.png){kind=icon}
Return
Optimal Sample Allocation for Multisite Partially Nested Individual-Randomized Trials
.png){kind=icon}
OA
AI
DOI：10.1080/00220973.2026.2719484.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
We examine the efficient and strategic design and analysis of multisite partially nested randomized trials. Multisite partially nested randomized trials arise when, for example, studies use individual randomization within sites to interventions that introduce intermediate clustering that is not present in a control condition. We begin by presenting a conceptual framework for multisite partially nested designs and illustrating it with a working example. We then examine estimation of the average treatment effect, derive power formulas for models with and without covariates, and develop optimal sampling expressions under both a conventional equal cost approach and a more flexible approach that allows unit costs to differ across treatment conditions. We conclude by examining the scale of well-powered studies under common conditions and the sensitivity of the optimal sampling plans to the misspecification of assumed scenarios. The results provide estimation and design principles that guide researchers in the efficient design and analysis of multisite partially nested randomized trials.
Keywords:
Multisite clustered randomized trials
optimal allocation
partially nested design
sample-size determination
Journal
J
IF:
1.7
Papers:
125
Citations:
2.8K
Organization
Cited Papers
No cited papers available