Randomized controlled studies have shown that male circumcision (MC) can reduce the lifetime risk of HIV infection by 53%–60%.1–4 The World Health Organization (WHO) and the Joint United Nations Program on HIV/AIDS recommend that MC be considered as an additional HIV prevention means in countries with low MC rates, high HIV prevalence, and predominantly heterosexual epidemics.5 There are more than 38 million adolescent and adult males in Africa who could benefit from MC for HIV prevention.6,7
In Rwanda, the percentage of all circumcised men is estimated at about 15%.8 The government of Rwanda plans to implement a nationwide voluntary MC program to reach 2 million adult men in 2 years and achieve the national goal of decreasing the HIV incidence rate by 29% by 2025.9 Given the current resource limitations, such a goal would be challenging to meet using the WHO-approved surgical methods in an acceptable time frame. Assuming that 4 surgical MC procedures per hour can be done, which would be 32 procedures in one 8-hour day, then in 1 year with 250 workings, a full-time physician can perform 8000 surgical MCs. To reach 1 million men each year, 125 full-time MC physicians are needed. However, assuming that a physician can allocate only 20% of his/her time to perform MC, 625 part-time MC physicians are needed. In 2008, there were only 395 physicians in Rwanda,10 which is just more than 60% of the MC physician needs. According to the Rwandan Ministry of Health in 2012, there is 1 doctor per 18,000 inhabitants and 1 nurse per 1476 inhabitants.11 Therefore, even with the physicians who are available to surgical MC, the fact that their numbers are so limited further reduces the amount of time they can allocate to the procedure. Thus, the government wanted to explore innovations in MC that would be more suitable for nonphysicians in nonsterile, rural settings, which would minimize the burden on the existing healthcare system that is already strained by life-saving procedures.
Using the WHO framework guidance for Clinical Evaluation of Devices for Adult MC,12 the government of Rwanda organized a research program to explore the safety, efficacy, and practicality of use of the PrePex device for a national MC program. The device works by stopping flow of blood to the distal foreskin, leading to necrosis of the tissue, which is removed with the device after 7 days.
Before conducting this current study, the PrePex safety and efficacy was successfully validated in Rwanda in a safety clinical study that was published in 2011.13 The PrePex superiority over surgical adult MC in both safety and efficacy was validated in a published randomized controlled trial (RCT).14 In preparation for scale-up, this current study assessed the safety and efficacy of the PrePex device when circumcision was performed by lower cadre nurses on healthy adult men scheduled for voluntary circumcision.
This 1-arm, open-label, nonrandomized study was initiated on July 25, 2011, at Kanombe Military Hospital in Kigali, Rwanda, and consisted of 3 following phases: (1) a 3-day nurse training phase (July 25 to July 27), (2) a pilot phase conducted on July 28, and (3) a pivotal phase, which was initiated after the pilot phase safety analysis. The study lasted 3 months, ending on October 31, 2011. Study period per subject was 9 weeks. Procedures were approved by the Rwanda National Ethical Committee and adhered to the ethical standards of the Helsinki Declaration of 1975, as revised in 2000.
There were 10 nurses selected from a local pool of 30 potentially eligible candidates to participate in the study based on the following selection categories: (1) employed at Rwanda Military Hospital, (2) live around Rwanda Military Hospital, (3) capable of understanding English, (4) availability to participate in the study, (5) certified to practice nursing by a recognized Rwanda authority, and (6) never seen or performed PrePex procedures. Selected nurses were approached by the study Principal Investigator to participate in the study, and once they consented, their managers were approached for authorization. Of the 10 nurses recruited, 5 nurses had previous surgical experience (4 of which had previous surgical MC experience), and 5 were regular nurses. Seven nurses were educated at the A2 level, which is the equivalent of 6 years of nursing studies in a secondary school and the level of education of the majority of the health workforce in Rwanda.15 Three nurses were educated at the A1 level, indicating that they had 3 additional years in a nursing institute with a university-level diploma. Nurses with an A0 level of education, which is the highest cadre and indicates a graduate degree, were excluded from this study because the objective focused on the performance of the PrePex procedure by lesser skilled health workers. The nurses had 2–14 years of previous working experience with the average number of years worked by the nurses approximately 7 years.
