affected the need for
seeking modem health care services (coded 1
for "yes" and 0 for "no"). The final variable
was based on the clinical evaluation of need.
This indicator sought to show whether
individuals tumed to modem health care
services on the basis of the type of problem
(infectious or chronic). The problem was
identified and based on hospital diagnosis.
Infectious problems were coded as 1; chronic
problems were coded as 0.
Interpersonal
Several indicators were used to assess the
infiuence of interpersonal relationships on the
dependent measure, time taken to seek
modem health services. The first variable
asked respondents whom they first consulted
for their problem: family, friends, others, or
no one (family and friends were coded as 1;
others and no one were coded as 2). Next,
three measures asked respondents who referred
them to the modem health services:
family, friends, or self (coded 1 for "yes" in
each response, and 0 for "no"). Similarly,
three variables asked who made the decision
for respondents to seek modem health services:
family, friends, or self (1 for "yes" in each
response, and 0 for "no"). The two final
questions in this category attempted to assess
whether the respondent was pressured into
seeking modem health service by family
and/or friends (1 for "yes" in each response,
0 for "no").
Pluralistic Therapy
As with the other independent variables,
several indicators were used to determine how
different factors related to pluralism affected
the time taken to use modem health services.
The first question asked respondents whether
they had delayed seeking professional help by
using home remedies (1 for "yes," 0 for
"no"). As mentioned earlier, several studies
have shown that people generally use various
types of folk or traditional medicine in home
treatment. Second, respondents were asked
which type of medical professional help (folk,
traditional, or modem) they sought first for
their perceived problem (responses indicating
folk or traditional help were coded as 1; those
for modem help were coded 2) and were
asked about satisfaction with the type of
416 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
professional help provided. The latter variable
was measured on a five-point scale: 1 was
"very satisfied" and 5 was "very dissatisfied."
The next two measures attempted to assess
whether respondents would use folk and
traditional health care for certain ailments (an
indication of competing systems). Respondents
were asked whether they could name
ailments for which they use folk and/or
traditional health care. TTie number of ailments
(0-4) for which they would see either
folk or traditional practitioners were recorded.
High scores indicated likely use of multiple
systems. The next question was designed
specifically to find out whether the respondents'
dissatisfaction with the previous help—
i.e., folk or traditional (1 for "yes," 0 for
"no")—had led them to tum to the modem
health care system. The last two indicators
were included in the medical pluralism model
to isolate the effect of contact with folk and/or
traditional professional practitioners from
other pluralism factors. Respondents were
asked whether they had been advised to come
to the modem hospital or whether the decision
was made by either a folk or a traditional
practitioner. Both responses were coded as 1
for "yes" and 0 for "no."
RESULTS
Table 1 provides the means and standard
deviations for all the variables that were
operationalized. Table 2 shows the effects of
the predisposing, enabling, and need variables
on the dependent measure (time taken
for entry into the modem health care system).
As Table 2 shows, the total model was
found to be significant (p = .OO). The Rsquared
was .067, showing that the HBM
variables explained only approximately seven
percent of the variance in the time taken to
use modem health services.
Several variables were found to affect
significantly the time taken to seek modem
health care services. In general, older persons
and women tended to delay seeking modem
health care longer than younger persons and
men. On the other hand individuals who had
been to school and/or had higher total
household incomes took less time to seek
modem health care services than individuals
who had never been to school and/or had
lower total household incomes.
TABLE 1. Means and Standard Deviations of
the Variables in the Model
VARIABLES
Predisposing
Age
Sex
Marital status
Education
Religion
Employment status
Family size
Belief in modem health care
Nature of the problem
Belief in indigenous health care
Enabling
Total household income*
Access
Distance**
Need
Severity of the problem
Length of the problem
Type of medical problem
Interpersonal
First consulted
Referred by family
Referred by friends
Referred by self
Decision by family
Decision by friends
Decision by self
Pressured by family
Pressured by friends
Pluralistic Therapy
Home remedies
Type of professional help
Satisfaction with type of help
Use of folk health care
Use of traditional health care
Dissatisfaction with previous help
Advised by folk/traditional professional
Decision by Folk/traditional
professional
Time***
Mean
35.68
1.37
.73
1.45
1.16
.73
7.28
1.12
.47
.12
Standard
deviation
15.67
.48
.45
.49
.37
.45
4.47
.69
.50
.33
2274.55 2102.41
3.53
108.37
2.96
.44
.79
1.22
.52
.13
.29
.32
.05
.67
.22
.13
.31
1.87
2.70
.90
.52
.29
.08
.13
238.92
1.15
417.68
1.03
.50
.41
.49
.50
.34
.45
.47
.22
.47
.41
.34
.84
.34
1.27
1.31
.98
.45
.27
.33
292.11
* Per month in rupees ($1.00 = Rs. 26.00).
