The study was conducted in Rakai, Uganda. Men eligible for this retrospective cohort study were uncircumcised men aged 15–49 years who were initially HIV-negative participants in the Rakai Community Cohort Study (RCCS) described below. These men were subsequently enrolled into the randomized trials of male circumcision and had measurement of their foreskin surface area following surgery. There were 965 initially HIV-negative men under surveillance in the RCCS for a maximum of 4 years prior to surgery. We determined HIV acquisition in these men and assessed the association between foreskin size measured after surgery and the incidence of HIV acquisition while under surveillance prior to male circumcision.
The RCCS is an open cohort with an annual follow-up of approximately 12 000 adults residing in Rakai district who provided written informed consent at enrollment and at each follow-up visit. At each visit, participants provided information on social–demographic characteristics and sexual behaviors. Samples including venous blood for HIV testing were collected. Voluntary HIV counseling and testing, health education on HIV prevention strategies [abstinence, be faithful and use condoms (ABC)] and free or subsidized condoms were provided at every visit.
We conducted two randomized trials of male circumcision for HIV prevention in men enrolled between 2003 and 2006. The trials have been described elsewhere [8
]. Men in both trials were randomized to either receive male circumcision immediately or have male circumcision delayed for 2 years. Informed consent and a questionnaire were administered at baseline and follow-up at 6, 12 and 24 months. Thus, trial participants who had previously been enrolled in the Rakai cohort study afforded an opportunity for assessment of HIV incidence during cohort surveillance in relation to the foreskin surface area measured after male circumcision.
Both the RCCS and the male circumcision trials were approved by four institutional review boards: the Science and Ethics Committee of the Uganda Virus Research Institute (Uganda), the National Council for Science and Technology (Uganda), the Committee for Human Research at Johns Hopkins University Bloomberg School of Public Health (USA) and the Western Institutional Review Board (USA). The trials were registered with Clinical.Trials.Gov
numbers NCT00425984 and NCT00124878.
HIV status was determined using two enzyme immunoassays (EIA): Veronostica HIV-1 (Organon Teknika, Charlotte, North Carolina, USA) and Welcozyme HIV-1 and HIV-2 (Murex Diagnostic, Dartford, UK). Concordant positive EIA tests were interpreted as HIV positive, and concordant negative EIA tests were interpreted as HIV negative. Discordant EIA tests and all incident HIV cases were confirmed by western blot (Calypte Biomedical Corp., Rockville, Mississippi, USA). Men who tested HIV positive were referred to the Rakai President’s Emergency Plan for AIDS Relief (PEPFAR)-funded HIV care and treatment program.
Male circumcision was conducted by trained physicians using the sleeve circumcision method, which has been described elsewhere [8
]. Measurement of foreskin surface area was done by physicians immediately after completion of male circumcision. The foreskin was placed on a flat surface, and mild tension was applied at all four corners of the foreskin using artery forceps to straighten the edges. A tape calibrated in centimeters was used to measure the length and width of the foreskin at the midpoint, and results were recorded.
Person-time for estimation of HIV incidence was accrued from the date of enrollment into the Rakai cohort until the date of seroconversion or male circumcision. The timing of HIV infection was estimated at the midpoint between the first HIV-positive and last HIV-negative serologic test. There were no losses to follow-up, as we purposefully selected initially HIV-negative men who were under continuous observation until time of surgery.
Foreskin surface area in square centimeters (cm2) was calculated by multiplying the length by the width of the foreskin. Mean and median foreskin surface areas were stratified by age and by HIVacquisition or persistent HIV-negative status. The t-test was used to assess differences in the mean foreskin surface areas by age and between men who seroconverted and those who remained uninfected. Foreskin surface area was also categorized into quartiles [ ≤25% (7.0–26.3 cm2), 26–50% (26.4–35.0 cm2), 51–75% (35.1–45.5 cm2) and >75% (45.6–99.8 cm2)]. We then assessed HIV incidence by quartiles of foreskin surface area. Incidence rate ratios (IRRs) and 95% confidence intervals (95% CIs) of HIV acquisition associated with quartiles of foreskin surface area were estimated by Poisson multiple regression after adjustment for age, education, religion, number of sex partners and condom use.
The analyses for this study used Stata version 8.0 (StataCorp Inc., College Station, Texas, USA).