The study recruited 590 male subjects aged 21–54 years out of a potential participation pool of 766 from the communities near the hospital based on the inclusion criteria, exclusion criteria, and withdrawal criteria (Table 1). Due to a relatively low sample size, mass sensitization methods were not used to recruit the subjects. Rather, mobilization methods were carried out with the study hospital community via the following: word-of-mouth, notification provided to the nearby health centers, a visit to the secondary school located outside the hospital grounds, a preexisting waiting list of subjects from previous PrePex studies, and a surgical MC waiting list at the study hospital. The age range of 21–54 years is in accordance with the national decision for clinical studies in Rwanda.
The 3-day training program was designed to train lower cadre nurses in performing PrePex procedures and was divided into 3 PrePex procedural phases as follows: screening, PrePex placement, and PrePex removal. Nurses underwent theoretical training (including practice on genital models) and clinical training and evaluation on subjects.
Each clinical training phase started with a step-by-step demonstration of the procedure by a physician, who was an expert on performing the PrePex procedure. The training manager selected 5 candidates from the 10 trainees to be the PrePex nurse operators and 5 to be the PrePex nurse assistants in each team based on their knowledge and skills with the genital model. Their roles changed during hands-on training, depending on how they improved their performance and knowledge of the PrePex device and procedure. During the hands-on training, the male subjects were allocated randomly (although not using a formal randomization procedure) to 1 of 5 PrePex teams. Each team was assigned 13–16 subjects to perform the MC procedure. The numbers varied because all PrePex procedures in this group were closely monitored and tutored by the physician who approved a nurse’s participation in the study once he deemed his/her PrePex skills acceptable.
The procedure was carried out in one large, clean, nonsterile inspection room with 5 beds and followed the same protocol of the RCT12 for PrePex placement, removal, and follow-up except that dermal anesthetic cream (5% lidocaine) was used during the placement procedure to avert 1 hour post placement discomfort. Also, after PrePex placement, subjects in this study were discharged in a formal discharge session and given ibuprofen to use at home in case of discomfort. They were instructed to return on day 7 for device removal or in any unsustained situation.
Pilot and Pivotal Phases
Although formal randomization process was not used, the subjects were one-by-one randomly allocated to each PrePex team. A subject next in line was referred to the next team that completed a PrePex procedure and was ready to receive a new subject.
The pilot phase, which included 101 subjects, was designed to ensure the safety of the PrePex procedure when performed by nurses. An interim report validated the procedure’s safety as a precondition to the pivotal phase.
The pivotal phase, which included an additional 417 subjects, proceeded after the interim analysis deemed the nurses’ PrePex procedures safety.
Safety Endpoint and Statistical Considerations
The primary safety endpoint of the adverse event (AE) rate was analyzed based on the total subjects in the study, including those from the training, pilot, and pivotal phases. The training phase data were used to evaluate the training direct measured efficacy endpoint. The samples size calculation (at least 590 evaluable subjects, including the participants in nurses’ training) was driven by the primary safety objective. The assumption was that the total AE rate when nurses perform PrePex may be around 2%, based on the results from the RCT. For an AE rate of 2% to be detected with 95% confidence and for an accuracy level of ±1.5% to be obtained, a minimal sample of 503 subjects was required. An additional sample size of up to 15 subjects per PrePex team for training was added, bringing the total sample size needed to 590 subjects.
Randomization and Statistical Analysis
This is a 1-arm, open-label, nonrandomized study. Statistical analyses were performed using SAS v9.2 or higher (SAS Institute, Cary NC) for the primary and safety endpoints. Binary data such as AE success/fail criteria were presented as a count and percentage together with an exact 95% confidence interval (CI). AE rates were compared descriptively with the reported rates of similar treatments. Continuous data were represented by a mean, standard deviation, median, minimum and maximum together with 95% CIs for the means. Safety endpoints include all AEs related and unrelated to device use.
There was 23 staff assigned to the study, including 10 nurses performing the PrePex procedures, 3 physicians on rotation to oversee the nurses, 8 nurse coordinators, and 2 administrators. There were 666 subjects enrolled in the study, 57 who did not return for their scheduled PrePex placement and 4 duplicate subject files. Of the 605 subjects who went to their procedure, 15 were in fact not eligible upon further examination (Fig. 1). The average age was 25 years among the 590 subjects who underwent the procedure. Five hundred eighty-five were healthy males, and 5 were HIV positive, as verified by their HIV tests. Each of the 5 nurse teams performed the PrePex produce on approximately 15 subjects during training, on 20 subjects during the pilot phase, and on 83 subjects during the pivotal phase, totaling 590 procedures (72 procedures during training and 518 procedures during the pilot and pivotal phases).