** In kilometers.
*** In days.
Similarly, persons who tumed to the
modem health sector on the basis of their own
subjective assessment of the problem took
longer to come to the modem hospitals than
those who did not seek such services because
of the nature of the problem. According to
Andersen (1968), need is the most important
determinant of health service utilization. Yet
among the three need variables in this study,
only the duration of the problem was found to
have a positive effect on time: persons who
HEALTH SURVEY IN KATHMANDU, NEPAL
TABLE 2. Multiple Regression for the Components
of the HMB: Predisposing,
Enabling, and Need Variables with
Time to Service (N=658)
VARIABLES Beta
Predisposing
Age
Sex
Marital status
Education
Religion
Employment status
Family size
Belief in modem health care
Nature of the problem
Belief in traditional health care
Enabling
Total household income
Access
Distance
Need
Severity of the problem
Length of the problem
Type of medical problem
2.86***
62.78**
11.68
59.08*
9.26
25.30
2.30
23.85
56.95**
5.79
.1522
.1048
.0179
.1022
.0118
.0385
.0344
.0569
.0991
.0065
- .02**
9.97
-8.68
-.1107
.0401
-.0013
•12.56
71.84***
3.48
.0452
.1246
.0049
Total
^ Change:
Predisposing
Enabling
Need
.067***
.045***
.011
.016**
* = P < .05.
** = P < .01.
*** = P < .001.
B = unstandardized regression coefficient.
Beta = standardized regression coefficient.
came to the modem hospitals because of the
long duration of the problem had taken longer
to seek modem care than those who stated
that they had not sought modem health
services because of the duration of the
problem. This finding may suggest that
persons who said that they had come to the
modem hospital because of the long duration
of the problem had relied initially on other
sources of health care, and tumed to the
modem sector because the problem persisted.
In the next stage, when the medical
pluralism variables were added along with the
HBM indicators and were regressed on time,
only three (age, education, and the nature of
the problem) of the six HBM variables which
were found to be significant earlier remained
important in the extended model (see Table
3). The total model now was found to be
significant and to explain almost 13 percent of
the variance in time taken to use modem
health care services.
Further, some significant medical pluralism
417
factors were found to affect the time taken to
use modem health services. Among interpersonal
factors and influences, the results
showed that individuals who first consulted
either family or friends upon recognizing the
illness delayed seeking modem medical help
TABLE 3. Multiple Regression Model for the
HBM and Pluralism Variables with
Time to Service (N = 644)
VARIABLES B Beta
Predisposing
Age 2.35** .1221
Sex 3.15 .0518
Marital status 22.43 .0342
Education -67.61** -.1163
Religion 17.91 .0229
Employment status 25.16 .0382
Family size l.OO .0149
Belief in modem health care 8.14 .0192
Nature of the problem 64.47** .1115
Belief in indigenous health
care -55.23 -.0621
Enabling
Total household income - .01 - .0945
Access 12.76 .0510
Distance .00 .0042
Need
Severity of the problem 18.26 .0653
Length of the problem 46.55 .0803
Type of medical problem -12.41 - .0174
Interpersonal
First consulted -65.94** -.0921
Referred by family 61.60 .1068
Referred by friends 2.44 .0028
Referred by self 20.08 .0323
Decision by family -36.77 -.0578
Decision by friends -25.59 -.0204
Decision by self -14.27 - .0227
Pressured by family -19.03 - .0269
Pressured by friends 42.39 .0505
Pluralistic Therapy
Home remedies 42.73** .1102
Type of professional help 29.76 .0352
Satisfaction with type 16.03 .0716
Use of folk health care 13.60 .0622
Use of traditional health care 22.08 .0738
Dissatisfaction with previous
help 69.93** .1097
Advised by folk/traditional
professional 104.99** .1207
Decision by folk/traditional
professional -11.33 - .0109
Total R2
R^ Change: HBM
Medical pluralism
.127***
.047**
.06***
* = P < .05.
