Sample size estimation for survival outcomes in cluster-randomized studies with small cluster sizes.

Area of Science:

• Biostatistics
• Clinical Trials Methodology
• Epidemiology

Background:

• Cluster-randomized studies are common in public health research.
• Calculating sample size for these studies, especially with many clusters, presents challenges.
• Multivariate survival data requires specific statistical approaches.

Purpose of the Study:

• To develop a method for determining appropriate sample size in cluster-randomized studies.
• To address scenarios with a large number of clusters and few observations per cluster.
• To accommodate multivariate survival data.

Main Methods:

• The proposed method focuses on computing sample size for cluster-randomized designs.
• It is specifically tailored for situations with numerous clusters and limited within-cluster observations.
• The approach assumes knowledge of the marginal bivariate distribution for multivariate survival data.

Main Results:

• A practical method for sample size computation is presented.
• The method is applicable to cluster-randomized studies with specific data structures.
• Discussion on the validity of the marginal bivariate distribution assumption is included.

Conclusions:

• The presented method offers a way to calculate sample sizes for complex cluster-randomized trials.
• It provides a statistical framework for handling multivariate survival data under defined assumptions.
• Further research may explore the implications of relaxing the distributional assumption.