At the start of training, every nurse underwent an initial exam that tested their knowledge of the PrePex device and procedure. At the end of training, every nurse passed the posttraining exam (Table 2). During training, 72 subjects underwent the procedure. Based on the clinical assessment, it was decided that all 5 teams successfully passed the PrePex training and were qualified to participate in the study.
The pilot phase enrolled 101 subjects. The pivotal phased began after a safety analysis report was issued. The procedure was performed on 417 subjects. Follow-up lasted 8 weeks.
There was an AE incidence rate of 0.96% (5 AEs on 4 subjects of 518) with a 2-sided 95% exact binomial CI of (0.31% to 2.24%). Therefore, the overall AE rate was less or equal to 2%. Four AEs were device related, and 1 AE was not related to device or procedure (Table 3). Therefore, the device or procedure related AE rate is less or equal to 1.5%. The AE percentage of the 518 procedures was 0.96% (0.31% to 2.24%), which was significantly less than the mean reported by physicians for the surgical MC in the RCT (Table 4) (P = 0.00003).
There were no severe AEs (95% CI: 0.00% to 0.71%), and all AEs were resolved and documented with their severity, relation to device/procedure, intervention, and final outcome. All AEs occurred between days 0 and 7.
Adverse Event 1 and 2
The subject arrived on day 7 for the scheduled removal. The PrePex device was already removed, and the subject had no foreskin. There was mild bleeding from his MC wound, which stopped after 30 seconds of direct pressure. He admitted to removing the device with a razor on the night of day 6 after feeling intense pain accompanied by compression by the inner ring (AE1; Table 3). Although removing the device by himself, he created a wound caused by cutting his foreskin with the razor. The subject felt no pain after the removal, and there was no bleeding. Mild bleeding began when he rode a motorbike to the hospital for the scheduled removal. The pain was related to the device. The bleeding was not related to the PrePex device but to the fact that the subject cut his own foreskin (AE2). The AE was resolved with application of the standard postremoval dressing according to the standard postremoval procedure. No special follow-up other than the regular follow-up plan was required.
Adverse Event 3
The subject arrived to the hospital for an unscheduled visit 1 day after placement complaining that although he could urinate, the flow was disturbed. During examination, it was discovered that the inner ring was placed improperly causing the foreskin to stretch and partially disturb urine flow.
This AE was related to the device. Replacement was performed with a new device. The new elastic ring was placed over the same circumcision line of the original elastic ring, so the previous compressed foreskin was included. The removal on day 7 of the second device went well, and the regular follow-up plan proceeded.
Adverse Event 4
The subject arrived 2 days after placement for an unscheduled visit and explained he showered, had an erection, and touched the device. He felt a slight pain. He noticed that the elastic ring was partially moved out of the inner ring groove. The movement occurred while the subject had an erection and was probably masturbating. The subject was examined to confirm that only the elastic ring partially moved. Localized edema and initial necrotic tissue were observed.
This AE was related to the device. A coinvestigator released the elastic ring completely and replaced it on the previous observed circumcision line to let the process continue until day 7. The inner ring was not replaced or moved from its original location. A new placement ring was used to perform the standard PrePex placement. The subject returned for his scheduled removal visit.
Adverse Event 5
After the device was removed on day 7, the subject called the hospital that night and complained of bleeding from the penis, which began after urinating. The subject went to a local clinic for dressing, but the bleeding did not stop. He returned to the hospital to be examined, which revealed active mild bleeding from a small artery 90-degree angle from the frenulum. One suture was used to stop the bleeding and resolve the AE.
This AE is possibly related to the device. The subject returned the following morning for his next scheduled visit and continued with the regular follow-up plan.
It was assumed that a trained PrePex team should perform the PrePex procedure in a reasonable timeframe, which would improve with experience. Duration of PrePex preparation and procedure were taken from the last 125 subjects., after a nurse had gained experience. The mean placement preparation time was 1 minute 37 seconds (SD: 21 seconds, median: 1 minute 38 seconds), and removal preparation was 1 minute 11 seconds (SD: 8 seconds, median: 9 seconds). The mean placement procedure took 1 minute 8 seconds (SD: 26 seconds, median: 1 minute 2 seconds). The mean removal procedure took 1 minute 43 seconds (SD 47 seconds, median: 1 minute 25 seconds). The mean total procedure time, including preparation, was 4 minutes, 39 seconds (SD: 4 minutes 39 seconds; 1 minute 2 seconds), almost 2 minutes less than the average physician time and 5 times faster than the average time it took a surgical MC procedure in the RTC (Table 4).
