The influence of male circumcision for HIV prevention on sexual behaviour among traditionally circumcised men in Cape Town,South Africa

Abstract

We examined the relationship between HIV prevention beliefs related to male circumcision and sexual behaviour/sexually transmitted infection (STI) acquisition among traditionally circumcised men in Cape Town, South Africa. HIV-negative men (n = 304), circumcised for cultural/religious reasons, attending a health clinic in Cape Town, South Africa, completed cross-sectional surveys. Generalized linear models were used to analyse the relationships between unprotected vaginal sex acts, number of female sexual partners, STI diagnoses and male circumcision-related beliefs and risk perceptions. Men who were aware that circumcision offers protection against HIV (relative risk [RR] = 1.19, 95% confidence interval [CI] = 1.06-1.32, P < 0.01), endorsed risk compensation related to male circumcision (RR = 1.15, 95% CI = 1.11-1.12, P < 0.01) and perceived lower risk of HIV infection when circumcised (RR = 1.08, 95% CI = 1.04-1.12, P < 0.01) were more likely to report unprotected vaginal sex acts. Similar patterns were also identified when predicting number of female sexual partners. Men who were more likely to endorse risk compensation related to male circumcision were also more likely to be diagnosed with a chronic STI (odds ratio [OR] = 1.64, 95% CI = 1.06-2.53, P < 0.05). Our findings suggest that we must not overlook the effects of beliefs towards male circumcision for HIV prevention among men traditionally circumcised; doing so may undermine current efforts to reduce HIV transmission through male circumcision.

Keywords

HIV men prevention sexual behaviour sexually transmitted infections male circumcision South Africa risk compensation risk perceptions

Introduction

The evidence for the partial efficacy of male circumcision providing protection against female-to-male HIV transmission has been demonstrated in epidemiological studies,^1-3 three randomized controlled trials (RCTs),^4-6 and cross-sectional and prospective studies.⁷ There are at least 10 countries in sub-Saharan Africa actively involved in different stages of rollout of male circumcision, with Zimbabwe, Zambia, South Africa, Swaziland and Kenya currently offering it with varying degrees of accessibility.⁸ Although male circumcision for HIV prevention has demonstrated partial efficacy, concerns remain regarding its effectiveness in ecologically-valid settings,⁹ with some evidence suggesting that we should be cautious with regard to the external validity of male circumcision for HIV prevention.¹⁰ Changes in behaviour that could potentially result from new and emerging biomedical HIV prevention technologies need to be carefully monitored and understood.¹¹

Taking into consideration risk compensation or behavioural adaptation to new and evolving information about the protective benefits of male circumcision should be a priority.^12,13 If individuals engage in greater risk taking in response to a perceived lower risk for HIV, then benefits provided by a biomedical prevention technology could be compromised.¹⁴ As information about the preventive benefits of male circumcision becomes widely known, it is possible for behaviour to change in response to dissemination of this information.¹² Risk compensation is particularly relevant to male circumcision due to the partial protection that it offers. In order to effectively promote male circumcision, men need to understand that they are seeking this procedure to lower their likelihood of HIV infection, yet the same messages regarding the need for condom use must remain.

In the three completed male circumcision RCTs and in one other longitudinal study, on the whole, it appears that both circumcised and uncircumcised men reduced their risk behaviours during the course of study follow-ups^4,5,15-18 with a possible difference in the rate at which risk behaviour decreased: a sharper decrease among uncircumcised men in response to persistent risk reduction counselling. However, how risk compensation unfolds outside of a RCT – i.e. in the absence of monthly risk reduction counselling – remains to be seen, in particular among traditionally circumcised men. Given that risk reduction counselling during widespread rollout of male circumcision will not be provided to the extent it was in RCTs, conclusions about the effects of risk compensation cannot be made based on RCTs alone men.