** = P < .01.
*** = P < .001.
B = unstandardized regression coefficient
Beta = standardized regression coefficient.
418 JOURNAL OF HEALTH AND SOCL\L BEHAVIOR
longer than those who consulted others or no
one. This finding is similar to that of studies
of medical service utilization behaviors (Birkel
and Reppucci 1983; McKinlay 1973),
which found that reliance on close network
relationships for advice or help in a new
illness delays the use of professional medical
help.
Examination of the pluralistic therapy
indicators revealed some interesting features.
The use of home remedies emerged as a
significant factor in delaying use of modem
health service. Also, results showed that
when individuals tumed to the modem health
sector, they did so because they were
dissatisfied with the previous folk or traditional
professional consultation and/or because
folk or traditional practitioners had
advised them to seek modem hospital services.
Obviously, the respondents had contacted
these practitioners for the problem in
question.
Together, these variables suggest the influence
of medical pluralism. The time taken to
use modem health services is affected significantly
by interpersonal relationships, such as
who is consulted first regarding the problem.
This may be the case because—as Durkin-
Longley (1982) found in Nepal—when family
or friends are consulted for an illness
problem, they help initially by providing
home remedies. In the next stage they select
from a multitude of therapeutic options.
The use of home remedies is an important
element in the help-seeking process. Initially
individuals seem to delay seeking professional
help and to depend on home remedies
instead. Next, they tend to seek the indigenous
health services; it appears that when they
tum to the modem health services, they do so
because of their dissatisfaction with the
previous folk or traditional professional help
provided or because an indigenous health
professional advised them to seek modem
hospital services.
Thus even though the HBM can help us to
understand some individual determinants of
modem health service utilization, the inclusion
of the medical pluralism variables
enhances our understanding. These variables
show that the presence of medical pluralism
has an important effect on the use of modem
health care services: it delays such use.
DISCUSSION AND IMPLICATIONS
This study was based on two broad
assumptions. First, in developing countries
with pluralistic health care systems, widely
used models of health service utilization such
as Andersen's (1968) Health Behavioral
Model (HBM) will not be sufficient indicators
of the factors that affect the use of modem
health care services. Second, the presence
and the use of indigenous types of health care
in such countries have significant effects on
the use of modem health services. Hence they
must be taken into account in order to provide
a more complete understanding of factors
related to modem health service utilization.
Even though the HBM indicated a number
of significant individual determinants of
modem health service utilization, our results
suggested that it could not explain what
facilitates or delays an individual in seeking
modem health care service in a pluralistic
health care system. When we included
variables associated with the infiuence and
use of medical pluralism, the explanatory
power of the model was enhanced significantly.
The findings based on these variables
suggest that network infiuences and the
presence and use of indigenous types of
health care (home remedies and professionals)
were some of the significant factors which
delayed the time taken by an individual in
using modem health care services.
These findings help us to identify an issue.
The study was conducted in the capital city of
Kathmandu, where modem health care facilities
are available. In addition, the sample
consisted of persons living in and around the
capital city; as such, they had been exposed to
Westem or modem thought and culture
through the mass media, govemmental propaganda,
foreign experts, and the heavy influx
of tourists. Further, most of the modem
health care services in the hospitals are
provided by Nepali medical practitioners,
nurses, health assistants, and other paraprofessionals.
Yet the findings suggest that a major
reason why individuals delayed seeking
modem health care services was their reliance
on and use of indigenous health care before
coming to the modem hospital.
As mentioned earlier, the goal of most
govemments in developing countries has been
to raise the health status of the population by
promoting and increasing the outreach of
modem health care facilities. The results of
HEALTH SURVEY IN KATHMANDU, NEPAL
this study, however, suggest that the effectiveness
and the use of modem health care
services cannot be assured simply by increasing
their availability or by encouraging
persons in the community to use them.
The present findings show that the indigenous
health care system is a popular source of
health care and is here to stay. As long as
people are free to choose whichever type of
health care appeals to them and seems
beneficial, the probability of tuming to the
indigenous system remains high. Thus any
future policies and programs regarding the
implementation of modem health care services
must take this probability into account
because altemative types of health care are
readily available, less expensive, and socially
closer to the people in developing countries.
