Statistical Review – Intention-to-Treat Versus Per-Protocol Analyses

Published

Ever read a paper and feel a bit confused about all the statistical jargon and how to properly interpret the findings? At NEJM Resident 360, we’re trying something new and bringing you Statistical Review which aims to cover some key areas in research methodology and trial design. On this episode, we discuss intention-to-treat and per-protocol analyses, how to interpret studies using these approaches, and how they address bias introduced by study attrition and nonadherence.

Introduction: 0:00
Summary of Curbside Consult episode on preventing asthma exacerbations: 00:54
Introduction to intention to treat versus per protocol analyses: 1:58
Participant behavior in clinical trials and minimizing bias: 2:19
Attrition and non-adherence: 3:01
Definition and effects of attrition: 3:26
Definition and effects of non-adherence: 3:54
Intention-to-treat analysis definition and interpretation: 4:30
Limitations of Intention-to-treat analyses: 6:56
Per-protocol analyses definition and interpretation: 7:32
Dealing with bias in per-protocol analyses: 8:24
Benefits and limitations of per-protocol analyses: 9:25
Limitations of both methods: 9:43
Summary and teaching points: 10:00

This podcast is part of the series, Curbside Consults, which complements the foundational information in Rotation Prep by taking a deep dive into key topics with expert clinicians and educators. As we explore the details of pathophysiology and critique the evidence behind clinical practice, these conversations are intended to give you a better understanding of the topic and greater confidence when treating your patients.

 Angela Castellanos is a general pediatrician and editorial fellow at the NEJM.




 Amanda Fernandes is a 2018-2019 NEJM editorial fellow.