Little is known about how awareness of male circumcision for HIV prevention affects risk behaviour among men who have been circumcised for reasons outside of HIV prevention. Understanding these implications is not trivial, as globally an estimated 30% of men are circumcised for religious/cultural reasons.¹⁹ Priority should be given to understanding how these men may be affected by HIV prevention messages related to male circumcision. This idea is particularly important because these men receive no risk reduction counselling specific to male circumcision. Furthermore, due to variations in traditional circumcision procedures versus those procedures performed in RCTs, it is unclear if men who are traditionally circumcised would experience the same level of protection.^18,20 Major variations in practices regarding the procedure of traditional versus medical circumcision include traditional circumcision as generally being performed by individuals without formal medical training and without the use of analgesics or antiseptics, and many men can be circumcised with the same instrument without sterilization between procedures. Some evidence suggests that more often in traditional circumcision the foreskin remains partially intact potentially reducing the effectiveness of male circumcision for prevention.^20,21

In South Africa, 35% of men are circumcised for cultural/religious reasons.¹⁹ Circumcision in this country has a long history of being considered as a rite of passage into manhood and is socially significant in that partaking in it confers respect among fellow community members. This practice generally occurs among boys/men between the ages of 15 and 25 years attending circumcision school and may involve ritual sacrifices, seclusion of boys/men undergoing circumcision, and celebrations of entering into manhood. Although the practice of circumcision as a rite of passage is deeply rooted in Xhosa tradition, the meaning of this tradition has changed over the years. According to the work conducted by Vincent,²² traditional circumcision is ‘regarded as a gateway to sex rather than marking the point at which responsible sexual behaviour begins’, thus, ‘campaigns aimed at curbing HIV can be very confusing because they are layered onto other messages – of abstinence, sexual restraint and sexual responsibility on the one hand, and the rights and privileges that go along with Xhosa manhood on the other’. This work speaks to the necessity of fully understanding the awareness and beliefs about male circumcision for HIV prevention held by these men.

The current study assessed awareness of male circumcision for HIV prevention, risk compensation beliefs and perceptions of HIV risk as they relate to male circumcision among men in Cape Town, South Africa, who were traditionally circumcised. We believe this study to be the first of its kind to examine how male circumcision for HIV prevention may be influencing traditionally circumcised men. Specifically, we investigated influences of male circumcision awareness, risk compensation beliefs and perceptions of risk for HIV on reported sexual behaviours and chart-abstracted sexually transmitted infections (STIs).

Methods

Participants and settings

The study site was an urban health clinic in Cape Town, South Africa, that provides treatment in general medicine including STIs. Men and women (n = 749) attending the clinic were approached and asked to participate in a larger behavioural intervention which included, in part, an initial survey assessment. In total, 583 men and women agreed to participate. Of the 583 participants, 106 were women, 137 reported HIV-positive or unknown HIV status, 29 were uncircumcised men and seven men reported being circumcised for HIV prevention, leaving 304 HIV-negative, traditionally circumcised participants included in all further analyses. Women, uncircumcised men and HIV-positive/unknown persons were excluded in order to best understand what male circumcision-related expectations might exist among men culturally circumcised who possibly rely on male circumcision as a means of HIV prevention. A subsample of the 304 men consisted of 98 men for whom chart-abstracted STI data were available because of their agreement to partake in additional study opportunities.

Measures

All measures were administered in English, Xhosa and Afrikaans, the three languages spoken by nearly all clinic patients. Surveys were translated and back-translated to assure parallel forms. Measures were reviewed by staff familiar with local customs and piloted for feedback and clarity.

Demographic characteristics

Participants were asked their gender, years of education, age, ethnicity, whether they were circumcised, whether they were employed and information about their relationship status.

Participants were also asked to report their HIV status, the age at which they were circumcised and who performed their circumcision. A variable was created to differentiate between participants who were in a monogamous relationship versus those who were not. Specifically, participants reporting being in a committed relationship with one partner and no more than one sexual partner in the past three months were considered monogamous. All other participants were considered non-monogamous.

Sexual behaviour and STI outcomes

Participants were asked to report how many times they engaged in vaginal sex with and without a condom and number of female partners in the past three months.