If modem health care is to become broadly
available, it will be important to understand
the conditions that affect the use of such care.
SUGGESTIONS FOR
FUTURE RESEARCH
Future research may begin by compensating
for some of the limitations of this study,
such as by using a random sample and
conducting household surveys. Further, the
findings of this study were based on a single
region in Nepal where modem health care
services were readily available. Future work
must be undertaken in different, large areas of
Nepal to test the specific conclusions as well
as the general approach used here. It is
surprising how little research has been
conducted on determinants of modem health
care utilization in Nepal. Finally, researchers
must review and test more accurate measures
of medical pluralism factors in order to
develop a comprehensive model of health
service utilization.
NOTES
1. For the outpatients, every fifth person who
came to the outpatient clinic was selected as a
respondent. Due to the limited number of
hospital beds and inpatients, all available
inpatients were selected and interviewed continuously
for 15 days at each hospital.
2. Two questions provided the basis on which the
dependent measure of time was calculated. In
the first question individuals were asked
specifically how many days ago the particular
illness episode or problem had begun. The
419
second question asked both inpatients and
outpatients how many days ago rtiey had had
the first contact with the hospital. This number
then was subtracted from the number of days
mentioned in the first question. Thus the
dependent variable measured the time taken by
the individual from perception of illness to
seeking modem professional services at the
hospitals.
3. This category included never-married, widowed,
divorced, and other persons who were
not currently living with spouse.
4. Students, housewives, unemployed persons,
and retired persons were grouped together to
represent persons outside the labor force.
5. This was a continuous variable measured in
rupees. It included the amount earned from
land, rents, job, and other sources.
REFERENCES
Achard, Thomas. 1983. Primary Health Care in
the Hills of Nepal. HMG/SATA: Integrated Hill
Development Project (IHDP).
Aday, Lu Ann and Ronald M. Andersen. 1974.
Development of Indices of Access to Medical
Care. Ann Arbor: Health Administrative Press.
Andersen, Ronald M. 1968. A Behavioral Model
of Families' Use of Health Services. Chicago:
(]!enter for Health Administration Studies.
Bhardwaj, M. Surinder. 1980. "Medical Pluralism
and Homeopathy: A Geographic Perspective."
Social Science and Medicine 14B:209-16.
Birkel, R.C. and N.D. Repucci. 1983. "Social
Networks, Information Seeking, and the Utilization
of Services." American Joumal of Community
Psychology 11:185-205.
Chrisman, J. Noel and Arthur Kleinman. 1983.
"Popular Health Care, Social Networks, and
Cultural Meanings: The Orientation of Medical
Anthropology." Pp. 569-90 in Handbook of
Health, Health Care, and the Health Professions,
edited by David Mechanic. New York:
Free Press.
Djurfeldt, Goran and Steffan Lindberg. 1976. Pills
against Poverty: A Study of the Introduction of
Westem Medicine in a Tamil Village. Scandinavian
Institute of Asian Studies. New Delhi:
Oxford.
Durkin-Longley, S. Maureen. 1982. "Ayurveda in
Nepal: A Medical Belief System in Action."
Doctoral dissertation. Department of Anthropology,
University of Wisconsin, Madison.
ESCAP, 1980. Population of Nepal. Bangkok:
United Nations Economic and Social Commission
for Asia and the Pacific (ESCAP).
Fabrega, Horacio. 1980. Disease and Social
Behavior: An Interdisciplinary Perspective. Cambridge:
MIT.
Foster, George M. and Barbara G. Anderson.
1978. Medical Anthropology. New York: Wiley.
420 JOURNAL OF HEALTH AND SOCIAL BEHAVIOR
Frankenberg, Ronald and Joyce Larson. 1976.
"Disease, Illness and Sickness: Social Aspects
of the Choice of Healer in a Lusaka Suburb."
Pp. 223-49 in Social Anthropology and Medicine,
edited by J.B. Loundon. New York:
Academic Press.
Freidson, Eliot. 1970. The Profession of Medicine.
New York: Dodd, Mead.
Gesler, M. Wilbert. 1984. Health Care in
Developing Countries. State College, PA: Association
of American Geographers.
