|
Reproductive Health |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Issues in the Financing of Family Planning Services in Sub-Saharan AfricaChapter V: Cost of Services |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
While the previous chapters considered ways of mobilizing resources to meet family planning finance needs, this chapter considers ways of reducing funding requirements through the use of cost-saving strategies. Costs may be reduced in several ways:
The Costs of Methods and Distribution Systems: A Case Study of Kenya Before we can identify lower cost methods and distribution systems, it is necessary to assess current costs. We begin, therefore, by examining the current costs of methods and delivery systems, using a case study of Kenya. This case study illustrates that our knowledge of family planning costs is mixed and limited. This impedes efforts to identify and implement cost-minimizing strategies. Estimates of family planning costs in Kenya Many cost studies have been conducted in Africa, but their results are difficult to compare because of wide country variation in programs and economic conditions. We are focusing on only one country, Kenya, thereby eliminating some of the reasons for the variability in cost estimates. We examined nine studies of the cost of Kenyan family planning programs that have been conducted in the past few years. We began by using data from the five recent studies that provided method-specific visit cost estimates to construct estimates of method-specific costs per CYP. We make comparisons using cost per CYP, rather than cost per visit measures, because CYPs take into consideration the amount of time that a woman is protected against the risk of pregnancy by the method she receives. Method costs can best be compared using this standardized output measure. Although it has been criticized, the CYP measure remains the best means available of comparing costs and output associated with method provision.1 Each contraceptive method is assumed to have a number of CYPs associated with its distribution or use, corresponding to the average number of years of contraceptive protection that it provides. For example, because the average woman will use an IUD for 3.5 years and is assumed to be fully protected from pregnancy for this period, 3.5 CYPs are assigned to the IUD. Likewise, since a woman requires four injections to receive a year's contraceptive protection using injectables, one injection is allotted 0.25 CYPs. The cost per CYP is calculated by dividing the costs of providing and continuing to use a method (including all necessary visits and services) by the number of CYPs associated with that method. Standard conversion factors are shown in Table 5.1. In order to determine the costs of all visits, we made some assumptions about the programs' distribution policies for OCs and condoms specific to visits.2
Table 5.2 presents our results. In four of the five studies, costs per CYP are lowest for IUDs and second lowest for sterilization. In the fifth study, sterilization had the lowest costs per CYP. Costs per CYP are much higher for the other methods.
NCPD Kimunya Musau Twahir et al. AVSC
We then compared the costs of alternative distribution methods, using the Family Planning Association of Kenya (FPAK) program as an example. In that study, clinic-based delivery has lower costs per CYP than do CBD programs (U.S. $10.09 versus U.S. $16.30 (Chee 1996), data not shown). However, CBD programs provide only re-supply methods, and re-supply methods have a higher cost per CYP. The role of methodological differences There are large differences in the cost per CYP estimates reported in Table 5.2. The use of different methodologies probably accounts for much of this variation. Table 5.3 describes the methodologies used in each of the studies included in Table 5.2, as well as four other studies that provided non-method specific cost information. These additional studies were included here since they provide a useful point of methodological comparison. Unfortunately, many of these studies fail to provide a clear description of the research methods they employed. Some provide reasonable detail about the methods used to derive cost estimates, while others provide very little such information, making it difficult to interpret or compare results. The most common methodological problems appear to involve failure to include commodity and follow-up visit costs or method-specific cost estimate problems in the measurement of personnel time, and the allocation of personnel time to visit costs. Commodity costs Some of the studies reviewed include commodity costs in their estimates, while others leave them out (see Table 5.3, column 2). Since most family planning programs in Kenya receive their contraceptive commodities free of charge from donor organizations, several of the authors reasoned that they should not include the cost of these supplies in their analysis. They focused purely on the direct financial burden incurred by the individual program under examination. Failure to include the cost of commodities, however, results in an underestimation of the true cost of family planning service delivery. Even though these supplies may be free to the program, they still constitute a cost of providing services. If donor monies were withdrawn, programs would be forced to cover these costs themselves. Follow-up visit costs None of the studies examined included the costs of follow-up visits in their estimates of IUD, Norplant or tubal ligation costs. The cost per CYP for an IUD