Using AI/ML to Empirically Measure Intentional Dose Non-adherence: Impact on Clinical Trial Methodology and Data Interpretation

In this video Rich Christie MD PhD discusses:

Methodological questions being addressed: New technologies that can identify and quantify the scope of intentional non-adherence in clinical trials have the potential to improve data quality by ensuring more precise measurement of true treatment effects.
Introduction While the accurate estimation of intentional non-adherence rates in clinical trials is essential to determining the dose-response relationship, no accurate estimates have ever been produced. Several articles have characterized efforts by study volunteers to conceal non-adherence but there has never been a direct empirical measure of this behavior. This study used a platform combining artificial intelligence and human review to quantify the scope of these behaviors in 23 CNS trials conducted between December 2016 and May 2019.