In analyses using a subsample of the data (n = 98), STIs were coded from clinic medical charts from appointments occurring the same day as the survey assessment, and were used to form two variables. STI data were divided based on whether the infection could be established as being an incident infection or not. Chlamydia or gonorrhoea infections were summed and constituted an incident infection variable, while genital warts, genital herpes, syphilis and other STIs were summed together and treated as one composite STI variable indicative of chronic infection.

Awareness of male circumcision for HIV prevention

Participants were asked about whether they were aware of male circumcision providing protection against HIV. Specifically, the question was as follows: ‘I have heard that a man who is circumcised is safer from getting HIV'; responses were a dichotomous yes/no.

Risk compensation scale

Participants were asked their beliefs about male circumcision, sexual behaviour and HIV infection in order to assess how these factors affect their sexual risk behaviours (adapted measure²³). Participants were asked to respond to the following statements: (1) ‘condom use is not necessary if the man is circumcised'; (2) ‘if I am circumcised sex is safe without a condom'; (3) ‘being circumcised means a man can worry less about HIV/AIDS'; and (4) ‘if a man is circumcised he can have more sexual partners’. Responses ranged from 1 = strongly disagree to 6 = strongly agree. These items were used to form a male circumcision risk compensation scale that demonstrated internal consistency, Cronbach's alpha = 0.93.

HIV risk perception related to male circumcision

In order to assess participants’ perception of HIV risk in the context of male circumcision, a risk perception difference score was created from two items. The items asked participants to rate their perceived risk of HIV if a male partner engaged in the following sex acts with a HIV-positive woman: (a) ‘vaginal sex without condom’ and (b) ‘vaginal sex without a condom when the male partner is circumcised’. Participants rated the HIV risk associated with each activity on a 5-point Likert scale ranging from, very low risk = 1 to very high risk = 5. To create the risk perception difference score, item (b) was subtracted from item (a) resulting in a single measure of HIV risk perception related to male circumcision. Scores ranged from -4 to 4.

Drug and alcohol use

Participants were asked if they had used mandrax (methaqua-lone), marijuana, cocaine, heroin, methamphetamine, injection drugs or any other drug in the past three months. These responses were summed to create a composite score ranging from 0 to 7. For this scale, 0 = no drug use, and 7 = reporting using all listed drugs in the past four months. In addition, items from the Alcohol Use Disorders Identification Test²⁴ were used to assess alcohol frequency and consumption. Specifically, participants were asked to report how often they have a drink containing alcohol; responses ranged from never to more than 4 times a week; and how many drinks containing alcohol they have on a typical day when they are drinking; responses ranged from 1 don't drink to 10 or more.

Data analysis

Descriptive statistics were used to describe the sample population. Generalized linear modelling with a Poisson distribution for count data with a log-link function was used to model sexual behaviours, and binary logistic was used for STI variables. Parameter estimates from models were exponentiated and, therefore, results were reported as relative risks (RR) for Poisson variables and odds ratios (OR) for binary logistic variables, 95% confidence intervals (CI) were reported for both statistics. An analysis was conducted for the male circumcision HIV risk compensation scale to determine internal consistency using Cronbach's alpha. T-tests were run to establish any differences or similarities in the HIV risk perception variables and to test for differences on both the male circumcision risk compensation and HIV risk perception scales between men who were or were not aware of male circumcision for prevention. Variables were included in multivariate models due to their being important demographic controls (age, education) or their established relevance to sexual risk taking (drug and alcohol use) and, thus important to control for. PASW Statistics version 18.0 (SPSS Inc, Chicago, IL, USA) was used for all analyses.

Results

Participants averaged 10 years of education and 30 years of age. A majority of the participants were black, employed and in a non-monogamous relationship (Table 1). Drug use among the sample was minimal, with participants reporting an average of 0.22 drugs used in the past three months (SD = 0.54). Alcohol use varied, with 30.9% of the sample reporting never using alcohol, 14.1% monthly or less, 19.6% 2-4 times a month, 31.8% 2-3 times a week and 3.5% four or more times a week. During an average drinking occasion, 13.8% of participants reported one or two drinks, 24.1% three or four drinks, 20.9% five or six drinks, 6.4% 7, eight or nine drinks and 5.5% 10 or more drinks. On average, participants were circumcised at 19 years of age. The majority of men were circumcised at a traditional circumcision school.

