Bangladesh is one of many countries struggling to implement the expanded approach to family planning agreed to at the International Conference on Population and Development (ICPD)¹ in the context of broader health sector reforms.² The ICPD Programme of Action mandated, among other things, the integration of family planning with broader reproductive health services, an emphasis on client-centered approaches rather than on demographic targets, and improvements in the quality and range of services provided.

The government's Health and Population Sector Strategy, introduced in 1997, reflects many of the ICPD goals as part of a broader program of reforms to improve efficiency and expand access. The strategy proposes to integrate the previously bifurcated family planning and health sectors to provide family planning and reproductive health services as part of an "essential services package" to which everyone, including the most economically disadvantaged, would have access. The strategy explicitly emphasizes client choice over the achievement of method targets. It also mandates improvements in quality of care and, in a departure from the top-down population programs of the past, calls for ongoing stakeholder involvement in program design and monitoring.³

The government program charged with implementing the revised strategy—first as part of the Health and Population Sector Programme and then the Health, Nutrition and Population Sector Programme (HNPSP)—is supported by the World Bank and a consortium of international donors. In 1999, the program began the process of establishing community clinics throughout the country. These clinics created a network of primary level community-"owned" health facilities (for which community groups are expected to contribute land and resources for maintenance, and provide some degree of oversight) to deliver the essential services package in conjunction with higher-level services. (This process has been slow and, many feel, poorly coordinated.) The health and family planning workers who previously made home visits are being gradually redeployed and based in these facilities. They are being instructed to no longer visit each client, but instead focus outreach on potential clients who may have difficulty obtaining services.

In 1997, ahead of but aligned with this government program, a group of nongovernmental organizations (NGOs) began implementing the clinic-based essential services package model as part of a seven-year bilateral health and family planning program funded by the U.S. Agency for International Development (USAID). These NGOs officially discontinued door-to-door contraceptive distribution and started offering a wider range of services through fixed and satellite clinics. Village-level depots were established in rural areas to resupply contraceptives. To improve sustainability, USAID began to channel its support through a smaller number of NGOs and to expand the use of service charges. The NGOs also introduced a variety of measures to improve quality of services and responsiveness to client needs, such as upgrades in the technical capability of staff, training of staff, improved monitoring and management practices, and expanded facilities and services. With regard to family planning, the NGOs also sought to expand the range of methods available to couples and increase the use of long-term clinical methods.⁴

The changes these NGOs introduced are both pioneering and bold. The conditions that fostered a population crisis mentality in Bangladesh have not fully abated,* and the NGOs are experimenting with previously untested service delivery strategies that some believed, if taken to scale, might seriously jeopardize Bangladesh's impressive gains in population stabilization over the past 30 years. Previous policies focused more narrowly on family planning and sought to minimize the direct and indirect costs of contraceptive use by bringing information and methods to women's homes, at no user cost or for nominal fees. Domiciliary service provision is widely credited as a key factor in the success of these policies,⁵ and the prospect of discontinuing it to provide family planning in the context of other reproductive and family health services has provoked concerns in various policy circles. Some analysts predicted detrimental effects on contraceptive prevalence and fertility rates, and even on women's status.⁶

Thus, the new NGO service delivery approaches test the strength and nature of the demand for family planning, as well as the feasibility of sustaining contraceptive use while integrating family planning with broader reproductive health services. The NGO initiative also tests the extent to which clients, families and communities, as well as providers, can overcome the entrenched attitudes and practices that developed under the previous, demographically driven, vertical family planning program.⁷

The government of Bangladesh has an important opportunity to benefit from the lessons learned by the NGOs, in both program design and implementation. Recently, the current government (elected in 2001) has been reconsidering some of the core elements of its 1997 sector strategy. According to a recent independent review commissioned by the Bangladeshi government and its development partners, the government is shifting priority from community clinics to other modes of service delivery;⁸ informal sources, such as local newspaper accounts and statements by government officials, suggest that a return to widespread domiciliary services is being considered. Moreover, a government review of its national program cited stagnating health indicators, such as the total fertility rate (TFR),^† as evidence of failure.⁹ The review suggested that the transition to a "sectorwide approach," influenced by international donors, has been too rapid and was not based on sound evidence of the advantages and potential problems entailed in such an approach. The report does not appear to consider evidence from the NGOs' experience in implementing a similar approach, albeit on a much more limited scale, to be relevant.

