Comparing Performance of Methods Used to Identify Pregnant Women, Pregnancy Outcomes, and Child Mortality in the Iganga-Mayuge Health and Demographic Surveillance Site, Uganda

Background: In most low and middle-income countries vital events registration for births and child deaths is poor, with reporting of pregnancy outcomes highly inadequate or non-existent. Health and Demographic Surveillance System (HDSS) sites and periodic population-based household-level surveys can be used to identify pregnancies and retrospectively capture pregnancy outcomes to provide data for decision making. However, little is known about the performance of different methods in identifying pregnancy and pregnancy outcomes, yet this is critical in assessing improvements in reducing maternal and newborn mortality and stillbirths.

Objective: To explore differences between a population-based household pregnancy survey and prospective health demographic surveillance system in identifying pregnancies and their outcomes in rural eastern Uganda.

Methods: The study was done within the Iganga-Mayuge HDSS site, a member centre of the INDEPTH Network. Prospective data about pregnancies and their outcomes was collected in the routine biannual census rounds from 2006 to 2010 in the HDSS. In 2011 a cross-sectional survey using the pregnancy history survey (PHS) tool was conducted among women aged 15 to 49 years in the HDSS area. We compared differences between the HDSS biannual census updates and the PHS capture of pregnancies identified as well as neonatal and child deaths, stillbirths and abortions.

Findings: A total of 10,540 women aged 15 to 49 years were interviewed during the PHS. The PHS captured 12.8% more pregnancies than the HDSS in the most recent year (2010-2011), though between 2006 and 2010 (earlier periods) the PHS captured only 137 (0.8%) more pregnancies overall. The PHS also consistently identified more stillbirths (18.2%), spontaneous abortions (94.5%) and induced abortions (185.8%) than the prospective HDSS update rounds.

Conclusions: Surveillance sites are designed to prospectively track population-level outcomes. However, the PHS identified more pregnancy-related outcomes than the HDSS in this study. Asking about pregnancy and its outcomes may be a useful way to improve measurement of pregnancy outcomes. Further research is needed to identify the most effective methods of improving the capture of pregnancies and their outcomes within HDSS sites, household surveys and routine health information systems.