Randomized Control Trial of Shang Ring Versus Forceps-guided Adult Male Circumcision in Southwestern Uganda.

Overview

Adult male circumcision (AMC) can reduce the incidence of HIV transmission by 40-60% in sub-Saharan Africa (SSA). The World Health Organisation and Joint United Nations Programme on HIV/AIDS advocate AMC for the reduction of HIV transmission in heterosexual men, and feasibility studies have demonstrated that AMC programmes can be effectively delivered to hyperendemic communities, such as those in SSA. Despite these recommendations, the potential effects of AMC have been attenuated by low uptake, with only 5% of all men who could benefit undergoing surgery to date. Traditional, time-consuming methods of AMC that require formal surgical training limit uptake in low-income countries. Training and implementation by non-physician health providers is dampened by a need to perform approximately 100 procedures to achieve satisfactory operative time and adverse effect rates. The investigators performed a randomised controlled effectiveness study to compare the shang ring (SR) with standard forceps-guided (FG)-AMC in a publicly funded regional referral hospital with a locally trained surgeon in Mbarara, western Uganda.

Male undergoing elective AMC were recruited from the surgical outpatient department of Mbarara Regional Referral Hospital, Uganda. Patient age, marital status, smoking history and motive to seek AMC were recorded for each participant. The investigators performed block randomization by study day. Consenting participants selected an opaque envelope from a box for randomization to SR or FG groups. All AMC procedures were performed by a locally-trained surgeon who had performed over 100 prior FG procedures but who had no prior experience of the SR procedure. A study surgeon performed all procedures in the hospital operating theatres. In both study groups, participants were cleaned with povidone iodine solution and draped in a sterile fashion. Local anesthesia was administered to the dorsal penile nerve and penile ring blocks using 3mg/kg of 1% lidocaine. The surgeon measured participants in the SR group to determine ring size. Patients for whom a suitable ring size was not available crossed over to the FG group, but remained in the SR group for intention to treat (ITT) analyses. An assistant measured operative time using a stopwatch. For the SR group, timing began as the ring began to be fitted, and ceased when the foreskin had been excised. For the FG group, timing began at the first incision and ceased when the final suture was completed. Patients were discharged on the day of procedure, provided with sterile gauze dressings, and informed to change them twice daily until the first follow-up appointment. All pre- and post-operative assessments were performed by the lead investigator at one hour after the procedure, and on the 3rd, 7th, 14th and 21st post-operative days. The investigators measured incidence rates of all outcomes and estimated relative risk for the SR versus FG groups by fitting Poisson regression models with robust errors. The investigators compared time to healing with Kaplan-Meier survival curves and measured for a difference between groups using the likelihood ratio test. All analyses were initially performed using ITT analysis, and were repeated with an as-treated analysis by reallocating the 7 participants who crossed into the FG group.