Impact of Implementing Antenatal Syphilis Point-of-Care Testing on Maternal Mortality in KwaZulu-Natal, South Africa: An Interrupted Time Series Analysis.

Abstract

Background: Syphilis infection has been associated with an increased risk of HIV infection during pregnancy which poses greater risk for maternal mortality, and antenatal syphilis point-of-care (POC) testing has been introduced to improve maternal and child health outcomes. There is limited evidence on the impact of syphilis POC testing on maternal outcomes in high HIV prevalent settings. We used syphilis POC testing as a model to evaluate the impact of POC diagnostics on the improvement of maternal mortality in KwaZulu-Natal, South Africa. Methods: We extracted 132 monthly data points on the number of maternal deaths in facilities and number of live births in facilities for 12 tertiary healthcare facilities in KwaZulu-Natal (KZN), South Africa from 2004 to 2014 from District Health Information System (DHIS) health facility archived. We employed segmented Poisson regression analysis of interrupted time series to assess the impact of the exposure on maternal mortality ratio (MMR) before and after the implementation of antenatal syphilis POC testing. We processed and analyzed data using Stata Statistical Software: Release 13. (Stata, Corp LP, College Station, TX, USA). Results: The provincial average annual maternal mortality ratio (MMR) was estimated at 176.09 ± 43.92 ranging from a minimum of 68.48 to maximum of 225.49 per 100,000 live births. The data comprised 36 temporal points before the introduction of syphilis POC test exposure and 84 after the introduction in primary health care clinics in KZN. The average annual MMR for KZN from 2004 to 2014 was estimated at 176.09 ± 43.92. A decrease in MMR level was observed during 2008 after syphilis POC test implementation, followed by a rise during 2009. Analysis of the MMR trend estimates a significant 1.5% increase in MMR trends during the period before implementation and 1.3% increase after implementation of syphilis POC testing (p < 0.001). Conclusion: Although our finding suggests a brief reduction in the MMR trend after the implementation of antenatal syphilis POC testing, a continued increase in syphilis rates is seen in KwaZulu-Natal, South Africa. The study used one of the most powerful quasi-experimental research methods, segmented Poisson regression analysis of interrupted time series to model the impact of syphilis POC on maternal outcome. The study finding requires confirmation by use of more rigorous primary study design.