insertion, for example, should be the cost of the visit for insertion, one or more follow-up visits, and removal, divided by the average period of use to determine an annual cost. When these costs are not included, the costs per CYP of IUDs, Norplant and tubal ligation are artificially low. We corrected for the omission of both commodity and follow-up visit costs in the estimates shown in Table 5.2. For commodity costs, we used figures reported by Johnston (1994). For follow-up visit costs, we used the ratio of new to continuing client visit costs provided by Kimunya (1996) for IUD visits. Since this was the only follow-up visit cost estimate available, we also applied it to the Norplant and tubal ligation costs. We included the costs of two follow-up visits for both IUDs and Norplant (one of which would be for removal) and one follow-up visit for tubal ligations. Costs of personnel and the problem of underutilized capacity A key methodological issue, which we were not able to correct for in the estimates provided in Table 5.2, is the apparent variation in the treatment of personnel costs by the five studies reviewed. The amount of staff time devoted to various types of visits and supporting activities, as well as the amount of unused time, has a significant influence on cost estimates. The more unused provider time there is in a program, the higher its visit costs. None of the studies reviewed includes a detailed discussion of these costs, so a detailed assessment of this issue as it applies to estimates of family planning costs in Kenya is not possible. The third column in Table 5.3 indicates whether staff time was measured in the cost studies reviewed in this chapter. Most did not disaggregate the time use of clinic or CBD staff. Most of the studies deal with personnel costs only perfunctorily, generally by stating simply that the cost of staff time was included with other direct costs. Since personnel costs usually constitute the largest percentage of total costs in both clinics and CBD programs, they merit greater attention than they generally receive. There are two steps to developing accurate estimates of staff time costs: first, accurately measuring staff time use, and second, accurately applying staff time measurements to cost estimates.
Measuring personnel time Measuring staff time by activity is not a simple process. (For more information, see Janowitz and Bratt, 1994.) The most accurate method of time measurement is direct observation -- when a researcher actually watches and records the activities of personnel over a period of several days or weeks. This method is, however, both expensive and intrusive. Patient flow analysis provides accurate information on contact time, but provides no information on how non-contact time is allocated. It, too, is costly. Staff interviews can also be used to estimate time use. In this case, researchers ask providers for their estimates of how long they spend on various activities during a typical day. This method is prone not only to typical recall error, but also to the fact that providers are likely to be reluctant to report their own unproductive time. A third method is a self-administered timesheet on which providers record their own time use during the course of the day. This method, while possibly more reliable than provider interviews, is still prone to errors of underreporting of unproductive time. No method of time measurement is perfect. Nonetheless, estimates of staff time are necessary to assess the true costs of service provision; researchers must therefore choose the most appropriate among these imperfect measurement techniques. Allocating personnel time Once it has been gathered, information on staff time use must be used appropriately to determine labor costs per visit. All time, including unproductive time, must be included when deriving estimates of staff costs per visit. If only the staff time spent on direct client-related activities is used in deriving cost estimates, the actual cost of service provision is underestimated. The fourth column in Table 5.3 specifies whether all staff time was included in the study's cost estimates. For those studies that did not measure personnel time by activity, the total staff cost (i.e., salaries) is used for labor costs, so all time is correctly included -- including unproductive time. In the two studies that did assess staff time use, only the time spent per activity was included in the labor cost estimates, and the actual labor cost of family planning services is, therefore, likely to be underestimated. One study correctly included all staff time, while another did not clarify how labor costs were allocated to visits. Mobilizing underutilized capacity When family planning workers have a significant amount of time not spent with clients or carrying out other productive activities, there is underutilized capacity. Staff are being paid for non-productive time, and costs per visit are therefore high. If staff were more fully utilized, more clients could be served without increasing labor costs. Assessing the amount of underutilized capacity is especially important for the projection of future family planning costs. If demand for family planning rises, as it is expected to in all sub-Saharan African countries, the costs of providing family planning services need not rise substantially; the increased number of clients could be absorbed within the existing network of providers, and programs might not be forced to hire significantly more family planning workers or add to buildings and equipment. In