Misreporting of Pregnancy-Related Deaths in Siblings’ Survival Histories: A Validation Study in Niakhar, Senegal

Stephane Helleringer, Johns Hopkins University
Gilles Pison, Institut National d'Études Démographiques (INED)
Bruno Masquelier, Université Catholique de Louvain
Malick Kante, Columbia University
Géraldine Duthé, Institut National d'Études Démographiques (INED)
Laetitia Douillot, Institut de Recherche pour le Développement (IRD)
Cheikh Sokhna, Institut de Recherche pour le Développement (IRD)

Estimates of maternal mortality in countries with limited vital registration increasingly rely on data from siblings’ survival histories (SSH). These data are collected during retrospective household-based surveys such as the DHS. Demographers have expressed concern that such data may underestimate the proportion of deaths due to maternal causes. We conducted a validation study of SSH in Niakhar (Senegal), a locality where prospective data on adult mortality has been collected during demographic surveillance since 1962. We found that SSH collected using the DHS questionnaire were affected by a significant number of omissions of non-PR deaths, as well as misclassifications. SSH data collected with the DHS questionnaire thus significantly overestimated the proportion of deaths due to pregnancy-related causes. On the other hand, a modified SSH questionnaire (which incorporated recall cues and an event history calendar) yielded unbiased estimates in most settings where more than 10% of deaths are due to pregnancy-related causes.