DSMB Review for Cluster Randomized Trials of HIV Prevention

• Activity Number: 235
• Type: Topic Contributed
• Date/Time: Monday, August 5, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Victor De Gruttola*+ and Quanhong Lei and Rui Wang and Max Essex
• Companies: Harvard School of Public Health and Harvard School of Public Health and Harvard School of Public Health and Harvard School of Public Health
• Keywords: Cluster Randomized Trials ; DSMB review ; futility ; Epidemic Models

Abstract

The Botswana Harvard Partnership is developing a protocol for a cluster randomized study to investigate the population-level impact on HIV incidence of a combination of HIV prevention interventions compared with standard of care (SOC) in 30 villages in Botswana. The study will continue until all sites have 3 years of follow-up. Because it will take about 1 year to fully enroll all communities, the study will require at least 5 reviews, DSMB 0-4, which take place at baseline and at the ends of years 1, 2, 3 and 4. DSMB 1-2 will evaluate the extent of intervention rollout and safety. At DSMB 3, the DSMB can recommend stopping for efficacy or futility, or extending the study by 1 year, should there be evidence of poor intervention coverage or lower than anticipated incidence in the SOC arm. Estimates of conditional power based on simulation studies will be provided to the DSMB; such studies have demonstrated the plausibility of conditional power being less than 70% at DSMB 3 but exceeding 80% if the study is extended by 1 year. Such an extension would trigger a 2nd interim review for efficacy and futility at DSMB 4.