order to determine the extent of unused capacity, programs should conduct cost studies that address this issue and develop practical suggestions to refocus staff attention on direct client services and other productive activities. In order to reduce underutilized capacity, work performance must be strengthened. Two interrelated changes need to take place in order to increase the level of staff work effort: more clients need to demand services, and staff need to spend their time meeting this demand. Financial or technical incentives such as increased salaries or further training can enhance staff productivity, which will improve services and attract more clients. Though a potentially difficult process, management and supervisory structures need to encourage better job performance and increased attention to clients. The Costs of Medical Barriers In recent years, work has begun to address what are referred to as "medical barriers" to contraceptive use -- that is, practices that hinder clients' access to services. Seven categories of medical barriers have been defined: inappropriate contraindications, eligibility barriers, process/scheduling hurdles, provider bias, regulatory barriers, limits on who can provide services, and inappropriate management of side effects (Bertrand et al., 1995). Some of these barriers -- such as subjecting women to unnecessary tests and procedures, asking them to return for follow-up visits more frequently than is necessary (process hurdles), or only permitting medical personnel to provide contraceptive services -- also represent an inefficient use of scarce resources. Examples of process hurdles in sub-Saharan Africa abound. In parts of West Africa, for example, women must have a blood test performed before being prescribed combined OCs, to rule out the possibility of liver and cardiovascular disease (FHI et al., 1992). Research conducted by FHI found that more than half of family planning professionals considered these tests to be necessary. However, very few women are identified to be at risk through their use. In addition, the tests are expensive. In Senegal, for example, their cost was estimated to be between U.S. $55 and U.S. $216, as much as five times the monthly per capita income (Stanback et al., 1994). In most cases, taking a brief medical history is sufficient to identify those at risk. According to a World Health Organization (WHO) survey, requiring users to return for follow-up visits more often than necessary is perhaps the most common medical barrier to access (WHO, 1992). Most of WHO's collaborating centers, for example, prescribe OCs for only two to three months at a time, and recommend that IUD users return every four to six months. In Great Britain, by contrast, an IUD follow-up visit is required six to eight weeks after insertion, and annual visits are required thereafter (Cottingham and Mehta, 1993). It is also recommended that women be counseled to return whenever they experience side effects. Using data from clinical trials, Janowitz et al. (1994) found that a reduction in the number of recommended IUD revisits is safe. Limiting the types of personnel who can provide contraceptive services presents another barrier to access with significant resource implications. In some countries, only physicians -- the most costly category of health personnel -- are permitted to perform such family planning procedures as IUD insertions, despite a growing body of evidence from numerous countries that such procedures can be performed well by non-physician staff with appropriate training and supervision (Cottingham and Mehta, 1993). Other research has shown that nurse-midwives can perform postpartum sterilizations (Dusitsin and Satyapan, 1984; Kanchanasinith et al., 1990). Medical barriers and quality of care There are concerns that efforts to remove medical barriers might worsen the quality of care -- if, for example, necessary tests and procedures are also abandoned. According to Shelton et al. (1992), "many clinical practices both help to make the best contraceptive choice and provide secondary health benefits such as screening for sexually transmitted diseases (STDs). The challenge is to separate the wheat from the chaff." There is, however, disagreement on what constitutes the wheat, and what constitutes the chaff, as discussed further below. The WHO survey -- as well as other recent analyses (e.g., Adrian et al., 1992, Angle et al., 1993) -- found a complete lack of consensus as to the minimum but necessary elements of family planning care. Part of the problem is that many of the guidelines for contraceptive provision are based on information that is now out-of-date, or were designed based on contraceptives that have been radically reformulated in recent years (King et al., 1993). Service delivery guidelines Two efforts have been made to improve the guidelines and criteria for contraceptive service delivery. WHO has developed eligibility criteria for contraceptive methods (WHO, 1995), and USAID convened a group of international experts to develop technical guidelines on the provision of a wide range of methods (Technical Guidelines Working Group, 1994 and 1997). While these efforts will go some way toward counteracting the outdated criteria and procedures currently used, debate on appropriate practices continues in many key areas (Hardee et al., 1996). For example, while there is consensus on the benefits of non-physician provision of condoms, there is no consensus on whether or not OCs should be made available over-the-counter, or on whether a physician or