Novel strategies implemented to ensure high participant retention rates in a community based HIV prevention effectiveness trial in South Africa and Zimbabwe

Background: The identification of new HIV prevention methods that women can initiate themselves are urgently needed, particularly in high prevalence settings. HIV prevention trials must be designed with large sample sizes and/or substantial follow-up periods to ensure enough statistical power to measure product effectiveness. This paper describes the attendance rates of the Methods for Improving Reproductive Health in Africa (MIRA) trial, reasons for missed visits, and strategies used to retain participants; and examines demographic and behavioural predictors of retaining women. Methods: HIV negative, sexually active females were enrolled into the MIRA trial in Zimbabwe and South Africa. Once enrolled, women were expected to visit the clinic at 2 weeks and quarterly thereafter for 12 to 24 months. Tabulations of visit-specific retention rates are presented, along with a descriptive summary of retention strategies established prior to and in response to challenges incurred during implementation. Both univariate and multivariate logistic regression models were created in STATA to examine predictors of being retained vs. lost-to-follow-up. Results: At the three sites, the final retention rates were 94%, 93% and 89% for Zimbabwe, Durban and Johannesburg, respectively. This was achievable through intensive outreach efforts toward the latter part of the trial and a commitment from all staff. Each site implemented several retention strategies. Conclusion: The high retention rates were achievable in this trial through added staff efforts and resources. Community involvement was also crucial to achieve these rates. Retention of trial participants should be considered during trial design and implemented from the onset. © 2009 Elsevier Inc. All rights reserved.