Debates about the future of Bangladesh's government health program are not happening in an empirical vacuum; there is relevant evidence to be gleaned from the NGOs' experience. In this comment, we highlight some of the lessons learned from the NGO process that may inform the critical decisions and processes ahead as the government designs and implements reforms under the HNPSP. Our recommendations are based on data from a recently completed three-year, multimethod, qualitative study examining NGO program implementation, and the ways in which clients, communities and program staff have responded to policy changes and adapted to new service delivery norms and practices.¹⁰ For our study, we collected data from five USAID-run NGO sites, two urban and three rural, from January 1998 through July 2000 (with retrospective questions about the time frame prior to and during the initial transition period, which began in 1997). This comment is based on data from approximately 500 in-depth, semistructured interviews conducted with clients, clinic staff and community members (which included some husbands of female clients).

TO INTEGRATE OR NOT TO INTEGRATE?

Perceived Benefits of Integration

For example, many women at the NGO clinics appreciated having access to a broader range of contraceptive methods and services.¹¹ The pill had previously been the only viable option for many women,^‡ despite their frequent dissatisfaction with side effects. A number of women we interviewed had tried to switch to clinical methods such as the injectable, but found services to be irregular or inconveniently located, or suffered side effects that were not adequately addressed. Some women had alternated between the injectable and the pill or condoms, depending on availability, often with gaps in contraceptive protection.

We also heard that women appreciated having access to a broader range of health services—not "just pills"—through the NGOs' fixed and satellite clinics, and valued having greater access to more technically skilled providers for their family planning needs than they had in the past. The interview transcripts contain many complaints about the limited skills of the former door-to-door workers and the narrow range of services and advice they provided. Clients were often frustrated with the workers' responses to contraceptive side effects. Many complained that the advice given was inadequate or unrealistic, such as recommendations to eat more "rich" foods.¹²

Perceived Costs of and Barriers to Integration

• Discontinuing long-standing doorstep services. In Bangladesh, integrating family planning and basic health services at the point of delivery effectively meant giving up or limiting domiciliary pill distribution. Under the new NGO program, former clients of the home delivery system were expected to obtain contraceptives from pharmacies, fixed and satellite clinics or, in rural areas, "depot holders"—village women who sell pills, condoms and basic medicines and help organize satellite clinics in designated areas of approximately 120-150 households. Depot holders also do selective home visits for women with special needs or constraints.

Most of the women we interviewed would have preferred the convenience and privacy of home delivery in theory, but they recognized that the system in practice was far from ideal. The workers did not come regularly enough. As one woman put it, "What good is it if they do not visit us properly?" Many users appreciated the increased reliability of supplies under the new program. Now women were able to receive family planning services when they needed them, rather than wait for the often-irregular visits of the door-to-door worker.¹³

Contrary to expectations, women did not face insurmountable social barriers to obtaining services outside the home. Over the past 10 years or so, Bangladeshi women have deviated increasingly from traditional purdah norms by working outside the home, participating in credit programs and going to providers to obtain health care. Many women in our study were already accustomed to going out to get contraceptives when home delivery periodically lapsed. They tended to describe venturing out to obtain family planning and other types of health care as a routine, well-established norm within the community at large, if not within every family.

The complete cessation of contraceptive home delivery under the new program, however, tests the extent to which women are willing and able to travel and incur costs to obtain contraceptives and reproductive health services. The change also gauges how socially acceptable it is for them to do so regularly rather than sporadically, and assesses the value attached to women's health by families, who are now required to provide more resources and support to ensure women's access to services.

Although our qualitative data do not allow us to reliably quantify the impact of the program change on contraceptive continuation, numerous women in our study reported problems in maintaining supplies of pills around the time that domiciliary services ended. (Women also discontinued use of family planning during this time for reasons unrelated to supply issues, such as side effects, a desire for pregnancy, incorrect knowledge about their method and improper use.) In most cases, however, these problems appeared to be caused more by poor communication of the program change than by dependence on home delivery or weak demand. Family planning appears to be widely regarded by both men and women as one of the basic necessities of life. Asked how it was that no one in her community discontinued oral contraceptives after the apas^§ stopped coming, one rural woman explained: "The way people manage to get their rice and fish, they also manage to get pills."

