Tshimanga 2016.

Study characteristics
Methods	Design: randomized, open‐label RCT conducted in Zimbabwe and Malawi. Compared the performance of the PrePex device to FG surgical circumcision. 240 men randomized. Ethical approval: Medical Research Council of Zimbabwe's (MRCZ) and Battelle's Institutional Review Board approved this study (MRCZ: approval 5 September 2011, MRCZ/A/1628; Battelle: approval October 2011, IRB000284). All interviewers and nurses attended study protocol training sessions, as well as a good clinical practice (GCP) courses offered jointly by the Research Support Centre, College of Medicine of Malawi and MRCZ Consent: participants provided written formal consent and confirmed their understanding of their participation. Duration of study: 21 November 2011 to 18 January 2012
Participants	Inclusion criteria: males aged ≥ 18 years, uncircumcised and agreement to be circumcised, HIV‐negative, ability to understand study procedures and requirements, and freely give informed consent for participation in this study, agreement to abstain from sexual intercourse and directly rubbing circumcised area (up to 70 days after procedure) and until the end of the follow‐up, and to return to the healthcare facility for follow‐up visits (or as instructed). Exclusion criteria: active genital infection, anatomic abnormality, or other condition, which in the opinion of the investigator prevented the participant from undergoing a circumcision. Men who had phimosis, paraphimosis, warts under the prepuce, torn or tight frenulum, narrow prepuce, hypospadias, epispadias, known bleeding/coagulation abnormality, uncontrolled diabetes, did not agree to anonymous video and photographs of the procedure and follow‐up visits, HIV‐positive, refusal to take HIV test and ages < 18 years. Baseline characteristics: no differences in baseline characteristics between groups. Mean age: 29.1 (SD 9.0) years for the PrePex group and 27.6 (SD 7.6) years for surgical group (difference not statistically significant). 49.2% were single and 73% were employed.
Interventions	Providers: doctors and nurses trained by PrePex Masters from Rwanda. Intervention group (PrePex): 160 participants. Clean and non‐sterile preparation followed by dermal anesthetic cream (5% lidocaine). The device was then applied. Follow‐up: 7 days for device removal, 2 days later for dressing removal and weekly thereafter for 6 weeks. Control group (surgical circumcision): 80 participants. Surgical circumcisions were done under local anesthesia using the FG method consistent with the national guidelines adapted from the WHO Manual for Male Circumcision under Local Anesthesia. Follow‐up: assessed on the 3rd, 7th, 14th, 21st, 42nd postoperative days
Outcomes	Primary outcomes: procedure operative time Secondary outcomes: pain using VAS safety – adverse events (bleeding; hematoma, edema, infection, device‐related adverse events) satisfaction (participant and partner filled out questionnaire using a 5‐point scale) time to complete healing based on treating physician's objective wound assessment using specified parameters difficulty during procedure based on reports cost of procedure
Funding sources	United Nations Population Fund
Declarations of interest	The authors declared that they had no competing interests.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomization sequence generated using Excel; low risk of selection bias.
Allocation concealment (selection bias)	Unclear risk	Allocation concealment in the trial used numbered marbles picked out of a box. Participants picked 1 numbered marble from the box, which was then set aside. The participant was assigned that number. The nurse verified which study arm the participant was assigned to according to the randomization list. If a participant opted out of the assigned arm, he was excluded from the study. His marble was put back in the box. It is possible that this approach could result in changes to allocation. Unclear risk of selection bias.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Blinding of participants and personnel was not possible with this surgical intervention and it was, therefore, unclear whether this may have impacted the performance of either group.
Blinding of outcome assessment (detection bias) Subjective outcomes (severe adverse events, moderate adverse events, minor adverse events, pain, patient satisfaction)	High risk	There was no mention of independent outcome assessment. The lack of blinding of subjective outcome assessment could confer a high risk of detection bias. In addition, each arm received different follow‐up visit days which could also result in differential assessment of the outcomes. High potential risk of detection bias.
Blinding of outcome assessment (detection bias) Objective outcomes (operative time)	Low risk	As the operative time was an objective outcome, the lack of blinding was not considered a risk of detection bias.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Overall there was low attrition between groups with no differences in numbers lost to follow‐up; low risk of attrition bias.
Selective reporting (reporting bias)	High risk	Study outline in ClinicalTrials.gov (NCT01956370). The primary outcome was unchanged; however, reporting on pain was not explicitly mentioned in the registry record. Therefore, indicated a potential reporting bias.
Other bias	Low risk	No other biases identified.