The Decision Modeling Lab’s research informs health policy and clinical decisions through data-driven modeling and analysis. We use methods from decision science, optimization, epidemiology, health economics, and machine learning to enable the efficient, effective, and equitable utilization of resources. We collaborate with stakeholders in medicine and health policy to maximize our impact on policy and practice while extending the state of the art in data-driven decision modeling.

A major area of focus is data-driven decision analytic modeling, which integrates individual-level data into models that compare health intervention or policy options. Traditionally, decision analyses either model an ‘average’ patient or a relatively homogeneous cohort of synthetic individuals, extracting values from the literature or expert opinion to characterize the impacted population and estimate the impact of policy alternatives. This assumes risks and costs are not distributed across the population and interventions' treatment effects are homogeneous. Our lab is developing methdos to directly integrate individual-level data to reflect the true heterogeneity in patient populations and capture differences in expected outcomes under different policy alternatives. This enables more accurate estimation of the trade-offs involved with an intervention and allows us to look at the distributional impact of interventions to reveal potential inequities.

Current projects

Currently funded projects include:

  • Analyzing the operational impact of a patient portal with propensity score matching

  • Assessing the public health value of blood donor data with Bayesian modeling

  • Developing an individualized approach to managing risk of iron deficiency in repeat blood donors using machine learning, optimization, and simulation

Pending projects

Several research projects that are pending funding decisions will use methods from epidemiology, machine learning, health economics and mathematical modeling to inform decisions in health policy and clinical practice. Prospective group members should send an email to Alton with their CV so he can determine whether any of the pending projects are aligned with your interests.

Open science research philospohy

Our group works hard to produce research that is informative, rigorous, transparent, and reproducible. The Decision Modeling Lab Manual describes our approach to open research and dissemination.

Thank you to our funders

Research in the decision modeling lab is supported by grants from: