Very interesting session on methodological innovations for challenges in health policy and clinical care at ACIC2022!

The first talk, by Nick Seewald, was on policy evaluation in the real world. Based on work using synthetic controls to investigate the effect of opioid prescribing laws, tips were given for communication, focusing on the research question, and justifying the chosen methods.

The second talk, by Niko Ignatiadis (@nikosIgnatiadis), was on a new method for noise-induced randomization in regression discontinuity designs. It was applied to assess the effect of eligibility for antiretroviral therapy on retention in the presence of noisy measurements of CD4 count.

The final talk, by Liyang Sun, was on new methods for empirical welfare maximization subject to constraints that must be estimated. The methods were illustrated by a derivation of an optimal budget-constrained Medicaid expansion in the US.