Commentary: Estimands in cluster trials: thinking carefully about the target of inferenceand the consequences for analysis choice

Hemming, Karla; Taljaard, Monica

doi:10.1093/ije/dyac174

Cited by 3 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“… 34 , 77 Furthermore, depending on whether the target of inference is the typical cluster or typical individual, in conjunction with the possibility that the impact of the intervention might vary with cluster size (known as ‘informative cluster sizes’), an analysis approach should be chosen that appropriately targets the effect for the typical individual or the typical cluster (below). 79 , 80 …”

Section: Introductionmentioning

confidence: 99%

Key considerations for designing, conducting and analysing a cluster randomized trial

Hemming

Taljaard

2023

International Journal of Epidemiology

Self Cite

View full text Add to dashboard Cite

Not only do cluster randomized trials require a larger sample size than individually randomized trials, they also face many additional complexities. The potential for contamination is the most commonly used justification for using cluster randomization, but the risk of contamination should be carefully weighed against the more serious problem of questionable scientific validity in settings with post-randomization identification or recruitment of participants unblinded to the treatment allocation. In this paper we provide some simple guidelines to help researchers conduct cluster trials in a way that minimizes potential biases and maximizes statistical efficiency. The overarching theme of this guidance is that methods that apply to individually randomized trials rarely apply to cluster randomized trials. We recommend that cluster randomization be only used when necessary—balancing the benefits of cluster randomization with its increased risks of bias and increased sample size. Researchers should also randomize at the lowest possible level—balancing the risks of contamination with ensuring an adequate number of randomization units—as well as exploring other options for statistically efficient designs. Clustering should always be allowed for in the sample size calculation; and the use of restricted randomization (and adjustment in the analysis for covariates used in the randomization) should be considered. Where possible, participants should be recruited before randomizing clusters and, when recruiting (or identifying) participants post-randomization, recruiters should be masked to the allocation. In the analysis, the target of inference should align with the research question, and adjustment for clustering and small sample corrections should be used when the trial includes less than about 40 clusters.

show abstract

Section: Introductionmentioning

confidence: 99%

Key considerations for designing, conducting and analysing a cluster randomized trial

Hemming

Taljaard

2023

International Journal of Epidemiology

Self Cite

View full text Add to dashboard Cite

show abstract

Statistical analysis plan for the NU IMPACT stepped-wedge cluster randomized trial

Scholtens,

Lancki,

Hemming

et al. 2024

Contemporary Clinical Trials

View full text Add to dashboard Cite

Parenting, Child Maltreatment, and Social Disadvantage: A Population-Based Implementation and Evaluation of the Triple P System of Evidence-Based Parenting Support

Sanders,

Clague,

Zając

et al. 2024

Child Maltreat

View full text Add to dashboard Cite

Child Maltreatment (CM) is a widespread public health problem, with adverse outcomes for children, families, and communities. Evidence-based parenting support delivered via a public health approach may be an effective means to prevent CM. The Every Family 2 population trial applied a public health approach to delivering evidence-based parenting support to prevent CM in disadvantaged communities. Using a quasi-experimental design, 64 matched low socioeconomic communities in the Australian states of Queensland and New South Wales received either the full multi-level Triple P system (TPS) of parenting support, or Care as Usual (CAU). Two population indicators of CM, the number of substantiated cases of CM, and the number of notifications of CM to protective services were compared using Welch’s t-test to evaluate intervention effectiveness. After two years of intervention, medium to large effect sizes favoring TPS communities were found for substantiations (d = 0.57, p < .05) and notifications (d = 1.86, p < .001). These findings show the value of the TPS, deployed using a public health approach, in efforts to prevent CM in socially disadvantaged communities. A number of uncontrolled contextual factors are described that may have contributed to some of the differences detected between TPS and CAU communities.

show abstract

Commentary: Estimands in cluster trials: thinking carefully about the target of inferenceand the consequences for analysis choice

Cited by 3 publications

References 10 publications

Key considerations for designing, conducting and analysing a cluster randomized trial

Key considerations for designing, conducting and analysing a cluster randomized trial

Statistical analysis plan for the NU IMPACT stepped-wedge cluster randomized trial

Parenting, Child Maltreatment, and Social Disadvantage: A Population-Based Implementation and Evaluation of the Triple P System of Evidence-Based Parenting Support

Contact Info

Product

Resources

About