Expected Side Effects
Expected side effects are clinical findings which are expected in a normal circumcision procedure. They are not considered an AE because they are within the normal range of a well-healing wound of a circumcision procedure as reported in all circumcision-related literature.
There were 76 expected side effects observed (14.6% of the 518 subjects), which included 18 cases of oozing (3.4% of side effects), 50 cases of localized edema (9.6%), and 8 clear exudates (1.5%). The majority of side effects (57.3%) were observed on day 9, preceded by 18.6% of cases on day 7. The last side effects were observed on day 49. There were none when wearing the device. All side effects resolved naturally in short time.
A possible additional side effect noticed in 57 subjects (11.0%) on day 9 was that the wound had a white caramel-like cover over the exposed granulating tissue. There was no sign of infection, local increase in skin heat next to the wound, pain, or change in color around the wound or deviation from average time to complete healing. The slough was cleaned, but not removed, with simple sterile gauze to absorb clear exudates. The subjects were asked to wash the wound with water and soap at home, as was already requested with follow-ups. The problem was resolved in all subjects in 1–2 weeks. All 57 cases had no additional side effect or AE, and it was believed that the white slough observed in the 57 subjects was a result of the necrotizing tissue process.
None of the expected side effects delayed or complicated the normal MC healing process.
Pain measurements reported by 496 of 518 subjects during placement were an average of 1 on a VAS scale of 10. Thirty minutes postplacement, 123 subjects’ measurements averaged 0.5.
Subjects were given 2 tablets of ibuprofen to use if necessary at home to control possible pain or discomfort. When they came back for the removal, 387 subjects were questioned on pill usage. Three hundred and forty-seven (89.6%) used the pills at home, which effectively reduced their pain.
When subjects returned for device removal, they were asked if they had an erection although wearing the device; 446 (86.1%) reported that they had an erection, and 72 (13.9%) did not reply. Therefore, all subjects who reported on whether they had an erection, indeed, had an erection. The average pain during erection was 2.68, and 42.6% of subjects indicated that they felt pain of 2 at erection.
The average time to complete healing was 33 days postdevice removal on day 40. In the RCT, average healing time was 38 days.14 Healing was calculated for 334 of 518 subjects; 147 subjects (28.3%) did not return for their final visit, and, therefore, cannot be included in the analysis. Two phone calls, 1 week apart, were made to the subjects’ mobile phones to remind them of their final visit. Then, a third call was made to their friend/relative contact they provided at the time of enrollment. In any case, it should be noted that these subjects lost-to-follow-up who missed the final visit were all observed and followed through almost complete healing and showed very good healing progress for several weeks. This leaves no doubt in the investigators’ opinion that they achieved complete healing 5 days after the last visit attended by these subjects.
Confidence in Procedure and Nurses
For the last group of subjects to undergo the PrePex procedure, a questionnaire was filled out by these subjects and by all the nurses to assess the PrePex placement and removal. It should be noted that it was the investigators’ prestudy decision to collect data that best represented the view of the subjects and nurses on the procedure in an optimal manner that eliminated bias from less experience in conducting the procedure. Thus, by surveying the final procedures, the investigators felt it would best represent the procedure being conducted on a regular basis by regular staff. After placement, 123 subjects responded to a procedure questionnaire, 87% found the procedure easy, and 97.5% found painless. Almost all thought that the nurse was confident in the procedure with 68.3% saying very good and 28.4% good. Almost all thought the procedure went quickly (65.9% very fast, 25.2% fast).
After removal, 92.2% of 103 subjects thought the removal was easy (7.7% difficult), and all of them found it painless. Nurses were thought to have been more confident with the removal (78.6% very good, 16.5% good, and 3.9% medium). Only 1% thought the removal was slow. At the end of the study, 96.9% of 130 subjects would recommend a PrePex circumcision.
The nurses responded to similar questionnaires for the final group of procedures. No one thought the placement or removal procedure went badly or was unsafe. Among the operators, 89.1% felt they did well in performing the placement procedure, and 85.6% of the assistants thought they did well. Their confidence increased after the removal; 98.5% of the operators thought they performed the removal well and 97.8% of the assistants.
The primary objective of this study was to assess the safety and efficacy of the nonsurgical circumcision PrePex procedure when performed by nurses. Device safety was assessed by the rate of clinical AEs and device-related incidents attributed to the PrePex device and its procedures. Other measures such as procedure time, expected side effects rate, pain assessment, and time to complete healing were considered for the nurses’ performance.
