B1178
Title: Models for longitudinal ordinal outcomes
Authors: Thomas Stewart - University of Virginia (United States) [presenting]
Abstract: Many COVID-19 studies captured patient-reported symptom data over several days, including our trial of repurposed medications. We will discuss the estimands, models, and missing data strategy for the longitudinal ordinal outcome we proposed for our trial. We will discuss computational challenges and solutions for implementing the model and missing data strategy, including generating power and sample size calculations for trial planning.