Many women receiving services at home simply lacked sufficient information to distinguish the cessation of door-to-door services from the temporary disruptions they had dealt with frequently under the old program. There was no formal system to notify domiciliary clients of the program change in any of the sites included in our study, and many women simply did not know that they needed to find alternative sources of family planning.** In the few areas where individual family planning workers took the initiative to inform women of the program change and to suggest alternative sources of supplies, women did not wait needlessly for the door-to-door worker to arrive. These few informed clients were less likely than most to experience gaps in contraceptive access during the transition to the new program.¹⁴

To the extent that women in the research sites faced ongoing mobility-related barriers to health care, these had more to do with time and domestic responsibilities than with social restrictions per se (except perhaps for newly married women). Very few women who had received contraceptive supplies from door-to-door workers were subsequently unable to obtain them from other sources because their families prevented their leaving the house. Even among women with comparatively limited mobility, going to a "known" clinic was often the exception to familial restrictions and one of the few places to which they could travel with other women or even completely alone.¹⁵

• Barriers to using integrated services. Our findings suggest a number of barriers to effective use of "one-stop health shopping." For example, many respondents associated the new NGO satellite clinics primarily with family planning. They mentioned the possible availability of other services (such as oral saline drops or basic drugs) only in response to detailed probing.¹⁶

Unfortunately, staff appeared to be doing little to counter the prevailing assumption that the NGOs provide family planning only. Indeed, in some instances, they seemed to inadvertently reinforce that notion in their outreach by disproportionately emphasizing family planning in their "marketing" of the services available. As a result, many prospective clients did not have enough information to take advantage of the essential services package; we found only a few examples of clinic staff actively promoting use of the full service package by linking family planning clients to the available services.

More effective promotion of an integrated services package requires that providers no longer think in terms of a vertical family planning model, but actively look for opportunities to connect clients with the broader array of services. One ongoing challenge to the successful implementation of an integrated family planning-reproductive health approach is the question of how to bring about a change in the mentality of both providers and clients.¹⁷ For example, although the NGOs now seem to take more initiative than in the past in providing information to clients (e.g., on contraceptive side effects), staff appear to do so inconsistently, often on a selective "need-to-know" basis. This paternalistic attitude is an extension of the top-down orientation of the old program and is inconsistent with the new, more client-centered approach.

EXPANDING ACCESS FOR THE POOR

To achieve the goal of creating universal access to basic health services, the Bangladeshi government has to address a set of issues related to equity and accountability. These issues can be roughly categorized into the following four areas:

• Frustration among the poor who rely on free health care. Our research documented a widespread feeling among the poor that they lack adequate access to health care and face discrimination in most types of facilities. In interviews about health care utilization and decision-making, the poorest respondents often said they preferred to go to government facilities for "free medicine."^††

As one pregnant rural woman put it, "We are poor people. We will go wherever we can get medicine for the least amount of money." But women also commonly perceived that getting drugs and vitamins from government facilities depended on the whim of clinic staff, a situation seen as symptomatic of corruption and discrimination against the poor. Although we did not find the same level of acrimony toward NGOs when we asked about access to medicines in those clinics, some clients nonetheless suspected they had been taken advantage of because they were poor.¹⁸

• Reliance on personal relationships for access to services. Poor respondents in our study often depended on personal relationships for treatment because they believed they would receive substandard care or have to pay more unless a better-off patron or advocate connected with the facility intervened.¹⁹ Such fears were substantiated in a recent study on unofficial payments in government health facilities, which found that a patient's personal connections and family reputation often constrained the collection of these fees by staff.^‡‡20 Respondents in our study said that providers would not dare overcharge those who were economically influential, and that better-off, more educated persons "knew how to talk" with providers. The better-off also were more likely to be personally acquainted with doctors or have family ties to them.

• Perceived unfairness of user fees. The essential services strategy is meant to provide universal access to a minimum package of health services, regardless of ability to pay. The NGOs we studied were providing this package in selected geographic areas assigned to them by the government. They collected modest fees for these services in an effort to improve financial sustainability. Poor clients, however, felt it was arbitrary and unfair to have to pay for what the government provided free in neighboring jurisdictions.

• Lack of transparency in user charges. The NGOs waived fees on the basis of need, under the official policy that no client should be turned away because of inability to pay. They were often reluctant to make the availability of free services widely known, however, for fear that too many clients would demand them. Thus, the double mandate of increasing cost recovery while expanding access to services for the poorest clients put the NGO staff in a difficult position, and many said they would have preferred not to charge clients for essential services.²¹ The government does not formally charge users for most primary health care services, but "unofficial" charges are commonly collected.²²Such corruption seems to breed a general sense of distrust; thus, when services are not made available (often due to resource constraints or poor logistics) clients frequently assume that providers are deliberately withholding medicines and treatments to which poor people are entitled.

