Dressler Emily V, Pugh Stephanie L, Gunn Heather J, Unger Joseph M, Zahrieh David M, Snavely Anna C
Department of Biostatistics and Data Science, Wake Forest University School of Medicine, Winston-Salem, NC 27157, United States.
NRG Oncology Statistics and Data Management Center, American College of Radiology, Philadelphia, PA 19103, United States.
J Natl Cancer Inst Monogr. 2025 Mar 1;2025(68):56-64. doi: 10.1093/jncimonographs/lgae053.
Cancer care delivery research trials conducted within the National Cancer Institute (NCI) Community Oncology Research Program (NCORP) routinely implement interventions at the practice or provider level, necessitating the use of cluster randomized controlled trials (cRCTs). The intervention delivery requires cluster-level randomization instead of participant-level, affecting sample size calculation and statistical analyses to incorporate correlation between participants within a practice. Practical challenges exist in the conduct of these cRCTs due to unique trial network infrastructures, including the possibility of unequal participant accrual totals and rates and staggered study initiation by clusters, potentially with differences between randomized arms. Execution of cRCT designs can be complex, ie, if some clusters do not accrue participants, unintended cluster-level crossover occurs, how best to identify appropriate cluster-level stratification, timing of randomization, and multilevel eligibility criteria considerations. This article shares lessons learned with potential mitigation strategies from 3 NCORP cRCTs.
J Natl Cancer Inst Monogr. 2025-3-1
Cochrane Database Syst Rev. 2021-9-6
J Natl Cancer Inst Monogr. 2025-3-1
Cochrane Database Syst Rev. 2022-1-17
Cochrane Database Syst Rev. 2018-1-16
Cochrane Database Syst Rev. 2025-3-25
Cochrane Database Syst Rev. 2008-7-16
Cochrane Database Syst Rev. 2011-9-7
Cochrane Database Syst Rev. 2016-10-13
Cochrane Database Syst Rev. 2024-12-12
Int J Epidemiol. 2023-10-5