Testing the Feasibility of Using Routine Health Information Systems and Population Data to Understand Contraceptive Supply and Demand in Malawi Photo by HC4 This is the second blog post in a series to showcase Data for Impact (D4I) work and share lessons learned in measurement, data quality, and data use. In this post, we discuss using routine and non-routine data to better understand contraceptive use patterns and inform family planning programs in Malawi. With USAID Malawi, D4I set out to test the feasibility of using existing routine health system (RHIS) data in combination with non-routine survey data to analyze supply and demand for family planning (FP) services in the country. The information gathered in the study will support districts in Malawi to better understand contraceptive behaviors and assess supply chain challenges to meet the contraceptive demands of their populations. An Innovative Approach This was an innovative approach, notes Dr. Mai Do, lead author of the case study, from Tulane University. “RHIS and population-based data are rarely looked at together,” said Dr. Do, although they can combine to provide extremely useful information for program decision making. For example, combining RHIS data and non-routine data from surveys in Malawi can potentially offer a bigger picture of supply and demand for contraceptives in the country. This can provide more insight into FP outcomes by helping district-level decision makers understand usage of these methods and whether discontinuation is linked to stockouts, side effects, or other factors. Where did the data come from? FP supply data were obtained from Malawi’s DHIS2, the primary routine health data collection and aggregation tool used by the Malawian Ministry of Health (MOH) since 2012. FP demand data originated with questions about contraceptive behavior included in the most recent Malawi Demographic and Health Survey (DHS), conducted in 2015–16 (National Statistical Office [NSO/Malawi] and ICF, 2017). Data from the DHS showed that 45% of women of reproductive age in Malawi used contraceptives, with injectables and implants being the most frequently used methods. Since over 80% of users reported receiving these contraceptive methods from public sector facilities, Malawi’s DHIS2 has the potential to capture these two methods’ supply environments relatively comprehensively and implement measures to meet the population’s needs more fully. Noting how impactful this data could be, D4I focused its analysis on these two methods. Data consistency and quality The study identified significant fluctuations in facilities’ DHIS2 data reporting on supply of injectables and implants, as well as inconsistencies in results. Contraceptive supply decision making in Malawi is informed via a “push-pull” mechanism, wherein service delivery points (SDPs) and higher-level facilities ideally report monthly on the number of contraceptive units used or distributed. These data will then be used to inform contraceptive distribution to facilities in the following month and plans for procurement at the central level. SDPs include community-based providers, pharmacies, and some hospitals, whereas higher-level facilities include district level health centers and major hospitals at the district or central level. SDPs at the community level report their commodity amounts and stockouts to higher-level facilities. The table below shows considerable month-to-month fluctuations in reporting data on injectables (DMPA-IM) stockout by district. The teams also identified a disconnect between the number of injectables received and the number used across facilities. Comparisons within districts also showed no correlation between stockouts and the quantity of injectables used. This suggests that stockout was not driven by commodity use recorded at the facility level, nor was it indicative of the insufficient quantities received. “In addition, we don’t know from the data how many facilities in these districts are providing FP services, and therefore should be reporting on commodities,” notes Dr. Do. A closer examination of the reporting processes may help to identify gaps in reporting tasks and quality of data reported within each district. Recommendations The case study highlights a critical need for high quality, consistently available RHIS data to enable supply and demand analyses to understand contraceptive use patterns and inform FP programs. D4I proposed two key recommendations to improve the usability of commodity supply data and enhance the feasibility and usefulness of this type of data analysis in the future: 1. Establish clear criteria for which data service delivery points and health facilities need to report, and how to report it. These criteria also need to clearly distinguish different distribution mechanisms, i.e., facility-based versus community-based, depending on the specific FP method. 2. Establish a feedback mechanism for the district and central levels to review and incorporate supply and demand data into FP commodity procurement and distribution decision making. This mechanism should allow sufficient time for the district and central levels to review data and enhance data quality to improve reporting and data use. Dr. Do notes that many stakeholders in Malawi were already aware of and concerned about these data quality issues. She is hopeful that documenting data quality challenges will serve as a first step for the MOH and its partners to identify and promote solutions. Read the full report, Using Routine Data in Combination with Population Surveys to Understand Patterns of Contraceptive Use: A Case Study of Malawi, here: https://www-dev.data4impactproject.org/publications/using-routine-data-in-combination-with-population-surveys-to-understand-patterns-of-contraceptive-use-a-case-study-of-malawi/ The authors are grateful to the support and cooperation of the USAID/Malawi Mission team throughout the process of this exercise, colleagues at the Malawian Ministry of Health for allowing us access to the Health Management Information System, as well as several individuals who provided critical inputs to the analysis.
