Public Health

Improving Infectious Disease Models with Real-World Data
Infectious disease is a leading cause of global morbidity and mortality. Transmission dynamic models often help guide public health response to infectious disease, projecting how potential interventions (e.g., therapeutics, vaccines) can help curtail its spread. However, they typically scale effects from small studies and may produce overly optimistic estimates of population-level intervention effectiveness. Post hoc observational studies could help address this by measuring intervention effectiveness in real-world settings. Observational causal inference approaches, including difference-in-differences and synthetic control methods, estimate the impact of an intervention based on empirical counterfactuals: comparing outcomes of interest between treated units with those of similar untreated units.  However, applying these to infectious disease is not straightforward, as they can produce misleading estimates with nonlinear outcomes.  Even where observational methods perform well, it remains challenging to transport estimates to new settings to predict the impact of future interventions. To address these issues, this project will develop architecture to synthesize transmission dynamic models with observational causal inference – employing empirical counterfactuals while accounting for complex population dynamics. We will illustrate our methods re-analyzing prior studies as well as applying them to new questions about respiratory illness control, in collaboration with partners in state and local public health institutions.
PI: Alyssa Bilinski, Peterson Family Assistant Professor of Health Policy, Assistant Professor of Health Services, Policy and Practice and Biostatistics