Table 1

Demographic characteristics of self-reported HIV-negative circumcised men (n = 304)

	Mean	SD
Education	10.4	2.3
Age (years)	29.7	6.3
Age when circumcised (years)	19.0	5.06
Ethnicity	n	%
Black	296	96.4
White	7	2.3
Indian	0	0
Coloured	4	1.3
Other	0	0
Employed
Yes	246	79.4
No	64	20.6
Relationship status
Monogamous	76	24.8
Non-monogamous	230	75.2
Who performed circumcision
Medical doctor	27	8.7
Traditional healer	38	12.2
Circumcision school	241	77.5
Other	4	1.3
Where circumcision occurred
Eastern cape	250	80.6
Western cape	36	11.6
Other	24	7

Participants reported 1.96 (SD = 1.30) female sexual partners and 5.59 (SD = 5.12) protected and 5.15 unprotected vaginal sex acts (SD = 5.10) in the past three months. Among men with chart-abstracted STI information, the majority was identified as having an incident STI (n = 64, 66%) and fewer men presented with a chronic STI (n = 12, 12.3%).

Twenty-five percent (n = 77) of men had heard that male circumcision provides protection from HIV infection. Participants reported on average 2.00 (SD = 1.14) for a score on the male circumcision risk compensation scale. This average corresponded with a somewhat neutral stance on this scale. Participants’ perceptions of HIV risk averaged -0.42 (SD = 1.27), meaning that, on the whole, participants perceived lowered HIV risk when the male partner is circumcised. In a separate analysis, a paired samples t-test assessing a possible difference in perceived risks between ‘vaginal sex without condom’ (M = 4.8, SD = 0.72) and ‘vaginal sex without a condom when the male partner is circumcised’ (M = 4.4, SD = 1.22) demonstrated that participants reported an overall reduction in risk for HIV when the male partner is circumcised (t [304] = -5.89, P < 0.001). Men who reported being aware that male circumcision makes them safer from getting HIV were significantly more likely to endorse male circumcision risk compensation beliefs (t [304] = 2.20, P < 0.05, [M = 2.14, SD = 1.40]; [M = 1.81, SD = 1.07]). These means correspond approximately to slightly disagreeing and neutral beliefs towards risk compensation in response to male circumcision. There were no differences in the perceived HIV risk variable between men who were aware versus those who were not aware that male circumcision provides protection against HIV.

Univariate models predicting number of unprotected vaginal sex acts and number of female sexual partners

Awareness of male circumcision for HIV prevention, endorsement of the risk compensation scale, and lowered HIV risk perception when the male partner is circumcised were all significantly related to reporting greater numbers of unprotected vaginal sex acts. Endorsement of the risk compensation scale and lowered HIV risk perception when the male partner is circumcised were significant predictors of greater numbers of female sexual partners (Table 2).

Table 2

Univariate generalized linear models predicting unprotected sex acts and number of female sexual partners

	Unprotected vaginal sex acts			Number of female sexual partners
	RR^*	95% CI	P	RR	95% CI	P
Has heard that circumcised men are safer from HIV	1.19	1.06–1.32	<0.01	1.10	0.93–1.33	0.26
Male circumcision risk compensation scale	1.15	1.11–1.12	<0.01	1.08	1.01–1.15	<0.05
Lowered HIV risk perception when man is circumcised	1.08	1.04–1.12	<0.01	1.07	1.01–1.14	<0.05

Significant relative risks (RRs) larger than one correspond with greater numbers of unprotected vaginal sex acts and female sexual partners

CI = confidence interval

Multivariate model predicting number of unprotected vaginal sex acts

Being in a monogamous relationship, awareness of male circumcision for HIV prevention, lowered HIV risk perception when the male partner is circumcised, and endorsement of the risk compensation scale were associated with reporting greater numbers of unprotected sexual acts (Table 3). Age, education, alcohol consumption and frequency, and drug use were non-significant predictors for this model.

