Bayesian probabilistic projections of proportions with limited data: An application to subnational contraceptive method supply shares

Hannah Comiskey, Monash University

Co-authors: Niamh Cahill, Maynooth University; Leontine Alkema, University of Massachusetts Amherst; David Fraizer, Monash University; Worapree Maneesoonthorn, Monash University

Abstract: Engaging the private sector in contraceptive supply is crucial for equitable healthcare. Subnational data on where women access contraceptives is essential but scarce, with poor data availability hindering decision-making in some countries. We propose a Bayesian hierarchical model to project subnational public and private contraceptive supply shares in low- and middle-income countries, addressing data scarcity. Drawing on previous work at the national level, this approach leverages spatial-temporal relationships to deliver nuanced estimates for policymaking and intervention evaluation.