nurse should evaluate clients before OC use is initiated. Similarly, while there is consensus that STD screening is important for IUD users in areas with high prevalence of STD, there is no consensus on the number of follow-up visits IUD users should be required to make, or on how often, if at all, OC users should be required to make a follow-up visit for weight or blood pressure measurement and other assessments. Hardee et al. (1996) recommend that programs regularly update their guidelines where consensus exists on the "necessity or superfluity of a specific practice." Where no consensus exists, programs should make decisions about the practices they will endorse based on the most up-to-date scientific information, their program goals, the context in which they operate, and resource availability, among other factors. There is a need to find ways to ensure that such an effort is undertaken by family planning programs in sub-Saharan Africa. An assessment of existing service guidelines can serve as a starting point. In Africa, however, explicit guidelines often do not exist. This was the case in eight countries that developed national standards with assistance from the International Program for Training in Health (INTRAH, 1993). The effort began with a detailed examination of current practices "as espoused through scant documentation," and on current scientific information on contraceptives. This enabled policy-makers and physicians in each country to reach consensus on new guidelines. Some of the changed standards will not only improve access to services but also represent more efficient use of scarce resources -- for example, lifting restrictions for non-clinical distribution of OCs. A different approach was used in Kenya, where group discussions with policy-makers and providers were convened during a contraceptive technology update seminar in an effort to elucidate medical barriers and possible steps to overcome them (Huber and Jesencky, 1993). One of the efficiency-enhancing outcomes of this exercise was the elimination of the pelvic examination requirement for women to initiate or continue use of OCs. It should be noted that changing guidelines will not necessarily change practices in a manner that reduces medical barriers. In Cameroon, for example, a study was conducted to assess provider adherence to MCH/FP service policy standards and medical protocols. For the most part, practices did not change (Thompson et al., 1995). Improved monitoring and supervision are key to ensuring that improved guidelines translate to improved practice. Costs of Providing Additional Reproductive Health Services In response to calls for the provision of more comprehensive reproductive health care, many family planning programs are diversifying their services to include the management of RTIs, including the subset of STDs, as well as other reproductive health program components. These services are critical to sub-Saharan African women, given the high prevalence of RTIs in the region and their devastating sequelae, which can include pelvic inflammatory disease (PID), chronic pelvic pain, neonatal disease, infertility, social ostracism, and, in some cases, death. RTI management is particularly important in the context of family planning services, and particularly IUD services, insofar as the insertion of an IUD in a woman with an untreated infection may lead to upper reproductive tract infection with serious consequences. In addition, family planning programs can play a critical role in STD prevention, through condom provision, education, and counseling. Without adequate attention to RTIs, family planning services risk worsening the reproductive health of their clients. Mayhew (1996) suggests that the critical question for decision makers is "whether the greater costs of staff training, drugs and clinical equipment will be outweighed by the money saved from reducing (the prevalence of) STDs and sequelae needing treatment." It is often suggested that extensive screening and treatment of family planning clients is likely to be cost effective, given the high costs -- both to individuals and the health system -- of treating the sequelae of RTIs. The results of limited research to date on the costs and cost savings of alternative RTI management strategies, as well as the costs of providing integrated RTI and family planning services, are presented below. Costs and cost savings of alternative RTI management strategies Alternative strategies for RTI management include laboratory testing and treatment, presumptive treatment of all women, syndromic management, no screening or treatment, or a combination of these strategies. The costs and cost savings associated with alternative approaches hinge on the prevalence of RTIs in a population and on the costs of screening and treatment. There is a large literature on the costs and cost savings of alternative strategies for detection and treatment of STDs in the context of family planning programs in developed countries. For example, Marrazzo et al. (1997) compared the cost-effectiveness of universal, selective and no screening for chlamydia trachomatis infection among women attending family planning clinics. The study found that universal screening was most cost effective unless the prevalence of infection was very low. This finding was due to the high costs associated with failing to treat existing infection, both to the entities providing treatment and to the individual. No complete analysis of the costs and cost savings of alternative RTI screening and treatment strategies for