Table 3

Multivariate generalized linear models predicting number of unprotected vaginal sexual acts and female sexual partners (n = 304)

	RR^*	95% CI	P
Model 1 - Unprotected vaginal sex acts
Sociodemographics
Age (years)	0.99	0.99–1.01	0.15
Employed	0.95	0.83–1.09	0.46
Committed relationship	1.33	1.18–1.49	<0.001
Substance use
Alcohol frequency	0.96	0.91–1.02	0.22
Alcohol consumption	1.02	0.97–1.08	0.50
Drug use	1.00	0.94–1.07	0.91
Has heard that circumcised men are safer from HIV	1.13	1.01–1.26	<0.05
Male circumcision risk compensation scale	1.14	1.10–1.19	<0.001
Lowered HIV risk perception when man is circumcised	1.05	1.01–1.09	<0.001
Model 2 – Number of female sex partners
Sociodemographics
Age (years)	0.99	0.97–1.01	0.07
Employed	0.98	0.80–1.20	0.84
Substance use
Alcohol frequency	1.03	0.93–1.13	0.59
Alcohol consumption	1.02	0.94–1.11	0.65
Drug use	1.08	0.99–1.18	0.10
Has heard that circumcised men are safer from HIV	1.08	0.90–1.30	0.43
Male circumcision risk compensation scale	1.06	0.99–1.14	0.07
Lowered HIV risk perception when man is circumcised	1.06	1.00–1.13	0.05

Significant relative risks (RRs) larger than one correspond with greater numbers of unprotected vaginal sex acts and female sexual partners

CI = confidence interval

Multivariate model predicting number of female sexual partners

Greater endorsement of the risk compensation scale and lowered HIV risk perception when the male partner is circumcised were marginally associated with reporting more female sexual partners (Table 3). Age, education, alcohol frequency and consumption, and drug use, were not associated with number of female sexual partners.

Univariate models predicting STI

Greater endorsement of the male circumcision risk compensation scale was associated with a greater likelihood of reporting a chronic STI (Table 4). Other relationships were non-significant for these models.

Table 4

Logistic regression models predicting sexually transmitted infections (STIs) among a subsample of study participants (n = 98)

	OR	95% CI	P
Variables predicting incident STI
Has heard circumcised men are safer from HIV	1.13	0.44–2.87	0.80
Male circumcision risk compensation scale	0.75	0.52–1.07	0.11
Lowered HIV risk perception when man is circumcised	1.27	0.94–1.73	0.12
Variables predicting chronic STI
Has heard circumcised men are safer from HIV	1.27	0.35–4.62	0.72
Male circumcision risk compensation scale	1.64	1.06–2.53	<0.05
Lowered HIV risk perception when man is circumcised	1.21	0.81–1.80	0.36

OR = odds ratio; CI = confidence interval

Note: significant OR larger than one correspond with an STI diagnosis

Discussion

The current study draws attention to men in South Africa circumcised due to cultural practice and the sexual behavioural implications of male circumcision-related HIV risk reduction beliefs among this group. Unlike male circumcision RCTs, in which men are explicitly circumcised to reduce HIV transmission and provided extensive risk reduction counselling, the current study sample had undergone circumcision as a cultural rite of passage and/or for religious reasons. We found that among these men, those who were aware of the HIV preventive benefits of male circumcision and held positive beliefs about the reduced risk of HIV infection as a result of male circumcision were more likely to engage in risky sex; specifically unprotected sex and sex with multiple sexual partners. Given that the prevalence of male circumcision is 35% in South Africa, the study findings bear critical implications for the large number of men who would never participate in male circumcision HIV prevention programmes, but are certainly susceptible to behavioural adaptation in response to prevention information regarding circumcision. Moreover, population-based data have shown that the prevalence of HIV among circumcised and uncircumcised men in South Africa is the same, which suggests that we must carefully evaluate the effectiveness of male circumcision during large scale rollout.¹⁰ No study that we know of has focused on the beliefs of male circumcision among men circumcised for cultural/religious reasons. Further study in the area should focus on garnering a more in-depth understanding of how populations on the whole, not just men in male circumcision RCTs, are affected by biomedical technologies for HIV prevention. Results relating to STI findings warrant further investigation. Considering that the presence of an STI was related to risk compensation beliefs for some STIs but not others, this relationship should be researched with larger sample sizes.

