Assessment of a New Approach to Family Planning Services in Rural Pakistan

Abstract

With 137 million inhabitants, Pakistan is the world's seventh most populous nation. According to United Nations projections, the population will grow to 285 million by 2050, at which time Pakistan will rank as the world's fourth largest country.1 The main reason for this huge projected increase and the rise in relative ranking is the slow pace of fertility decline. The level of childbearing started to fall in the 1980s, from about 7 births per woman; for the period 1995 to 2000 it was estimated to be 5.0 births per woman, a value considerably higher than for other countries in the region (3.1 for India, 4.4 for Nepal, 3.1 for Bangladesh).1,2

Socioeconomic factors offer no obvious explanation for this divergence between Pakistan and its neighbors. For instance, the country is considerably richer and more urbanized than Nepal and Bangladesh, and levels of literacy and life expectancy are similar.3 Furthermore, there is no evidence that Pakistani couples want particularly large families. The 1975 Pakistan Fertility Survey showed that, on average, married women wanted 4 children, a number typically found in Asian surveys at that time.4 The most recent national survey (1996–1997) found that the average desired family size was 3.5 children.5 It also confirmed the existence of a large unmet need for contraception: 38% of married women wanted no more children but were practicing no method of birth control. What distinguishes Pakistani couples most clearly from their neighbors is a reluctance or inability to translate reproductive preferences into appropriate behavior. The wide gap between preferences and practice stems partly from a prevailing (but erroneous) belief that Islam is opposed to contraception, and concerns about side effects and health hazards of modern methods.6,7 It also reflects the low political priority given to family planning for much of the last 30 years.

In the 1960s, President Ayub Khan initiated a vigorous family planning program that was widely applauded as a model for other Islamic countries. However, it had serious design defects. It relied heavily on one method, the intrauterine device, and on financial incentives to clients and providers. It achieved little and was discredited when Ayub Khan fell from power.8 For the next 20 years, first under the regime of Prime Minister Zulfiqar Ali Bhutto (1971–1977) and then under that of President Zia-ul-Haq (1977–1988), family planning was a low priority. During this period, several different approaches were tried, but implementation was poor and a sense of urgency was lacking. The reasons stemmed partly from domestic political considerations. Family planning was a low priority for Bhutto probably because it was so closely identified with his bitter political rival, Ayub Khan. Zia was reluctant to promote family planning vigorously because he drew much of his political support from conservative religious elements.9 During his tenure, for instance, expenditure on family planning was sharply reduced and television advertising of family planning messages was banned.

The political climate improved in the 1990s. Family planning has received steady support from successive regimes, and there have been several encouraging developments. Social marketing of contraceptives has expanded, and efforts are being made to involve private medical practitioners more closely in service provision. Perhaps the most promising initiative is the deployment of specially trained literate women to provide contraceptive information and basic services in their own and surrounding villages. A large body of international evidence shows that community-based initiatives are often effective at raising contraceptive use,10,11 and the results of small-scale projects in Pakistan have been positive.12 This approach has worked particularly well in Bangladesh and is thought to be largely responsible for the unexpectedly large decline in fertility there.13 In April 1992, a team of senior officials from the Ministry of Population Welfare, Pakistan's lead agency for family planning, visited Bangladesh. Later that year, a plan was announced to recruit and train 12 000 female village-based family planning workers by 1998.

Recruitment criteria were specified. Family planning workers are married women aged 18 to 50, with at least 10 years of schooling, who currently reside in a rural area. Recruits are trained for 7 months; the training comprises 4 months of classroom instruction interspersed with field practice. Duties include the compilation of a register of local married women of reproductive age who should be visited at home at regular intervals. Workers are supplied with oral and injectable contraceptives and condoms for distribution, together with a range of medications for the treatment of sick children. For this work, staff receive a monthly salary of 1500 rupees, equivalent to about US $25. Supervision takes the form of monthly visits from female managers, who are provided with transport. Workers also visit district centers to collect supplies and salary.