family planning clients has been carried out in any developing country to date. Such an analysis would need to focus on the costs and cost savings of alternative strategies in two areas: 1) identifying women with RTIs; and 2) treating women with RTIs. The costs of both screening and treatment are anticipated to be lower in sub-Saharan Africa than in developed countries, given the high proportion of these costs that are made up by the wages of health personnel and lower wage rates in these countries. However, the cost savings associated with treating the sequelae of infections will also be lower, since women in the developing world are less likely than their developed country counterparts to seek treatment for such problems as pelvic pain and infertility. Cost savings are also lower because foregone earnings associated with women's ill health are lower in sub-Saharan Africa than in the developed world. However, analyses of cost savings do not take into account such intangible but very real benefits as reductions in women's physical and emotional suffering related to RTIs. If these were taken into account, the benefits of screening and early treatment would be far more significant. Miller (1998) suggests using cost-effectiveness analysis to decide who should receive lab testing. He argues that the only relevant costs to consider are those for lab testing incurred by the organization that provides these services. One possible strategy is to use lab tests on only those women at moderate risk, empirically treat those women at high risk and not test women who have no or very low risk. This approach needs to be evaluated. Because lab facilities are scarce in the developing world, and because lab testing is costly, management strategies that do not rely on lab tests are of interest. Such strategies, known as the syndromic approach to STD management, generally involve initiating treatment based on symptoms, rather than on a definitive diagnosis. This approach was developed to treat STD clients. Recently, it has been used, sometimes in combination with risk assessment, for clients at family planning clinics. Unfortunately, recent research suggests that this approach, as it is currently used, may not be effective in treating family planning clients for a cervical infection like chlamydia. For example, positive predictive values are very low, indicating that many women who are treated for cervical infections are not infected. Thus, costs are unnecessarily incurred. Moreover, the approach fails to identify a high percentage of women who are infected (for more information, see Mayaud et al., 1995 and Dallabetta et al., 1998). Costs of providing integrated RTI and family planning services There is limited literature on the costs of providing RTI services to symptomatic family planning clients in developing countries. Recent research does suggest that these clients are most efficiently served in integrated programs, where they can receive both family planning and RTI services during the same visit, rather than during separate visits. We refer here to the combined provision of existing services, rather than the addition of new services to an existing package. Two recent studies of integrated family planning and STD services in the sub-Saharan region did conclude that this approach was more efficient. The first (Twahir et al., 1996) concluded that integrating family planning and STD services for symptomatic clients in Mombasa, Kenya, reduced costs per visit -- i.e., the cost per integrated STD and family planning visit was lower than the joint cost per visit of separate consultations for each service. However, no objective information was obtained by this study on the amount of time providers spent with clients; instead, this information was obtained from interviews with providers, which was not validated using observation or other methods. Since staff time constitutes a significant proportion of costs, any inaccurate assessment of time spent with clients will have a significant impact on cost estimates. A similar analysis of two clinics in Botswana (Maribe and Stewart, 1995) also found that integrated services were less costly. While some effort was made to allocate staff costs based on observations of provider-client interactions as well as interviews with staff, it is not clear how these data were used. Other methodological problems (e.g., failure to take into account the impact of capacity utilization) make it impossible to draw firm conclusions on the cost of integrated services from these studies. Perhaps more importantly, neither of these two studies nor other research conducted elsewhere in the developing world has considered the costs of providing RTI services to asymptomatic clients. Many women with RTIs present without symptoms, and it is much more difficult, and more costly, to provide these clients with appropriate screening and treatment. In this case, it is not a question of integrating the provision of existing services, but of adding a new service to an existing care package. Conclusions An important way to reduce the gap between needs and resources is to reduce the cost of meeting needs. Too often, the focus has been on ways to increase resources. Insufficient attention has been paid to the efficiency with which services are produced. In this chapter, we have raised a number of issues relevant to the potential for reducing the costs of service provision. 