DESIGN, ANALYSIS, AND REPORTING OF PILOT STUDIES IN HIV

Abstract

Pilot studies, a subset of feasibility studies, are essential in determining the feasibility of a larger study. This is especially true when targeting populations that are difficult to recruit, such as people with HIV. Designing high quality pilot studies can help limit waste by informing researchers how to proceed.

We conducted a meta-epidemiological review of pilot studies in the HIV literature published until November 25, 2020 using Medline, Embase and Cochrane Controlled Register of Trials (CENTRAL). We extracted bibliometric information, including the region and income of the country where the study was conducted, study design, using the pilot label, source of funding, nature of intervention, whether feasibility was the primary objective, progression criteria, protocol registration and sample size estimation. We used descriptive analysis to evaluate how pilot studies are designed and conducted, the outcomes assessed and how are they defined.

Our search retrieved 10,597 studies, of which 248 were included in our final review. The number of pilot studies has increased with time, with 25, 55, and 44 HIV studies published in 2018, 2019, and 2020, respectively. We found that 128 studies (70.39%) used the pilot or feasibility labels in their title, however 20.31% used these titles interchangeably. 5 studies in this review included progression criteria, all of which were published in 2020. Sample size estimation was only found in 59 studies (23.9%).

Pilot studies in the HIV literature are mislabeled. Sample size estimations are seldom included, and progression criteria are used. Formal guidance on the design and reporting of pilot studies in the HIV literature is necessary.