Heggenhougen, H.K. 1980. "Bomohs, Doctors
and Sinsehs—Medical Pluralism in Malaysia."
Social Science and Medicine 14B:235-44.
Horwitz, Allan V. 1978. "Family, Kin, and Friend
Networks in Psychiatric Help Seeking." Social
Science and Medicine 12:297-304.
Jaspan, M.A. 1976. "The Social Organization of
Indigenous and Modem Medical Practice in
South West Sumatra." Pp. 227-42 in Asian
Medical Systems: A Comparative Study, edited
by Charles M. Leslie. Berkeley: University of
Califomia.
Justice, Judith Ann. 1981. "Intemational Planning
and Health: An Anthropological Case Study of
Nepal." Doctoral dissertation. Department of
Anthropology. University of Califomia, Berkeley.
Kleinman, Arthur. 1980. Patients and Healers in
the Context of Culture: An Exploration of the
Borderland between Anthropology, Medicine,
and Psychiatry. Berkeley: University of California.
Kosa, John and Leon S. Robertson. 1969. "The
Social Aspects of Health and Illness." Pp.
35-68 in Poverty and Health: A Sociological
Analysis, edited by John Kosa, Aaron Antonovsky,
and Irving K. Zola. Cambridge:
Harvard University.
Kroeger, A. and H.P. Franken. 1981. "The
Educational Value of Participatory Evaluation of
Primary Health Care Programs: An Experience
with Four Indigenous Populations in Ecuador."
Social Science and Medicine 15B:535-39.
Leeson, J. 1974. "Social Science and Health
Policy in Preindustrial Society." International
Journal of Health Science 4:429-40.
Leslie, Charles M. 1976. "The Ambiguities of
Medical Revivalism in Modem India." Pp.
356-67 in Asian Medical Systems: A Comparative
Study, edited by Charles M. Leslie.
Berkeley: University of Califomia.
.. 1977. "Pluralism and Integration in the
Indian and Chinese Medical Systems." Pp.
511-17 in Culture, Disease, and Healing:
Studies in Medical Anthropology, edited by
David Landy. New York: Macmillan.
1980. "Medical Pluralism in World
Perspective." Social Science and Medicine
14B:191-95.
McKinlay, John B. 1973. "Social Networks, Lay
Consultation, and Help Seeking Behavior."
Social Forces 51:275-92.
Mechanic, David. 1962. "The Concept of Illness
Behavior." Joumal of Chronic Diseases 15:
189-94.
Nichter, Mark. 1978. "Pattem of Resort in the Use
of Therapy Systems and Their Significance for
Health Planning in South Asia." Medical
Anthropology 2(2):29-58.
Press, Irwin. 1978. "Urban Folk Medicine: An
Overview." American Anthropologist 80:71-84.
1980. "Problems in the Definition and
Classification of Medical Systems." Social
Science and Medicine 14B:45-57.
Rosenstock, M. Irwin. 1966. "Why People Use
Health Services." Milbank Memorial Fund
Quarterly 44:94-127.
Shah, Moin, P. Mathura Shrestha, and Robert
Parker. 1978. Rural Health Needs: Study No. I.
Report of a Study in the Primary Health Care
Unit of Tanahu District, Nepal. Kathmandu:
Institute of Medicine.
Streefland, Peter. 1985. "The Frontier of Modem
Westem Medicine in Nepal." Social Science
and Medicine 24( 11): 1151 -59.
Suchman, Edward. 1965a. "Stages of Illness and
Medical Care." Journal of Health and Social
Behavior (i.\\4~2%.
1965b. "Social Pattems of Illness and
Medical Care." Journal of Health and Social
Behavior 5:2-16.
Taylor, Carl E. 1976. "The Place of Indigenous
Medical Practitioners in the Modemization of
Health Services." Pp. 285-99 in Asian Medical
Systems, edited by Charles M. Leslie. Berkeley:
University of Califomia.
Young, Allan. 1983. "The Relevance of Traditional
Medicine Cultures to Modem Primary
Health Care." Social Science and Medicine
17:1205-11.
JANARDAN SUBEDI is an assistant professor of sociology at Miami University in Oxford, Ohio. His
current research interests focus on various strategies to maximize the effectiveness of the modem health
care system and of family planning programs in developing countries.