Evaluating the Feasibility and Quality of Data Collected through Rapid Mortality Mobile Phone Surveys in Malawi

Jethro Banda, Malawi Epidemiology Intervention Research Unit
Albert Dube, Malawi Epidemiology and Intervention Research Unit
Malebogo Tlhajoane, London School of Hygiene and Tropical Medicine
Estelle McLean, London School of Hygiene and Tropical Medicine (LSHTM)
Stephane Helleringer, New York University Abu Dhabi
Amelia Crampin, Karonga Prevention Study
Sarah Walters, .
Georges Reniers, London School of Hygiene and Tropical Medicine (LSHTM)

Background: Mobile phone surveys (MPS) are a promising alternative to face-to-face surveys for collecting demographic events data. However, its quality remains uncertain. Methods: We conducted a nested MPS (non-random sample, N=1995) within the Karonga Health and Demographic Surveillance System (KHDSS), Malawi. We validate MPS using individual-level comparisons to understand misreporting and omissions. Results: Respondents’ MPS age data showed little evidence of heaping, with over 86% reporting own ages within 2 years of KHDSS record. MPS captured HDSS deaths of parents, and births and deaths of own children with >90% specificity and sensitivity, and child deaths with 87% sensitivity and 98% specificity, but overestimated household sizes and deaths. There was quality variation within MPS data, with higher odds of age misreporting among those older; interviewed at workplaces/school; or interviewed late afternoon. There were also differences in age data quality by interviewer characteristics. Conclusion: Despite differences, MPS produced data of comparable quality.