Patient Expectations of Assigned Treatments Impact Strength of Randomised Control Trials.

Roberto Truzoli,Phil Reed,Lisa A Osborne

doi:10.3389/fmed.2021.648403

Abstract

Patient engagement with treatments potentially poses problems for interpreting the results and meaning of Randomised Control Trials (RCTs). If patients are assigned to treatments that do, or do not, match their expectations, and this impacts their motivation to engage with that treatment, it will affect the distribution of outcomes. In turn, this will impact the obtained power and error rates of RCTs. Simple Monto Carlo simulations demonstrate that these patient variables affect sample variance, and sample kurtosis. These effects reduce the power of RCTs, and may lead to false negatives, even when the randomisation process works, and equally distributes those with positive and negative views about a treatment to a trial arm.

Highlights

The Randomised Control Trial (RCT) is a flexible tool, capable of addressing many questions in clinical settings, grounded on the notion that outcomes have not been differentially determined by factors other than allocated treatment
These considerations may be addressed in a variety of ways in different fields, and there are aspects of good practise that may mitigate some of these deleterious effects on RCT outcomes
This may be regarded as a “brute” objection to RCTs that cannot be controlled away, and the best that may be achievable is to measure the influence, and recognise the advantages of alternative assessment-methodologies in some situations

Summary

Introduction

The Randomised Control Trial (RCT) is a flexible tool, capable of addressing many questions in clinical settings, grounded on the notion that outcomes have not been differentially determined by factors other than allocated treatment. The present perspective-article suggests that views around the usefulness of RCTs, within many clinical settings, may benefit from consideration of issues related to patient psychological characteristics. Patient expectations and motivation may determine aspects of sample statistics and trial results in ways that cannot be accommodated by RCTs. In particular, they may reduce statistical power relating to patients who complete the trial.

Results

Conclusion