Clinical trials are often implemented to test the safety and efficacy of particular prevention, diagnostic or therapeutic strategies and require a multi-faceted approach for development, operation and analysis. The focus of Dr. Deborah Donnell, VIDD principal staff scientist and SCHARP statistician, is designing protocols for and analyzing data from HIV prevention clinical trials, with the ultimate goal of identifying preventative measures that can reduce HIV acquisition and transmission. The proper design of a clinical trial is crucial to its success. Because these experiments take years to complete and generate massive amounts of data, scientists and statisticians collaborate to predict whether a certain trial protocol will address a hypothesis and, if so, how to make sense of trial results. This process is complex and depends on excellent statistics.
As Donnell puts it, “the role of a protocol statistician is to make sure the design is truly going to answer the question at the end of the day.”
There are so many details that go into a clinical trial design it is not surprising it takes a “small army of people,” Donnell explains, just for this initial process. Every clinical trial protocol undergoes intense scrutiny to ensure that when the trial is over the results are meaningful and help answer the original question. The decision of who to enroll in a trial itself has many components; how many subjects are needed, which communities have an adequate disease incidence, what are the enrollment criteria and how will the subjects be randomized are just a few examples of the process. In other words, initiation of an HIV clinical trial starts long before study participants begin receiving treatment or placebo interventions.
In August of this year, the HIV Prevention Trials Network (HPTN) published a groundbreaking study in The New England Journal of Medicine that reported a 96% reduction in HIV transmission between serodiscordant couples when early antiretroviral therapy, compared to delayed, was administered. Many VIDD researchers contributed to this trial (HPTN 052), including a large team of SCHARP scientists that provided extensive statistical analyses.
Donnell and her team at SCHARP, in addition to a multitude of partner organizations, are currently working on HPTN 043, a phase III randomized clinical trial taking place in South Africa, Zimbabwe, Tanzania and Thailand, which addresses whether a community-based mobile testing program, compared to the standard testing program, can reduce HIV seroincidence in these communities. This trial, which began in 2003, implemented the intervention for a total of 3 years and resulted in approximately 50,000 specimens for analysis of HIV seropositivity. Results on uptake of HIV testing and detection were published in The Lancet Infectious Diseases in July of this year; the team found that the community-based testing program increased the proportion of clients undergoing HIV testing by approximately 40 percent and almost four times more HIV cases were detected. These data suggest that implementation of this intervention strategy could heighten HIV detection in these communities. Even though the intervention component of this trial has been completed, a substantial amount of work is left to be done. Donnell’s team is currently working on deriving special algorithms for sample evaluation and statistical analyses for the comparison of HIV incidence rates in the trial arms.
What drew Donnell to the Hutchinson Center was the opportunity to do applied work with biologists, and HIV has been the focus of Donnell’s work since joining the Center. When asked what led her to the statistics field, Donnell answered, “I love to make sense out of numbers.”