RAPID COLLECTION OF ADULT MORTALITY DATA FROM MOBILE PHONE INTERVIEWS: AN EVALUATION OF SURVEYS CONDUCTED IN BURKINA FASO, MALAWI AND THE DEMOCRATIC REPUBLIC OF THE CONGO

Kassoum Dianou, Catholic University of Louvain (UCLouvain) and Joseph Ki-Zerbo University (UJKZ)
Shammi Luhar, London School of Hygiene and Tropical Medicine (LSHTM)
Bruno Lankoandé, Joseph Ki-Zerbo University (UJKZ)
Abdramane Soura, Université de Ouagadougou
Hervé Bassinga, Institut Supérieur des Sciences de la Population (ISSP) / Université Joseph KI-ZERBO (UJKZ) de Ouagadougou
Ashira Menashe-Oren, Universite catholique de Louvain (UCL)
Kelly McCain, London School of Hygiene and Tropical Medicine
Malebogo Tlhajoane, London School of Hygiene and Tropical Medicine
Georges Reniers, London School of Hygiene and Tropical Medicine (LSHTM)
Bruno Masquelier , Louvain University (UCL)

Reliable statistics regarding adult mortality in many low- and middle-income countries remain elusive due to the incompleteness of death registration. Mortality levels and trends are mainly derived from retrospective surveys and censuses conducted through in-person interviews. Face-to-face interviews are however expensive, time-intensive, and impractical during health crises or complex emergencies. The expansion in cell phone network coverage has opened up new possibilities for collecting demographic data through mobile phone surveys (MPS), but there is an important gap in our understanding of selection bias and reporting errors associated with this new interview modality. This study aims to assess adult mortality levels obtained through MPS conducted in these three countries during the period spanning 2021-2022. To limit respondent fatigue, we used a shortened version of the set of questions typically used in demographic surveys to ask about the survival of respondents’ siblings. We found substantial discrepancies between the mortality levels obtained from MPS and those extracted from censuses and previous DHS conducted face-to-face, with survey estimates approximately half those expected from UN estimates. The observed underestimation can be attributed in part to errors in reporting of ages and the timing of death of siblings. After imputing ages and dates based on full sibling histories collected in previous face-to-face surveys, mortality rates were much higher and consistent with alternative data sources. MPS offer promising potential for the rapid measurement of adult mortality in settings where face-to-face surveys are challenging, but they are susceptible to substantial reporting errors, particularly concerning age and date-related data.

See paper

 Presented in Session 28. Flash session New and Critical Perspectives on Data Collection and Measurement