METHOD-MIX TARGETS: WORTH THE RISK?

Ongoing government and donor concerns about the stagnating TFR have translated into a renewed desire to achieve an "appropriate method mix" and, more specifically, to increase "the proportion of clinical and terminal methods," as stated in the government's conceptual framework of its program for 2003-2006.²³ The emphasis on specific types of methods coexists with concern for high-quality, client-centered family planning services.²⁴ These dual priorities were also evident in the NGO program and, accordingly, many women interviewed in our study switched to long-term clinical methods after the program transition. By making a broader range of family planning methods more accessible to women, and by providing adequate follow-up care in case side effects develop, the new program may be filling a previously unmet need. However, our findings also suggest that the norms and behaviors developed under the previous top-down, target-oriented family planning program are difficult to shed²⁵ and may be undermining women's ability to make meaningful choices.

The historic emphasis on recruitment of contraceptive "acceptors" has lingering effects on client-provider interactions. Although we did not encounter examples of overtly coercive behavior, some providers clearly thought they had been charged with method-specific promotion, and some clients felt they were being strongly directed toward a particular (clinical) method. The persistence of a "motivation mentality" on the part of providers may manifest itself in relatively benign efforts to "market" certain methods, or it may take the form of aggressive strategies, which clients do not explicitly identify as coercive (perhaps because such strategies are so familiar). Nonetheless, such aggressive strategies perpetuate assumptions and expectations that can compromise quality of care and client choice and, ultimately, undermine program goals.²⁶

Given the new program's emphasis on meeting clients' needs and increasing clinical method use in the context of broader, better-quality services, it is unlikely that field-level staff are officially instructed to focus on recruiting women to use specific methods and, indeed, the financial incentives for staff to work this way are now substantially reduced. Nevertheless, field-level staff may be slow to fully understand and reconcile the potential contradictions between the goals of increasing the use of long-acting methods and implementing a client-centered service delivery strategy based on rights and choice.

DISCUSSION AND CONCLUSIONS

Our findings suggest that the changes in the Bangladeshi family planning program per se were neither inappropriate nor premature, although there have been problems with implementation, many of which could have been avoided. Both official and unofficial sources indicate that the pres-ent government's hesitation in moving forward and its revisitation of old service delivery approaches is related to bureaucratic and political struggles at both the ministerial and operational levels. For example, the integration of health and family planning functions has created tensions among frontline staff.²⁷ A recent government strategy paper²⁸ maintains that the former family planning fieldworkers have become "dysfunctional," implying that the plan to integrate family planning with basic health services may have been inappropriate. Unofficial sources, such as statements by local and international experts who have advised and evaluated the program, suggest that such dysfunction is most likely due to poor supervision^§§ and lack of political will to implement a service delivery model that integrates family planning with general health. (Bangladesh is not unique in dealing with the challenges of implementing a major program shift; as the case of India demonstrates, such shifts entail risks of backsliding in the face of institutional inertia or resistance, competing priorities and interests, and chronic resource constraints.²⁹)

Besides the areas allocated to programs run by the USAID-funded NGOs, the Bangladeshi government has also allocated areas to NGOs supported by the British government. Observers have noted that delivery of the essential services package in these areas is working well, but that the Bangladeshi government does not seem to be drawing lessons from this experience.³⁰ The same can be said for services supplied by the USAID-supported NGOs, as documented in a recent strategy paper produced by the Ministry of Health and Family Welfare.³¹ A powerful contingent in the Bangladeshi government is currently arguing for a reversion to the previous, nonintegrated service-delivery model on the grounds that introduction of the essential services package approach has not produced adequate progress in reducing fertility and improving other health indicators.

Reverting to the previous model would include a revitalization of door-to-door contraceptive distribution. What the officials making these arguments do not acknowledge is that the essential services package model has not been given a fair test because all the necessary components are not in place. To be sure, there is evidence that the program is not performing as well as expected. But many problems, such as persistently high discontinuation rates, are not new, and our data and other sources suggest that inadequate quality of care—particularly the absence of active and client-responsive counseling about contraceptive side effects—underlies the high discontinuation.³² Such inadequate care contributes to discontinuation more than insufficient user motivation or women's inability to leave their homes to get contraceptives.

Where an integrated essential services package model has been tested, as in the areas we studied where the USAID-funded NGO sites were providing services, the results have been promising. Problems emerged that need to be addressed, but they primarily reflect the weaknesses of the prior program and the inevitable difficulties in bringing about fundamental changes in a long-standing service delivery system.

