Prioritization and Design of Clinical Trials Using Value of Information Analysis

The Department of Epidemiology Seminar Series

Open to the public.

Speaker:

Myriam Hunink, MD, PhD
Professor and Principal Investigator of Assessment of Radiological Technology (ART)
Erasmus Medical Center
Adjunct Professor, Harvard T.H. Chan School of Public Health

Abstract:

In this talk I will present a framework for clinical trial design based on value considerations rather than consideration of statistical significance. Subsequently, I will illustrate the ideas with a practical example in the area of treatment for hospitalized COVID-19 patients during the early phases of the pandemic.

Key points are:

• The traditional error-driven approach, based on controlling type I and II errors can be inefficient or even lead to false research results, resulting in wasted research resources and delayed implementation of beneficial interventions.
• The goal of a value-driven trial design approach is to optimize the use of research resources and can result in earlier adoption of novel interventions compared to the traditional approach.
• Estimating the value of new information prior to performing a clinical trial can inform a decision maker whether a clinical or health policy decision can be made with current information or if collection of extra evidence is justified. Additionally, estimating the value of new information guides study design, data collection choices, and sample size estimation.
• During the COVID-19 pandemic time-sensitive policy and implementation decisions regarding new therapies needed to be made in the face of uncertainty. We quantified the consequences of approving therapies versus pursuing further research.

Bio: Myriam Hunink, MD, PhD, is Professor of Clinical Epidemiology and Radiology at Erasmus MC, Rotterdam, and Adjunct Professor of Health Decision Sciences at Harvard T.H. Chan School of Public Health, Boston. She has experience in randomized controlled trial design in particular for non-pharmaceutical interventions and diagnostic testing, comparative effectiveness studies using decision modeling, and cost-effectiveness analysis.