Sample Size Calculations for Clustered and Longitudinal Outcomes in Clinical Research

Chul Ahn · Moonseoung Heo · Song Zhang

Dec 2014 · Chapman & Hall/CRC Biostatistics Series Book 71 · CRC Press

About this ebook

This book explains how to determine sample size for studies with correlated outcomes, which are widely implemented in medical, epidemiological, and behavioral studies. For clustered studies, the authors provide sample size formulas that account for variable cluster sizes and within-cluster correlation. For longitudinal studies, they present sample size formulas that account for within-subject correlation among repeated measurements and various missing data patterns. For multiple levels of clustering, the authors describe how randomization impacts trial administration, analysis, and sample size requirement.

About the author

Chul Ahn, PhD, is a professor in the Department of Clinical Sciences and the cancer center associate director for biostatistics and bioinformatics in the Simmons Comprehensive Cancer Center at the University of Texas Southwestern Medical Center. He is also director of Biostatistics and Research Design for the NIH-sponsored Clinical and Translational Science Award (CTSA). He has published more than 370 peer-reviewed papers addressing the design and analysis of clinical trials and epidemiological studies as well as the evaluation of repeated measurements and correlated data.

Moonseong Heo, PhD, is a professor in the Department of Epidemiology & Population Health at the Albert Einstein College of Medicine. His research includes sample size determinations for clinical trials, meta-analysis, longitudinal data analysis applying mixed-effects models, handling attrition problems in clinical trials data, and epidemiology in the fields of obesity and psychiatry.

Song Zhang, PhD, is an associate professor in the Department of Clinical Sciences at the University of Texas Southwestern Medical Center. He has extensive experience in the design of clinical trials with correlated outcomes, addressing challenges that involve different correlation structures, missing data patterns, financial constraints, and historical controls. He is also interested in Bayesian statistical methods and their application in longitudinal and survival data analysis, high-throughput data analysis, disease mapping, adaptive design for clinical trials, and missing data imputation.