In 1994, the Ministry of Health launched a very similar but larger program involving female community-based health workers, called “lady health workers.” Recruitment criteria, training, remuneration, supervision, and method of service delivery are almost identical to those adopted by the Ministry of Population Welfare. While the main emphasis of the Ministry of Health program is on maternal and child health, the provision of family planning services is part of the package and lady health workers are supplied with oral contraceptives and condoms. The 2 ministries collaborated to minimize overlap in the placement of workers. By the end of 1996, about 5500 village-based family planning workers and 30 000 lady health workers had been trained and were operational. While the Ministry of Population Welfare program has been carefully monitored by several small-scale studies, which have documented practical problems of implementation14,15 but also encouraging signs of success in 4 Punjabi communities,16 the Ministry of Health program remains unevaluated.

This report assesses the impact of this new approach to the provision of family planning services in Pakistan. Specifically, we sought to determine whether or not use of modern reversible methods of contraception is higher in rural localities served by these community-based workers than in other localities. Attention was restricted to rural areas because the family planning workers, unlike the lady health workers, do not operate in towns and cities. The focus on reversible methods (thereby excluding sterilization) stemmed from the fact that many sterilizations preceded the start of the outreach programs. Moreover, the primary emphasis of the programs is on reversible rather than permanent methods.

METHODS

The data for this assessment come from the rural portion of the Pakistan Fertility and Family Planning Survey, a nationally representative survey conducted in late 1996 and early 1997 by the National Institute of Population Studies, Islamabad, in collaboration with the London School of Hygiene and Tropical Medicine. The detailed methodology of the survey has been previously published.5 In brief, the sample frame was prepared by the Federal Bureau of Statistics. In the rural domain, 175 geographical clusters were selected and a household census was carried out in each. From the household lists, 31 households were selected randomly in each cluster. Because of its small population, Baluchistan province was oversampled, and all results reported below have been weighted to adjust for this overrepresentation. Selected households were visited by specially trained female staff, who first identified and then interviewed ever-married women aged 15 to 49. The contact–response rate was 89%. Interviews with women covered a wide range of fertility, family planning, and health topics, including current contraceptive status and the number of household visits, if any, received from health or family planning workers in the previous 12 months.

Concurrently with the survey of women, an inventory was made of all health and family planning facilities and staff within a radius of 5 km of the center of each rural cluster. All facilities and medical practitioners were visited to ascertain the precise nature of family planning services offered. All data on service access, including access to community-based workers, are derived from this facility survey. Because Pakistani villages are typically compact, a single cluster-level measure of access is a good approximation of individual women's access. In order to obtain a complete profile of each selected cluster, the presence of other types of modern facilities or services (e.g., post office, bank, schools, and electricity) was also noted.

For the analysis, the 2 data files were linked and 2-level statistical modeling was performed with the software package Stata17 to estimate the effect of access to services, including those from community-based workers, on the use of reversible methods of contraception. As the outcome is binary (i.e., use or nonuse of a modern reversible method), logistic regression with a random effect term for the intercept is applied.18 The purpose of the random effect term is to take account of unobserved heterogeneity at the cluster level. Results are shown as unadjusted and adjusted odds ratios, with 95% confidence intervals. Widowed, divorced, and separated women (n = 171), along with sterilized couples (n = 224), were excluded. In addition, 372 women living in 12 clusters where the facilities were either closed or there was no staff to be interviewed were excluded from the analysis because key access variables could not be measured. These exclusions reduced the sample size to 4676 women residing in 163 rural clusters.

