A new educational offering from the Statistics section of the curriculum wheel has been posted (1-1.5 hours of primary open access content).
This website will be updated every Monday (by 12:00 PM Eastern) or Tuesday (if Monday is a holiday). Given that the design, implementation, and management of pragmatic trials is a non-linear process, featured modules will relate to various sections of the curriculum wheel over time.
Statistics Section
Covariate adjustment in randomized trials
Holmberg MJ, Andersen LW. Adjustment for Baseline Characteristics in Randomized Clinical Trials. JAMA. 2022 Dec 6;328(21):2155-2156. (2-page paper)
Summary: The purpose of randomization is to ensure that there are no systematic differences between intervention groups with respect to measured and unmeasured baseline confounders. In a randomized trial without selection or information bias, an unadjusted analysis (i.e., an analysis that does not take baseline covariates into account) will provide an unbiased estimate of the intervention effect. However, there may be additional benefits to covariate adjustment. i.e., Depending on the type of outcome, adjustment may lead to increased statistical power (i.e., the ability to detect an intervention effect when present) and may increase precision in the estimated effect. This brief paper provides guidance related to covariate adjustment in randomized trials.
Adjustment for Baseline Characteristics in Randomized Clinical Trials (JN Learning - Research, Methods, Statistics) (December 7, 2023): 18-min podcast.
Summary: JAMA Statistical Editor Dr. Roger Lewis discusses adjustment for baseline covariates in randomized clinical trials with Dr. Lars Andersen (one of the authors of the related JAMA “Guide to Statistics and Methods” 2-page paper).
U.S. Food and Drug Administration. Adjusting for Covariates in Randomized Clinical Trials for Drugs and Biological Products: Guidance for Industry (May 2023): 12-page guidance document (2-page snapshot) & 8-min podcast (transcript).
Summary: Baseline covariates refer to demographic factors, disease characteristics, or other information collected from participants before the time of randomization. In many randomized trials, the primary analysis used to estimate treatment effects of a new drug (or other intervention) might not adjust for baseline covariates. However, adjusting for prognostic baseline covariates can improve statistical efficiency. This guidance describes general recommendations (applicable to trials not related to drugs, and those outside of the United States) regarding adjusting for covariates in the analysis of randomized trials.
Kahan BC, et al. The risks and rewards of covariate adjustment in randomized trials: an assessment of 12 outcomes from 8 studies. Trials. 2014 Apr 23;15:139. (7-page paper)
Summary: Adjustment for prognostic covariates can lead to increased statistical power in the analysis of randomized trials. However, adjusted analyses are not often performed in practice. This paper describes a simulation study that assessed the increase (or decrease) in power - from covariate adjustment - across several outcomes in a variety of disease areas. It was shown that adjustment for known prognostic covariates can lead to substantial increases in power, with recommendations that covariate adjustment should be routinely incorporated into the analysis of randomized trials.
European Medicines Agency. Guideline on adjustment for baseline covariates in clinical trials (EMA/CHMP/295050/2013) (February 26, 2015): 11-page guidance document.
Summary: Similar to the general recommendations found in the US FDA guidance, this guideline clarifies when and why baseline covariates should be included in the primary analysis specified in the statistical analysis plan/trial protocol. This guidance has elements relevant to non-drug trials (including trials being conducted outside of Europe).
Morris TP, et al. Planning a method for covariate adjustment in individually randomised trials: a practical guide. Trials. 2022 Apr 18;23(1):328. (17-page paper)
Summary: There are several methods one can use to adjust for baseline covariates in randomized trials. From the perspective of writing a statistical analysis plan, this paper discusses how to choose between the three most promising approaches: (1) direct adjustment (e.g., in a regression model), (2) standardization, and (3) inverse probability of treatment weighting.
Van Lancker K, et al. Covariate adjustment in randomized controlled trials: General concepts and practical considerations. Clin Trials. 2024 Aug;21(4):399-411. (13-page paper) * As this is an optional resource, an institutional login (e.g., university or research institute e-mail address) is required to access this material.
Summary: With the growing interest in covariate adjustment in the analysis of randomized trials, the US FDA and other agencies have issued guidance regarding this approach. Relatedly, the importance of distinguishing between conditional and marginal effects has been emphasized. This article describes when and how to use covariate adjustment to enhance precision in randomized trials, describes the differences between conditional and marginal estimands, and emphasizes the need to align statistical analysis methods with the chosen estimand.
Covariate adjustment in trials - MRC Clinical Trials Unit at UCL (Page last updated: February 23, 2024): 1-page website.
Summary: Describes covariate adjustment and overviews why it is pertinent in the context of randomized trials. A list of relevant publications is also provided.