Forecasting & decision support for public health supply chains
Forecasting & decision support for public health supply chains
A session on learning, engagement, and research from the Data Lab for Social Good, Cardiff University, UK.
This session will provide an overview of the learning, engagement and research at the Data Lab for Social Good, Cardiff University UK, related to public health supply chain. It will then focus on a portfolio of projects aimed at improving forecasting and decision-making in this domain. The work covers three interconnected areas. First, we present an empirical evaluation of forecasting methods for family planning and the development of a hybrid approach to contraceptive demand forecasting. Second, we examine methods for estimating censored demand in family planning supply chains and analyse how forecast accuracy influences inventory decisions and public health outcomes. Third, we introduce a recent collaboration with the HISP Centre and DHIS2 teams to implement forecasting models within DHIS2 through CHAP, with an emphasis on practical and scalable deployment for country staff. Finally, we outline planned work exploring whether foundation and large language models can meaningfully operate on public healthcare data, given its sparsity, bias, and operational constraints.