Our research highlights issues in three areas that may be relevant for the Bangladeshi government if it is to proceed with health sector reform and its attempts to further expand access to basic health services:

• Integration of services and the discontinuation of doorstep contraceptive distribution. The new NGO program tested two important assumptions underlying the policy shift from domiciliary services—that is, that demand for family planning is strong enough to withstand the withdrawal of home-supplied methods and that changing social norms will allow women to utilize services outside the home. At our research sites, it was clear that women and men are strongly committed to fertility control, and that women are no longer stigmatized for visiting clinics. Although the NGOs have been able to capitalize on these changing norms, they are also developing strategies—such as selective home visitation by local women who run the contraceptive supply depots—to address the exceptions that persist and the remaining barriers to access.

Without regular home visitation to monitor use, however, it may be difficult to target women who fail to obtain facility-based services under the much larger government program. Other, more stable criteria, such as education or economic level, may be needed to identify those most at risk of not obtaining services,³³ but such monitoring and outreach should not be limited to encouraging contraceptive continuation. To fully realize the advantages of the essential services package and "one-stop shopping" for health care, additional efforts are needed—by both the government and the NGOs—to increase communities' awareness of the availability of more than "just pills," and to encourage providers to understand their mandate to address the full range of clients' health and family planning needs rather than promote a specific service.

• Access for the poorest. There is a lack of transparency in both NGO and government facilities regarding the availability and allocation of resources to meet the health care needs of the poor. In the context of widespread perceptions among the poor that they are discriminated against and must depend on individual connections and good will to obtain services and receive quality care, greater transparency and accountability to communities is clearly needed. The concept of a "community-owned" clinic system is an important step toward this end, but mechanisms still need to be developed to achieve it. Sustained efforts are needed to build a sense of trust in providers and a perception of fairness in the way health care resources are allocated.

In addition, the respective roles of the government and the NGOs in health service delivery need to be clarified and rationalized. As we suggest elsewhere,³⁴ if the NGOs are to function as models for improving access, quality and sustainability in the government health system, they need to demonstrate that they can reach everyone (including the poorest) in areas where they have replaced government services. In this framework, NGOs should not give high priority to the recovery of costs from users. On the other hand, if the NGOs are meant to fill a particular market niche (with the government providing a safety net), then they need to be designed to reach everyone (including the poorest).

• Attitudes and practices that evolved in a population-control framework. Our findings suggest that in light of the previous emphasis on motivating "acceptors," the NGOs' dual mandate of expanding contraceptive choice and promoting clinical methods may breed confusion and perpetuate practices and expectations that undermine program quality. Increased access to a broader range of methods afforded by the shift to a clinic-based program is a welcome change for many women, and it may improve both the reliability and the quality of family planning services. However, programs need to take care as they introduce new methods to avoid giving providers the idea that their role is to "recruit" clients to use them.³⁵

The NGOs' experiences offer lessons for the national program as it moves forward with the essential services package approach. We recommend that when implementing new policies, attention needs to be paid at all levels of service delivery to the tensions and contradictions that may emerge. Frontline service providers as well as clients may see the new strategies as being contrary to prevailing norms as well as to concurrent mandates. Without adequate and ongoing orientation, providers may therefore fail to fully understand new service delivery goals and strategies, and may respond by falling back on ingrained modes of behavior. Strategies to communicate policy changes to clients and communities are also needed to avoid unnecessary confusion, dissatisfaction and gaps in service access and utilization.

Our findings also suggest that further efforts are needed to make health services accessible and accountable to the poor. In areas the government allocates to NGOs that collect user charges, need-based exemptions should be widely publicized. The barriers presented by unofficial charges in government facilities might be reduced by activating the committees for community oversight that have already been established within each subdistrict to oversee government Family Welfare Centers.

The NGO initiative is an important test case of new sectorwide strategies being tried in Bangladesh, because of the nature of the changes involved and the speed with which they were introduced. In pioneering the difficult transition from door-to-door family planning to clinic-based essential services, the NGOs have demonstrated both the potential of the new program strategies and the challenges inherent in such a major reorientation. Clients and communities have responded favorably to many aspects of the new model, and no intractable social barriers to service utilization have emerged. Thus, our findings indicate that the policy changes made by the NGOs, reflecting both ICPD goals and new countrywide health sector reforms, were warranted and, for the most part, were implemented successfully. As the process of reform continues, both the Bangladeshi government and donors have a valuable opportunity to learn from the NGOs' experience with the new clinic-based essential services package.