Findings also suggest that as male circumcision HIV prevention programmes are considered for wider rollout, already circumcised men must be targeted in information, education and communication (IEC) campaigns regarding HIV risk reduction and that circumcision does not imply HIV risk elimination. Moreover, given that men circumcised for cultural reasons already appear to be internalizing messages about male circumcision for prevention, we stress that IEC campaigns should address these issues now, particularly among men who are at high-risk. However, to successfully provide accurate campaign messages for this population of men, the study findings underscore that further systematic research is needed on men already circumcised. Such studies will inform IEC campaigns and efforts to train traditional circumcisers regarding how to appropriately counsel men circumcised under their care.^25-28 The influence of these factors on already circumcised men with respect to HIV prevention is critical to assess.

A limitation of the current study is the cross-sectional design. We are, therefore, unable to determine the direction of causation in the relationship between male circumcision risk compensation beliefs, HIV risk perception and sexual risk behaviour. Future research in this area should employ longitudinal study designs. Data relating to STIs need replication with larger sample sizes and our findings may not be generalizable to men not attending a general health clinic. Moreover, future studies should examine basic knowledge around male circumcision for HIV prevention. For men in our study it is unknown exactly what they may or may not have heard about male circumcision for prevention.

Major study strengths include the notion that information garnered from the current study is novel and timely given current plans for widespread rollout. Given the millions of men already circumcised in high HIV prevalence African countries,¹⁹ the study findings have the potential to inform broader research and programming efforts in these contexts to ensure effective male circumcision for HIV prevention. Additional study strengths include the perception and risk compensation measures used; these measures can provide useful information for explaining cognitive processing around messages relating to male circumcision. Previous research simply assesses whether participants in male circumcision RCTs are more likely to engage in risky sex; however, the use of scales to assess men's risk compensation-related male circumcision beliefs may help identify high-risk groups that should be targeted in male circumcision HIV prevention programmes for risk reduction.

In conclusion, our study findings suggest that advances and novel application of medical technology in health prevention have the potential to reduce individuals’ perceptions of risk and, thus, motivation to engage in risk reduction behaviours. When drugs become available and medical procedures are developed to prevent disease and illness, behaviour change should be a closely monitored and well-understood factor, as it has the potential to undermine prevention and/or treatment efforts. Therefore, in the case of male circumcision for HIV prevention, evaluation of programme impact outside the context of male circumcision RCTs is necessary and public health messages may need to be tailored to counteract risk compensation and shape accurate HIV transmission beliefs among already-circumcised men.

Footnotes

Acknowledgements

National Institute of Mental Health (NIMH) grant number, 5R01MH074371, T32MH074387, T32MH020031 supported this research. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIMH or the National Institutes of Health.

References

Lavreys

, Rakwar

, Thompson

ML.

Effect of circumcision on incidence of human immunodeficiency virus type 1 and other sexually transmitted diseases: a prospective cohort study of trucking company employees in Kenya.

J Infect Dis 1999; 180: 330–36.

Reynolds

, Shepherd

, Risbud

AR.

Male circumcision and risk of HIV-1 and other sexually transmitted infections in India.

Lancet 2004; 363: 1039–40.

Shaffer

DIM

, Bautista

, Saferen

WB.

The protective effect of circumcision on HIV incidence in rural low-risk men circumcised predominantly by traditional circumcisers in Kenya: two-year follow-up of the Kericho HIV Cohort Study.

J Acquit Immune Defic Syndr 2007; 45: 371–79.

Gray

, Kigozi

, Serwadda

Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial.

Lancet 2007; 369: 657–66.

Auvert

, Taljaard

, Lagarde

Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial.

PLoS Med 2005; 2: e298.

Bailey

, Moses

, Parker

CB.

Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial.

Lancet 2007; 369: 643–56.

Siegfried

, Muller

, Deeks

HIV and male circumcision - a systematic review with assessment of the quality of studies.

