Menu
Menu
Menu
Menu
From 2001 to 2011, modern contraceptive prevalence in Uganda increased from 18% to 26%. However, modern method use, in particular use of long-acting reversible contraceptives (LARCs) and permanent methods (PMs), remained low. In the 2011 Uganda Demographic and Health Survey, only 1 of 5 married women used a LARC or PM even though 34% indicated an unmet need for contraception. Between 2011 and 2014, a social franchise and family planning voucher program, supporting 400 private facilities to provide family planning counseling and broaden contraceptive choice by adding LARCs and PMs to the service mix, offered a voucher to enable poor women to access family planning services at franchised facilities. This study analyzes service trends and voucher client demographics and estimates the contribution of the program to increasing contraceptive prevalence in Uganda, using the Impact 2 model developed by Marie Stopes International. Between March 2011 and December 2014, 330,826 women received a family planning service using the voucher, of which 70% of voucher clients chose an implant and 25% chose an intrauterine device. The median age of voucher users was 28 years; 79% had no education or only a primary education; and 48% reported they were unemployed or a housewife. We estimated that by 2014, 280,000 of the approximately 8,600,000 women of reproductive age in Uganda were using a contraceptive method provided by the program and that 120,000 of the clients were “additional users” of contraception, contributing 1.4 percentage points to the national modern contraceptive prevalence rate. The combination of family planning vouchers and a franchise-based quality improvement initiative can leverage existing private health infrastructure to substantially expand family planning access and choice for disadvantaged populations and potentially improve contraceptive prevalence when scaled nationally.
This study presents service trends and voucher client demographics from the family planning voucher program in Uganda. Routine service delivery and client data were collected on every voucher client through a voucher management information system, with client demographic data recorded at the point of voucher distribution and cross-checked by the service provider. To reduce error and fraud, MSU conducted a medical plausibility review of all claims, data verification audits of sampled claims, and intermittent client follow-up checks. All data collection and analysis were conducted according to international principles of maintaining privacy and confidentiality of personal information. Using the Impact 2 model developed by Marie Stopes International (MSI), the study estimated the health impact of the contraceptive services, such as pregnancies and maternal deaths averted, as well as contributions to contraceptive prevalence rate (CPR) growth and the contribution to national-level additional users of contraception in Uganda. Impact 2 is a publicly available Excel-based model designed to use service provision data (http://mariestopes.org/impact-2).29 Impact 2 converts service data to the estimated number of contraceptive method users in a year, rather than the total number of clients who received services each year. Because LARCs and PMs offer multiple years of contraceptive coverage, the women who use these methods may not receive services annually. For example, some women who receive a LARC in 2012 could still be using the method in 2013, without receiving another service in 2013. The model factors in discontinuation of LARCs. For short-acting methods, the model estimates the number of services required for one year’s worth of use. From the number of users of contraceptive methods, the model estimates the number of pregnancies averted and the resulting adverse health and economic outcomes averted, using best-available data on probabilities of these outcomes. The model also takes into account data on who the program is reaching—for example, some women who are “new” to a provider may not be new to contraception—and estimates how these distinctions contribute to national-level additional users of contraception, in line with goals established by the global Family Planning 2020 (FP2020) initiative.30 While it is important that the social franchise and voucher program offers quality services and a fuller choice of methods, providing clients who were already using contraception from another provider with contraception services will not result in national-level increases in contraceptive use. Impact 2 addresses this by setting a “client profile,” which categorizes clients as: Impact 2 does not allow provider-changers to contribute to national-level growth in contraceptive use. Continuing clients are important to maintain the baseline of users, while adopters offset declines in user-numbers and contribute to national additional users. However, reliable data on the proportion of voucher clients who were adopters, continuers, and provider-changers were not available from the voucher client data set. Instead, the client profile used to generate Impact 2 additional user and CPR change data was estimated from client exit interviews carried out on a random sample of family planning clients using services from MSI’s Social Franchise channel in Uganda in 2012 and 2013. The short, interviewer-administrated standardized questionnaire gathered information about the client’s demographics and recent use of contraception. In the absence of a client exit interview survey for 2014, the 2013 client profile estimate was used. Because the family planning clients surveyed included both voucher and non-voucher users, and the exit interviews were not carried out in 2011 and 2014, the exit interview client profiles were proxies for the proportion of voucher clients who were adopters, continuers, and provider-changers. Exit interview data were used for the CPR change and additional user estimates only; all other findings were based on the routine voucher client data collected as part of the voucher management process. After service data and the client profile were entered into the model, Impact 2 was run in “service life-span” mode to estimate the impact of services provided in a given year over the full life span of the methods—given that LARC and PM services will continue to provide contraceptive protection in future years. The service life-span concept applies to LARC and PM services only; for short-acting methods, there is no carry forward into future years. Using the service life-span mode ensures that the contribution of LARCs and PMs made in the first year is carried forward into subsequent years by including a modelled reduction in LARC use over time to reflect estimated discontinuation of methods use by current users.
N/A