1. What are the current and projected costs of family planning service provision? If programs are to develop plans for meeting their resource needs, they need information on current and projected service delivery costs. Our analysis for Kenya indicates that existing data are deficient and that additional work is needed to develop reasonably accurate estimates of both current and projected costs. 2. What are the costs of various method-delivery system combinations? Program managers need to quantify the costs of various method-delivery system combinations to enable them to understand the cost implications of their programming decisions. Our results indicate that costs per CYP are lower for long-acting methods (IUD and sterilization) and lower in clinics than in CBD programs. While costs are but one criteria upon which decisions about method-delivery system combinations are made, they do need to be considered. Programs need to understand the trade-offs they face -- in terms of costs, service quality, access, and reproductive choice -- when they consider broadening or narrowing the method mix, introducing or reducing the provision of high-cost methods, and expanding or contracting more expensive distribution systems. For example, ensuring that rural populations have access to contraceptive services may mean that more costly CBD programs need continued funding. In addition, younger and lower parity women's choices would be constrained if they did not have access to re-supply methods. 3. How can excess capacity in delivery systems be tapped? There is some evidence of excess capacity in family planning service delivery systems. If this capacity is tapped, then the additional costs of expanding services may be low: additional services could be produced using existing staff and infrastructure; added costs would include only those of contraceptives and supplies. However, it may not be easy to tap such capacity. Potential interventions -- which may be costly, and which often require changes in the "culture" of service delivery -- include salary increases to improve motivation and performance, additional staff training, and improved management and supervisory structures. 4. How can costly regulations and unnecessary procedures be eliminated or minimized? Some tests and procedures currently required before contraceptives are provided to clients are not necessary. These practices not only represent an inconvenience to users, but can add significantly to the costs of service delivery. Regulations may require that certain procedures be performed by a physician despite evidence that trained non-physicians can safely perform them. Use of non-physicians to provide additional services not only increases access, but also reduces the costs of service provision. One way to reduce adherence to costly and unnecessary regulations and procedures is to develop or revise guidelines for clinical practice that eliminate those that are unnecessary. Such efforts are now ongoing in several sub-Saharan African countries. 5. Will the integration of reproductive health services into family planning programs reduce costs? If additional reproductive health services are provided to family planning clients, then the costs of service provision will clearly rise. Adding RTI management to family planning services can be particularly costly -- particularly if services aim to treat asymptomatic as well as symptomatic women. However, there is evidence that existing services now provided separately could be provided at lower cost both to programs and to the clients if they were integrated. Integrating the provision of vertical family planning and RTI services to symptomatic clients provides an example. Programs need to consider the cost implications of integrating existing services as well as those of providing previously unavailable reproductive health services to family planning clients. POLICY ACTIONS 1. Efforts should be made to mobilize underutilized capacity in family planning services in the region. This capacity is substantial, and could be used both to expand services and improve their quality, without substantial new resources. However, the costs, financial and otherwise, of capacity mobilization should not be underestimated. 2. In light of the potentially high costs but low effectiveness of treating family planning clients for cervical infections based on risk assessment or the syndromic approach, these approaches should either be revised and retested or phased out. 3. A major effort should be made to develop and promote standards and guidelines for family planning service provision, with an emphasis on eliminating unnecessary or outdated tests and procedures. Existing, updated international standards should be assessed, modified, and implemented at the national/local levels, based on local circumstances. Such an effort will not only reduce the costs of service provision, it will also enhance access. 4. While individuals should be given a choice of contraceptive methods, the costs of these methods cannot be ignored in determining the method mix. Given limited resources, the universal provision of methods based on demand and without regard to cost will restrict the number of individuals whose need for family planning services can be met. 5. Non-physician personnel should be trained and mobilized to play a greater role in family planning service provision, with due regard for safety and quality. There is evidence that nurses can be trained to insert IUDs safely and effectively. Similarly, lower level personnel can provide OCs and injectables. Not only will this approach diminish service costs, it will also enhance access.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
