| Abstract: |
The Radiation Therapy Oncology Group (RTOG) designed a random sampling process and observed its influence upon radiotherapy review mechanisms in cooperative group clinical trials. The method of sampling cases for review was modeled from sampling techniques commonly used in pharmaceutical quality assurance programs, and applied to the initial (on-study) review of protocol cases. 'In control' (IC) status is defined for a given facility as the ability to meet minimum compliance standards. Upon achieving IC status, activation of the sampling process was linked to the rate of continued patient accrual for each participating institution in a given protocol. The sampling design specified that ≥ 30% cases not in compliance would be detected with 80% power. A total of 458 cases was analyzed for initial review findings in four RTOG Phase III protocols. Initial review findings were compared with retrospective (final) review results. Of the 458 cases analyzed, 370 underwent initial review at on-study, while 88 did not require review as they were enrolled from institutions that had demonstrated protocol compliance. In the group that had both initial and final review, 345 370 (93%) were found to have followed the protocol or had a minor variation. Of the exempted cases, 79 88 (90%) were found to be per protocol or a minor variant. The sampling process proved itself to be cost-effective and resulted in a noticeable reduction in the workload, thus providing an improved approach to resource allocation for the group. Continued evaluation of the sampling mechanism is appropriate as study designs and participants vary over time, and as more data become available to study. Further investigation of individual protocol compliance is appropriate to identify problems specific to new trial investigations. © 1995. |