None of the nurses who participated in this study had previously seen or performed a PrePex procedure. The results demonstrate that the PrePex training course is effective in training nurses who have no experience with PrePex MC in becoming PrePex MC experts. The short total procedure time posttraining (4 minutes and 39 seconds), low AE rate (0.96%), and successful circumcision of the 518 subjects further demonstrate that the training course was effective in training nurses to perform safe, effective, and scalable PrePex MC procedures.
The overall AE rate in the study was 0.96% (95% CI: 0.31% to 2.24%), which is lower than the study hypothesis AE rate of 2% and lower than the RCT—reported 11% AE rate of surgical MC. It should be noted that the investigators believe that the AE results of this study are so much lower than the reported surgical AE rate for Rwanda because there is a lack of standardization of the exact definition of what constitutes an AE. The reported 11% rate (which is 9.5% if you omit 1 unrelated AE) is taken from a comparison study between the PrePex and surgical MC using the dorsal-slit method, in which the AE threshold was very low for both procedures.14 It is possible, therefore, that certain clinical occurrences were defined as AEs in that study, which may not be considered AEs in other studies.
In this current study, only 1 AE related to the device was also associated with the performance of nurse—the erroneous placement of the device that partially disturbed urine flow. Two AEs were related to subject behavior while wearing the device, although 1 was directly device related. The AE rate related to subject behavior could be further reduced with proper counseling and education after the placement procedure. There were no serious AEs, and all AEs resolved with minimal intervention and without any harmful outcome to subjects. It is therefore concluded that the PrePex procedure when performed by nurses is very safe and is safer than adult surgical MC performed by physicians, which strongly supports task-shifting MC from Physicians to nurses.
After gaining experience, the nurses performed the procedure on average at 1 minute and 56 seconds faster than the physicians performing PrePex in the RTC. The PrePex procedure of nurse was also much faster than surgical MC (by 5 times). This further attests to the successful nurse PrePex training program.
Pain was minimal, mainly during erection. Over-the-counter pain medication relieved pain the subjects felt.
The expected side effects did not delay or complicate the normal MC healing process. It was believed that the white slough observed in 57 subjects was a result of the necrotizing tissue process and should not be confused with wound infection because it does not have any accompanying signs of wound infection,16 such as change of skin color, sweet bacterial odor, increase in skin temperature, edema/erythema, exudates, etc. This side effect healed without any intervention, such as antibiotics.
Average healing time was 33 days from removal, as compared with 38 days from removal in the RTC, which further demonstrates that there is no difference in time to complete healing when PrePex MC is done by nurses or physicians and further validates the efficacy of the study.
A possible limitation of the study was not having the performance data of physicians as comparison. That said, the nurses had very good training and with experience were increasingly efficient and confident with the procedure. Another limitation may be that as many as one-third of the subjects did not return for their final follow-up visit, which would have determined their complete healing. Because they made no further contact with the hospital, it was assumed that their healing was complete. The low response rate on the questionnaires by the number of subjects and the number of procedures the nurses reported on is a final limitation of the study which bears impact on the strength of the conclusion without a broader overview of the progress and improvement of nurses and subject satisfaction over time. However, as previously explained with the results of the study, the investigators did not want to subject the nurses performance to subject bias over time.
It should be further clarified that narrow foreskin was included in the rigorous exclusion criteria to ensure a more efficient study flow because of the risk of potential pain from inserting the ring through a narrow foreskin. Nearly 11% of candidates were excluded due to their narrow foreskin, but it is assumed many could have still had the procedure done. A future study is needed to confirm the eligibility of subjects with narrow foreskin.
This study validates the safety and efficacy of task-shifting PrePex MC from physicians to nurses in Rwanda where physicians are limited, not necessarily available to perform surgical MC, and the demand for MC is very high. The procedure was minimally invasive and did not require injected anesthesia, sutures, or sterile settings. PrePex has the potential to facilitate rapid, safe, nonphysician MC scale-up programs for HIV prevention, an imminent need in Sub Saharan Africa where physicians are limited. The recommendation, pending approval by WHO, is that the Government of Rwanda move forward with scaling up the adult MC program in Rwanda and use the Prepex device and nonsurgical procedure where physicians are limited.
Kristen Eckert and Dr Marissa Carter (Strategic Solutions, Inc, Cody, WY) edited the article, and they have also given permission to have their name and information included in the acknowledgment.