Objectives: Participants who obtain an HIV test outside of an HIV vaccine efficacy trial could potentially unblind themselves which could result in differential behavior change and loss to follow-up based on assignment status. In a reanalysis of the VaxGen VAX004 data, the objectives were to determine: 1) the proportion of participants who were tested for HIV outside of the study (despite instructions not to do this) and reasons why; 2) demographic and risk factors associated with reported testing outside of the study; and 3) if outside testing was related to participant loss to follow-up.
Methods: Analyses were restricted to men who have sex with men (MSM) who completed a survey at one or more annual visits in a randomized, double-blind, placebo-controlled efficacy trial of a bivalent rgp 120 vaccine conducted from 1998-2002. A generalized linear mixture model assessed associations with outside testing.
Results: Despite instructions to the contrary, 16.9% (791/4670) of MSM reported being tested for HIV outside of the study, with the top two reasons being a) medical provider request (28.1%) and b) insurance requirement (17.1%). Increased odds of self-reported outside testing was associated with site location, reporting one or more sexually transmitted infections (STIs), joining the trial because of the belief that participation might confer some protection against HIV infection, engaging in unprotected anal sex, and being lost to follow-up. Decreased odds of self-reported outside testing was associated with perceived study arm assignment to vaccine or uncertainty about study arm assignment compared to placebo.
Conclusions: To avoid biases such as differential risk behavior and loss to follow-up based on perceived assignment status, initiating additional procedures to reduce the likelihood of outside testing will be important to assure the validity of future study results.