Challenges with routine data sources for PMTCT programme monitoring in East Africa: insights from Tanzania

Abstract

Routinely collected clinic data have the potential to provide much needed information on the uptake of services to prevent mother-to-child transmission (PMTCT) of HIV, and to measure HIV prevalence in pregnant women. This article describes the methodological challenges associated with using such data, based on the experiences of researchers and programme implementers in Tanzania and drawing from other examples from East Africa. PMTCT data are routinely collected in maternal and child health (MCH) clinics in East Africa using paper-based registers corresponding to distinct services within the PMTCT service continuum. This format has inherent limitations with respect to maintaining and accurately recording unique identifiers that can link patients across the different clinics (antenatal, delivery, child), and also poses challenges when compiling aggregate data. Recent improvements to recording systems include assigning unique identifiers to HIV-positive pregnant women in MCH clinics, although this should ideally be extended to all pregnant women, and recording mother and infant identifiers alongside each other in registers. The use of ‘health passports’, as in Malawi, which maintains the same antenatal clinic identifier over time, also holds promise. Routine data hold tremendous potential for clinic-level patient management, surveillance, and evaluating PMTCT/MCH programmes. Linking clinic data to community research datasets can also provide population-level estimates of coverage with PMTCT services, currently a problematic but vital statistic for monitoring programme performance and negotiating donor funding. Enhancements to indexing and recording of routine PMTCT/MCH data are needed if we are to capitalise on this rich data source.