Evaluating tests for cluster‐randomized trials with few clusters under generalized linear mixed models with covariate adjustment: A simulation study
Evaluierung von Tests für cluster-randomisierte Studien mit wenigen Clustern unter generalisierten linearen Mischmodellen mit Kovariatenanpassung: Simulationsstudie
Statistics in Medicine · 2024-01-30 · 43 (2) · 201-215
DOI: 10.1002/sim.9950 ↗Auf einen Blick
Die Simulationsstudie untersuchte Testverfahren für den Behandlungseffekt in cluster-randomisierten Studien mit 10 oder 20 Clustern. Bei 0,01 Intraklassenkorrelation und 2 Kovariaten lagen Likelihood-Ratio-Tests mit between-within denominator degree of freedom nahe am nominalen Niveau.
Fragestellung
Welche GLMM-basierten Tests sind bei cluster-randomisierten Studien mit wenigen Clustern und Kovariatenanpassung für binäre und Zähldaten geeignet?
Methode
Simulationsstudie zu parallelgruppigen cluster-randomisierten Studien mit 10 oder 20 Clustern; binäre und Zähldaten; Kovariatenanpassung mit 1 oder mehr personen- oder clusterbezogenen Kovariaten.
Ergebnisse
Bei 0,01 Intraklassenkorrelation und 2 Kovariaten zeigten Likelihood-Ratio-Tests mit between-within denominator degree of freedom Typ-I-Fehlerraten nahe am nominalen Niveau. Bei 5 Kovariaten schwankte die Leistung stark; keine Methode war durchgehend gut.
Schlussfolgerung der Autor:innen
Die Autor:innen empfehlen, nur für wenige Kovariaten zu adjustieren und Likelihood-Ratio-Tests mit between-within denominator degree of freedom zu verwenden.
Kernaussagen
- 10 oder 20 Cluster wurden in Simulationen untersucht
- Bei 2 Kovariaten und 0,01 ICC waren Likelihood-Ratio-Tests am besten
- Bei 5 Kovariaten zeigte keine Methode durchgehend gute Leistung
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
At a glance
Journal article in Statistics in Medicine, published online on 2023-11-07. Written by Hongxiang Qiu and 2 co-authors. Institution: Department of Epidemiology and Biostatistics, Michigan State University, East Lansing, Michigan, USA. Major MEDLINE topics: Research Design. Cited 14 times so far (Crossref); 24 references. The full text is freely available in PubMed Central. Funded by: NIDA NIH HHS, National Institute on Drug Abuse.
Abstract (original language)
Generalized linear mixed models (GLMM) are commonly used to analyze clustered data, but when the number of clusters is small to moderate, standard statistical tests may produce elevated type I error rates. Small‐sample corrections have been proposed for continuous or binary outcomes without covariate adjustment. However, appropriate tests to use for count outcomes or under covariate‐adjusted models remains unknown. An important setting in which this issue arises is in cluster‐randomized trials (CRTs). Because many CRTs have just a few clusters (eg, clinics or health systems), covariate adjustment is particularly critical to address potential chance imbalance and/or low power (eg, adjustment following stratified randomization or for the baseline value of the outcome). We conducted simulations to evaluate GLMM‐based tests of the treatment effect that account for the small (10) or moderate (20) number of clusters under a parallel‐group CRT setting across scenarios of covariate adjustment (including adjustment for one or more person‐level or cluster‐level covariates) for both binary and count outcomes. We find that when the intraclass correlation is non‐negligible (0.01) and the number of covariates is small (2), likelihood ratio tests with a between‐within denominator degree of freedom have type I error rates close to the nominal level. When the number of covariates is moderate (5), across our simulation scenarios, the relative performance of the tests varied considerably and no method performed uniformly well. Therefore, we recommend adjusting for no more than a few covariates and using likelihood ratio tests with a between‐within denominator degree of freedom.
Subjects
MeSH terms (MEDLINE)
Author keywords
Authors and institutions
- Department of Epidemiology and Biostatistics, Michigan State University, East Lansing, Michigan, USA.
- Biostatistics Unit, Kaiser Permanente Washington Health Research Institute, Seattle, Washington, USA.
- Department of Biostatistics, University of Washington, Seattle, Washington, USA.
ORCID profiles: Hongxiang Qiu, Jennifer F Bobb
Funding
- NIDA NIH HHS
- National Institute on Drug Abuse
Cite
APA
Qiu, H., Cook, A. J., & Bobb, J. F. (2024). Evaluating tests for cluster‐randomized trials with few clusters under generalized linear mixed models with covariate adjustment: A simulation study. Statistics in Medicine, 43(2), 201-215. https://doi.org/10.1002/sim.9950
Vancouver
Qiu H, Cook AJ, Bobb JF. Evaluating tests for cluster‐randomized trials with few clusters under generalized linear mixed models with covariate adjustment: A simulation study. Statistics in Medicine. 2024;43(2):201-215. doi:10.1002/sim.9950
BibTeX
@article{Qiu2024,
author = {Qiu, Hongxiang and Cook, Andrea J and Bobb, Jennifer F},
title = {Evaluating tests for cluster‐randomized trials with few clusters under generalized linear mixed models with covariate adjustment: A simulation study},
journal = {Statistics in Medicine},
year = {2024},
volume = {43},
number = {2},
pages = {201--215},
doi = {10.1002/sim.9950}
}
EUVIMED