RESULTS

The unweighted results of the 1996–1997 survey show that 7% of married women in rural Pakistan were using a reversible modern method. After data are weighted to adjust for the overrepresentation of Baluchistan (a low-use province), this estimate rises to 9%, with a fairly even spread between intrauterine devices, condoms, oral pills, and injectables. In addition, 4% of couples had been sterilized and 6% were using coitus interruptus or periodic abstinence. Thus, the overall prevalence of contraceptive use in rural areas was 19%, compared with 11% recorded by the previous 1994–1995 national survey.19 Nearly half (43%) of women lived within 5 km of a static facility with staff trained in family planning, and 24% had similar access to a private practitioner who provided contraceptive services. In terms of access to the new outreach programs, 37% of rural women lived within 5 km of a lady health worker or a village-based family planning worker and an additional 10% lived within 5 km of both types of worker. In rural Pakistan as a whole, 16% of women said that they had been visited at home by a health or family planning worker within the previous 12 months. This proportion rises to 20% in clusters with access to one type of worker and to 32% in clusters with access to both types of worker (P < .001). Partly because workers display boards outside their dwelling that indicate their institutional affiliation, most women were able to distinguish between Ministry of Health workers and those from the Ministry of Population Welfare, despite the considerable overlap in the nature of their services. When the most recent visit had been from a family planning worker, 92% of women recalled that contraception had been discussed and 73% that health matters had been raised. The corresponding figures for a health worker's visit were 79% and 86%.

The middle column of Table 1 ▶ shows the crude or unadjusted effects of access to services, other cluster characteristics, and household and individual characteristics on the use of modern reversible methods. Three types of access are examined: the availability of community-based workers, the presence of a static health or family planning center having at least 1 staff member with special training in contraception, and the presence of at least 1 private practitioner who offers contraceptive services. There is no relationship between access to a static center and contraceptive use, but the other 2 types of service access are significantly related.

Other features of clusters are represented by 3 variables. The indicator of modernization represents the number of modern institutions, such as banks and post offices, in the locality. Similarly, the indicator of school availability represents the number and types of schools within 5 km of the cluster. Along with the third variable, distance from the nearest town, these variables are included in the analysis because of theoretical or commonsense expectations that they may influence the uptake of contraception. Indeed, Table 1 ▶ shows that all have statistically significant associations.

The final block of factors to be considered is derived from the survey of individuals; it includes characteristics that are known from previous studies to act as powerful determinants of contraceptive uptake. As expected, contraceptive use rises in line with household wealth, extent of schooling of the wife, and exposure to television. It is also higher among older couples, those with larger families, and those who have more sons than daughters than among other types of couples.

These effects on contraceptive use were reassessed with a multivariate model; the results in terms of adjusted odds ratios are shown in the right column of Table 1 ▶. The effect of access to community-based workers is severely attenuated, suggesting that conditions for contraceptive uptake are favorable in localities served by such workers. Nevertheless, the net effect remains significant. In localities where both types of worker are present, the probability of using a modern reversible method is increased by 74%. The difference in use between localities served by one worker and those with no worker access is in the expected direction but is not significant. The effect of access to private practitioners offering contraceptive services is also significant. Among the other cluster-level factors, only the presence of schools retains a net effect on contraceptive use. All the household and individual factors (except women's age) remain significantly associated with use.

DISCUSSION

In a population as large and rapidly growing as Pakistan's, the design of effective family planning services is an issue of both national and international importance. The key to success in many countries has been to expand information and services beyond the restrictions of a clinic-based approach. In this regard, Pakistan appears to be no exception. For many years, family planning services have been available at over 1200 family welfare centers run by the Ministry of Population Welfare. Contraceptive services are also provided at many of the Ministry of Health's rural health centers and basic health units. These facilities are severely underutilized. For instance, one evaluation showed that, on average, a family welfare center received only 2 clients per day.20 One reason for this underutilization is the limited mobility of Pakistani women, which stems from deep-rooted traditions of female seclusion, or purdah. Specific questions were asked in the 1996–1997 survey about mobility. Only 15% of rural women had been outside their village for any purpose in the previous month without being accompanied by another adult, and only 20% said that they would be able to visit a hospital by themselves. Thus, accessing a health or family planning center in a nearby village or town is logistically complex because, typically, the husband, mother-in-law, or another adult family member must be persuaded to act as an escort, thereby preserving the family's izzat, or honor.

