Challenges and Solutions to Combining Data from Leading Global Health Surveys: An IPUMS Perspective

Anna Bolgrien, University of Minnesota
Miriam King, University of Minnesota
Devon Kristiansen, Minnesota Population Center

IPUMS Global Health freely provides integration and documentation for three leading global health surveys: Demographic and Health Surveys (DHS), UNICEF Multiple Indicator Cluster Surveys (MICS), and Performance Monitoring for Action (PMA), with nationally representative surveys in over 110 countries (42 in Africa). IPUMS Global Health has increased comparative scholarship by helping researchers analyze multiple samples within each survey collection. MICS, DHS, and PMA surveys often cover the same topics and employ similar questions and sampling. Pooling data across these IPUMS collections could extend analyses’ geographic and temporal scope, but surveys’ differences make such pooling labor-intensive and error-prone. In this paper, we identify the main barriers to combining data across IPUMS Global Health data collections and describe IPUMS’ ongoing work to increase the three surveys’ interoperability. We illustrate the gains from interoperability by presenting results for three Sustainable Development Goal (SDG) indicators, using IPUMS DHS, MICS, and PMA data from Africa.