Accounting for Residual Confounding in the Analysis

Analysis Plan


Section 4


Accounting for Residual Confounding in the Analysis

Contributors

Patrick J. Heagerty, PhD
Elizabeth R. DeLong, PhD
For the NIH Health Care Systems Research Collaboratory Biostatistics and Study Design Core

 

Contributing Editors

Damon M. Seils, MA
Jonathan McCall, MS

Despite incorporating design-based control for confounding, such as stratification, pair matching, or constrained randomization, it is sometimes advisable to also include in the analysis covariates that might still be unbalanced across the arms of the study. Depending on the goals of the study, these covariates might be at the cluster level or even at the individual level. However, depending on the sample size, the number of covariates might be limited.

Key Question: Residual Confouding

When the number of clusters is small, permutation tests might be recommended. Again, the study statistician, in collaboration with the investigators, will determine the appropriate design and analysis methods.

SECTIONS

CHAPTER SECTIONS


Version History

May 27, 2020: Added Heagerty to the contributors list and reordered the sections of this chapter as part the annual content update (changed made by D. Seils).

May 1, 2020: Added a Resources sidebar with a link to online course material as part of the annual content update (changes made by D. Seils).

January 16, 2019: Added a “key question” image, and made nonsubstantive changes to the text as part of the annual content update (changes made by D. Seils).

Published August 25, 2017

Citation:

DeLong ER. Analysis Plan: Accounting for Residual Confounding in the Analysis. In: Rethinking Clinical Trials: A Living Textbook of Pragmatic Clinical Trials. Bethesda, MD: NIH Health Care Systems Research Collaboratory. Available at: https://rethinkingclinicaltrials.org/chapters/design/analysis-plan-top/accounting-for-residual-confounding-in-the-analysis/. Updated May 27, 2020. DOI: 10.28929/018.