Lancet Infect Dis 2005; 5: 165–73.

World Health Organization. Country Experiences in the Scale-up of Male Circumcision in the Eastern and Southern Africa region: Two Years and Counting. Geneva: WHO, 2009. See http://www.who.int/hiv/pub/malecircumcision/meetingreport_june09/en/index.html (last accessed 12 January 2009)

Green

, Travis

, McAllister

RG.

Male circumcision and HIV prevention insufficient evidence and neglected external validity.

Am J Prev Med 2010; 39: 479–82.

10.

Connolly

, Simbayi

, Shanmugam

, Nqeketo

Male circumcision and its relationship to HIV infection in South Africa: results of a national survey in 2002.

S Afr Med J 2008; 98: 789–94.

11.

White

, Glynn

, Orroth

KK.

Male circumcision for HIV prevention in sub-Saharan Africa: who, what and when?

AIDS 2008; 22: 1841–50.

12.

Eaton

, Kalichman

Risk compensation in HIV prevention: implications for vaccines, microbicides, and other biomedical HIV prevention technologies.

Curr HIV/AIDS Rep 2007; 4: 165–72.

13.

Hogben

, Liddon

Disinhibition and risk compensation: scope, definitions, and perspective.

Sex Transm Dis 2008; 35: 1009–10.

14.

Blower

, McLean

AR.

Prophylactic vaccines, risk behavior change, and the probability of eradicating HIV in San Francisco.

Science 1994; 265: 1451–4.

15.

Agot

, Kiarie

, Nguyen

HQ.

Male circumcision in Siaya and Bondo Districts, Kenya: prospective cohort study to assess behavioral disinhibition following circumcision.

J Acquir Immune Defic Syndr 2007; 44: 66

16.

Eaton

, Kalichman

SC.

Behavioral aspects of male circumcision for the prevention of HIV infection.

Curr HIV/AIDS Rep 2009; 6: 187–93.

17.

Mattson

, Campbell

, Bailey

RC.

Risk compensation is not associated with male circumcision in Kisumu, Kenya: a multi-faceted assessment of men enrolled in a randomized controlled trial.

PLoS One 2008; 3: e2443.

18.

Lukobo

, Bailey

RC.

Acceptability of male circumcision for prevention of HIV infection in Zambia.

AIDS Care 2007; 19: 471–77.

19.

World Health Organization and Joint United Nations Programme on HIV/AIDS. Male Circumcision: Global Trends and Determinants of Prevalence, Safety, and Acceptability. Geneva: WHO, 2007

20.

Weiss

, Larke

, Halperin

, Schenker

Complications of circumcision in male neonates, infants and children: a systematic review.

BMC Urol 2010; 10: 2

21.

Bailey

, Egesah

, Rosenberg

Male circumcision for HIV prevention: a prospective study of complications in clinical and traditional settings in Bungoma, Kenya.

Bull World Health Organ 2008; 86: 669–77.

22.

Vincent

‘Boys will be boys’: traditional Xhosa male circumcision, HIV and sexual socialisation in contemporary South Africa.

Cult Health Sex 2008; 10: 431–46.

23.

Eaton

, Kalichman

, Cain

DN.

Serosorting sexual partners and risk for HIV among men who have sex with men.

Am J Prev Med 2007; 33: 479–85.

24.

Saunders

, Aasland

, Babor

TF.

Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption - II.

Addiction 1993; 88: 791–804.

25.

Ozdemir

Significantly increased complication risks with mass circumcisions.

Br J Urol 1998; 81: 652

26.

Meissner

, Buso

DL.

Traditional male circumcision in the Eastern Cape -scourge or blessing?

S Afr Med J 2007; 97: 371–73.

27.

Crabb

Male circumcision to prevent heterosexual HIV transmission gets (another) green light, but traditional circumcision in Africa has ‘shocking’ number of complications.

AIDS 2010; 24: Nl–2.

28.

Peltzer

, Nqeketo

, Petros

, Kanta

Traditional circumcision during manhood initiation rituals in the Eastern Cape, South Africa: a pre-post intervention evaluation.

BMC Public Health 2008; 8: 64