5: Religion, Politics, and Ordinal Outcomes

In this episode of "In the Interim," Scott Berry discusses the vital topic of ordinal outcomes in clinical trials—a subject as controversial as politics and religion at the dinner table. Using historical examples like James Lind's 1747 scurvy trial and Austin Bradford Hill’s pioneering randomized trial, the episode explores the complexities and ongoing debates about analyzing ordinal endpoints. Berry challenges conventional analysis methods and advocates for more refined, explicit approaches, delivering valuable insights for statisticians, clinicians, and anyone involved in clinical trial designs.

Key Highlights

• Examination of the historical context of ordinal outcomes, starting with James Lind's 1747 scurvy trial.

• Discussion of the first randomized human clinical trial by Austin Bradford Hill and its implications for ordinal endpoint analysis.

• Exploration of the Modified Rankin Score as a current example of ordinal outcomes in stroke trials.

• Critique of conventional methods like dichotomization and proportional odds models for analyzing ordinal data.

• Argument for adopting utility-based approaches in clinical trial analysis for meaningful outcomes.

Quotes

• "Almost every endpoint is ordinal. So you can't escape this." – Scott Berry

• "My claim is nobody has that weight. But yet, it's very commonly done." – Scott Berry

• "Hiding behind ad hoc ways to do this, I think just leads us to bad places." – Scott Berry