‘Shocking’ rates of adverse events seen with traditional and medical circumcision in Kenya

A rate of adverse events, infections and delayed healing described as ‘shocking’ and ‘unacceptable’ by the investigators has been revealed by a survey of both traditional and medically performed circumcisions amongst a traditionally-circumcised ethnic group in Kenya.

The study’s authors, who include Robert Bailey, the principal investigator of one of the three randomised controlled studies (RCTs) of circumcision as an HIV prevention method, urge training for practitioners, the provision of low-cost kits of circumcision materials, and the integration of circumcision into a full complement of HIV prevention and reproductive health services. They urge the adoption of a certification process for traditional and medical practitioners.

The rate of adverse events observed – 35% in traditional circumcision and 18% in medically performed ones – are an order of magnitude above those seen in the RCTs of circumcision, and in medically supervised circumcision in the developed world. Six per cent of operations resulted in adverse events described as permanent and irreversible.

As well as exacting significant levels of morbidity in the young male population, the authors say that poorly-performed circumcision, although often the result of lack of equipment and money, may end up costing families more than properly supervised circumcision would. It also represents a significant HIV risk in itself as 6.3% of the young men circumcised traditionally and 3% of those circumcised medically had already engaged in sex a mean of 60 days after circumcision even though in 24% of the traditional cases and 19% of medical cases the circumcision wound had still not healed properly by this time.

The contrast with the medically-supervised circumcision performed in the RCTs is most starkly highlighted by the fact that in the RCTs all but 4% of circumcision wounds had healed by 30 days after the operation whereas in a directly-observed subset of 12 traditional and 12 medical procedures in this survey, no wound had properly healed by this time.

The survey results were a result of interviews with 1007 boys and young men who had undergone circumcision in the Bungoma district of western Kenya, 445 of them traditionally and 562 with some kind of medical supervision, which meant in a hospital, health centre or private office “by anyone considered by the participant to be a clinician”.

This area is predominantly inhabited by the Bukusu ethnic group whose men are almost universally circumcised at adolescence. As indicated above, the first 24 procedures were directly observed by the investigators and when it became clear that there was a very high rate of adverse events, the last 298 men and boys in the survey to be interviewed were also given a penile examination, an average of 45-90 days after circumcision.

The median age for circumcision was 14 with the medically-supervised circumcisions being performed on a younger age group (90% under 16 for medical circumcision compared with 66% for traditional circumcision). About 40% had engaged in sex before being circumcised, at a median age of 15.

The main difference between medical and traditional circumcision, in the 24 operations directly observed, was that all medical ones featured some form of local or general anaesthesia compared with none in the traditional circumcisions, and in 75% of cases the wound was sutured, though often inadequately, whereas in traditional circumcision it was just left to heal. Not surprisingly bleeding was a common adverse event with 8% of medical circumcisions featuring bleeding described as “profuse, requiring IV fluids”.

In one in three traditional circumcisions observed and one in six of the medical ones, incomplete foreskin was removed and re-circumcision was required.

Infections, ranging from mild swelling and redness to life-threatening necrosis, were very common and – alarmingly – even frequent in the medical settings (50% versus 42%). Antibiotics were used in a high proportion of cases (50% medical and 42% traditional); the most common was a brand of talcum powder containing penicillin which was applied to the wound and then bandaged. The authors comment: “Whether it prevented infections we cannot be sure, but it tended to…delay healing and result in thick scarring.”

Permanent adverse events included torsion (bending) of the penis, injuries to the glans, loss of penile sensitivity caused by scarring and erectile dysfunction.

Among the 298 boys and men examined post-operation only 21% of traditional and 10% of medical circumcisions had fully healed an average of 45-89 days after the operation.

The cost of a traditional circumcision was about 345 Kenyan shillings ($5.00 or £2.80), though additional payments were often required in cash and in kind when there were complications and may have ended up costing more than medical procedures.

Private medical facilities produced an adverse event rate of 22.5% compared with 11% in government-run ones. The most important lack of medical equipment was lack of autoclaves for sterilisation, with instruments being sterilised by boiling, and lack of sutures. In many cases, especially in traditional circumcision, instruments were not sterilised between several operations, creating an infection and HIV transmission risk, and lack of sharp scalpels resulted in ragged wounds in a quarter of traditionally circumcised and 17% of medically circumcised boys/men.

The investigators also interviewed 21 traditional and 20 medical circumcisers. Traditional circumcisers had performed more operations, with nine (versus five medical) having performed over one hundred. When asked if they felt they were adequately trained to perform circumcision, only one (medical) practitioner said “no” but when asked if they would like further training 50% from each group felt they would profit from it.

No death was reported as directly due to circumcision though the authors comment on one case who, if not taken to the district hospital by the investigators, “would very likely have died without our intervention.”

The authors comment: “The levels of morbidity and mortality from circumstances documented as occurring in this study community are unacceptable,” and they add that there is sufficient anecdotal evidence to indicate that Bungoma is not unique, especially in east and southern Africa where circumcision is performed on adolescents rather than infants.

They say: “Our results…should serve as an alarm to ministries of health and the international health community that focus cannot only be on areas where circumcision is low…it must address the safety of circumcision in areas where it is already widely practised.

“If the practices in these communities continue to be largely ignored,” they conclude, “the gains to be achieved by promotion and provision of circumcision for HIV prevention may well be undermined by further accounts of unnecessary suffering.”