There are many ways to expand access beyond static clinics and, over the past 30 years, many of them have been tried in Pakistan. Mobile units have proved to be expensive and ineffective, partly because of the practical difficulties of vehicle maintenance.21 The training of traditional birth attendants, or dais, has been tried on many occasions in South Asia, but their low social status prevents them from being plausible agents of social change, and their impact on family planning has been negligible.22–24 The most interesting of Pakistan's attempts to improve access occurred in the early 1970s with the deployment of 2-person teams (1 man and 1 woman) to provide contraceptive advice and supplies at the doorstep.22,25 Unfortunately, recruitment of workers became excessively politicized and the program was dismantled. It remains uncertain whether proper and sustained implementation would have had an impact. Unlike the current community-based programs, there was no insistence that staff reside in their work areas. Moreover, the deployment of mixed-sex teams in conservative rural Pakistan is problematic.

The current approach may be a sociologically more appropriate way of presenting contraception in rural Pakistan than that begun in the early 1970s. The workers are members of the communities they serve. Being literate in a society where nearly 90% of married rural women cannot read or write means that workers typically belong to respected and influential families and can thus act as bridges between village life and the outside world, which may be regarded as alien and threatening. By allowing wives or daughters-in-law to work for either ministry, male members of workers' families implicitly endorse family planning. Indeed, it is not an uncommon sight to see husbands actively helping their wives by transporting them to nearby villages on their motorbikes.

The results reported here suggest that a steady expansion of current community-based programs may be the most effective way in the short term of meeting the huge potential demand for contraceptive services in rural Pakistan and in facilitating fertility decline. The deployment of thousands of salaried workers is not a cheap option, nor should the difficulties of maintaining adequate logistical and supervisory systems be understated. The Ministry of Population Welfare currently spends about US $37 million per year on family planning; much of this expenditure is on types of service that clearly do not meet the needs of women. The village-based family planning program accounts for only 18% of expenditure. A strong case exists for increasing this fraction. In the long term, of course, integration of the health and family planning outreach programs is both desirable and inevitable. Since it lost direct responsibility for family planning in the mid-1960s, the Ministry of Health has made rather little effort to provide contraceptive advice and supplies as part of its overall health service. One of the most encouraging results to emerge from the 1996–1997 survey is that the Ministry of Health workers are actively promoting contraception. It is to be hoped that an era of closer collaboration between the 2 ministries has begun.

The results also suggest that expansion of schools in rural Pakistan is likely to have a profound effect on reproductive behavior. While the effects of parental education on fertility have been intensively studied, the influence of children's schooling opportunities has attracted comparatively little empirical attention. The positive link between access to schools and contraceptive use found in this analysis is consistent with theoretical expectations that new opportunities to invest in children will encourage parents to choose to have fewer children.26,27

CONCLUSION

This analysis shows that use of reversible modern methods of contraception is significantly higher in localities having good access to literate, female community-based workers than in localities with little or no access. Because this result is not based on an experimental study design, causal attribution has to be cautious. Nevertheless, the link between access and contraceptive use persisted despite the introduction of a wide range of controls into the analysis. It is thus highly probable that the new community-based programs of the ministries of health and population welfare are starting to have an impact. Access to private practitioners who offer contraceptive services also had a significant effect on use. In contrast, this analysis detected no impact from access to static family planning services.

The availability of schools also exerts a powerful influence on contraceptive uptake, but the presence of other modern institutions, or proximity to a town, had no effect.

Acknowledgments

Notes

M. Sultan designed the study, supervised data collection, and reviewed all drafts. J. G. Cleland was responsible for drafting and interpretation. M. M. Ali analyzed the data and reviewed all drafts. All authors designed the analysis.

Peer Reviewed

References