Comparing cluster-level dynamic treatment regimens using sequential, multiple assignment, randomized trials: Regression estimation and sample size considerations

TL;DR

This paper introduces a regression-based method and sample size calculations for comparing cluster-level dynamic treatment regimens in sequential, multiple assignment, randomized trials, enhancing analysis and design in cluster-level interventions.

Contribution

It proposes a weighted least squares regression approach for analyzing cluster-randomized SMARTs and derives sample size calculators for comparing treatment regimens.

Findings

Regression approach allows covariate adjustment in cluster SMARTs.

Sample size formulas for continuous outcomes in cluster SMARTs.

Application to psychiatry trial demonstrates practical utility.

Abstract

Cluster-level dynamic treatment regimens can be used to guide sequential, intervention or treatment decision-making at the cluster level in order to improve outcomes at the individual or patient-level. In a cluster-level DTR, the intervention or treatment is potentially adapted and re-adapted over time based on changes in the cluster that could be impacted by prior intervention, including based on aggregate measures of the individuals or patients that comprise it. Cluster-randomized sequential multiple assignment randomized trials (SMARTs) can be used to answer multiple open questions preventing scientists from developing high-quality cluster-level DTRs. In a cluster-randomized SMART, sequential randomizations occur at the cluster level and outcomes are at the individual level. This manuscript makes two contributions to the design and analysis of cluster-randomized SMARTs: First, a weighted least squares regression approach is proposed for comparing the mean of a patient-level outcome between the cluster-level DTRs embedded in a SMART. The regression approach facilitates the use of baseline covariates which is often critical in the analysis of cluster-level trials. Second, sample size calculators are derived for two common cluster-randomized SMART designs for use when the primary aim is a between-DTR comparison of the mean of a continuous patient-level outcome. The methods are motivated by the Adaptive Implementation of Effective Programs Trial, which is, to our knowledge, the first-ever cluster-randomized SMART in psychiatry.