Tropical Medicine and International Health
doi:10.1111/j.1365-3156.2011.02744.x
volume 16 no 5 pp 589–597 may 2011
Prevalence and factors associated with knowledge of and willingness for male circumcision in rural Zimbabwe Webster Mavhu1, Raluca Buzdugan2, Lisa F. Langhaug2, Karin Hatzold3, Clemens Benedikt4, Judith Sherman5, Susan M. Laver5, Oscar Mundida6, Godfrey Woelk7 and Frances M. Cowan1,2 1 2 3 4 5 6 7
Zimbabwe AIDS Prevention Project, Department of Community Medicine, University of Zimbabwe, Harare, Zimbabwe Centre for Sexual Health & HIV Research, University College London, UK Population Services International-Zimbabwe, Harare, Zimbabwe UNFPA, Harare, Zimbabwe UNICEF, Harare, Zimbabwe National AIDS Council, Harare, Zimbabwe Research Triangle Institute, Durham, NC, USA
Summary
objective To explore male circumcision (MC) prevalence, knowledge, attitudes and intentions among rural Zimbabweans. methods Representative survey of 18–44 year olds in two provinces, as part of an evaluation of the Zimbabwe National Behaviour Change Programme. We conducted univariate, bivariate and multivariate analyses. Linear regression was employed to predict knowledge of MC (composite index) and logistic regression to predict knowledge that MC prevents HIV, willingness (oneself or one’s partner) to undergo MC, and willingness to have son circumcised. results Two thousand seven hundred and forty-six individuals participated in the survey (87% of eligibles). About two-thirds were women (64%). Twenty per cent of men reported being circumcised, while 17% of women reported having a circumcised partner. Knowledge of MC and its health benefits was low. Attitudes towards MC were relatively positive. If it could prevent HIV, 52% of men reported that they would undergo MC and 58% of women indicated that they would like their partners to be circumcised. Seventy-five per cent of men who reported being HIV positive were willing to undergo MC, against 52% of those who reported HIV negative status. Reported acceptability of neonatal circumcision was high with 58% of men and 60% of women reporting that they would have their sons circumcised if it protected them against HIV. Fear of adverse effects was highlighted as a barrier to MC acceptability. conclusion More knowledge about MC’s health benefits positively affects people’s attitudes towards MC. The relatively high MC acceptability suggests an enabling environment for the scale-up programme. keywords acceptability, HIV, male circumcision, prevalence, Zimbabwe
Introduction By 2007 three randomised controlled trials had demonstrated male circumcision’s (MC) efficacy in preventing female-to-male HIV transmission, estimating a protective effect of between 51% and 60% (Auvert et al. 2005; Bailey et al. 2007; Gray et al. 2007). Subsequently, UNAIDS and WHO recommended that MC be provided as an additional prevention strategy in high HIV prevalence countries (WHO 2007). Zimbabwe has one of the world’s most sustained HIV epidemics (Stirling et al. 2008; NAC & UNFPA 2009). The
ª 2011 Blackwell Publishing Ltd
country’s MC prevalence, one of the region’s lowest, is around 10.3% (Zimbabwe Central Statistical Office 2007). Mathematical modelling estimates that 750 000 HIV infections could be averted if 80% of Zimbabwean men are circumcised within seven years (USAID 2009). Initially prioritizing males aged 15–29 will lead to the greatest reduction in HIV incidence in the short-term, whilst scaling up MC to reach the targeted number will likely yield US$3.8 billion net savings over 15 years (USAID 2009). In 2007, Zimbabwe’s Ministry of Health and Child Welfare (MOHCW) and the National AIDS Council decided that MC should be integrated into the National 589
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
AIDS Prevention Strategy. With technical and financial support offered by Population Services InternationalZimbabwe, MOHCW established five MC pilot sites in 2009. The national MC programme, currently targeting men aged 15–29, aims to reach all willing, eligible men, after which MC will be extended to infants (MOHCW 2010). In 2009, a population-based survey was conducted to evaluate Zimbabwe’s National Behaviour Change Programme. The survey included MC questions; reported findings will contribute towards the planning and implementation of scale-up MC services.
Methods
Results Socio-demographic profile A total of 2746 individuals participated in the survey (87% of eligibles). Survey participants were mostly women (64%, n = 1747), young (mean age 28), of Shona ethnicity (96%, n = 2624), Christian (73%, n = 1997), currently (59%, n = 1614) or previously married (16%, n = 440), and unemployed (37%, n = 1010) or involved in small-scale farming (37%, n = 1029); 45% (n = 1237) had ‡11 years of education (Table 1). Compared to men, women were slightly older, mostly married (67%, n = 1164), less educated and more likely to be unemployed (39%, n = 675) and ⁄ or farmers (43%, n = 748).
Design and participants The National Behaviour Change Programme (NBCP) study was a representative interim survey of 18–44 year olds conducted in two Zimbabwean provinces (October– November 2009). The survey was conducted in six districts, four where the NBCP had been implemented since 2007 (‘focus’ districts, n = 1829) and two where it had not (‘nonfocus’ districts, n = 917), in order to evaluate NBCP’s impact on HIV-related knowledge, attitudes and norms. Data were collected using audio computer-assisted survey instrument (ACASI), a questionnaire delivery mode which had been used previously in these districts without problems (Langhaug et al. 2010a,b). Ethics approval was given by the Medical Research Council of Zimbabwe and the ethics board of University College London. Data analysis Data were analysed using stata 10. While we could not use stata survey functions correcting for district-level clustering (given small number of districts), we controlled for district in all significance tests. Analysis focused on four main dependent variables (DVs): MC knowledge, knowledge that MC prevents HIV, willingness (oneself or one’s partner) to undergo MC, and willingness to have one’s son circumcised. For each DV, we conducted univariate, bivariate and multivariate analyses. At the bivariate level, we examined associations between each DV and sex, age, ethnic group, education, religion, marital status, occupation, NBCP exposure, ever having had sex, HIV knowledge, and reported HIV status. MC knowledge was examined as a possible factor associated with attitudes towards adult and infant MC. Variables statistically significant at P < 0.10 were included in multivariate models (Tables 2–5). Linear regression was employed to predict MC knowledge and logistic regression for the other DVs.
590
MC prevalence Overall, 20% of men (n = 203) reported being circumcised, while 17% of women with at least one sexual partner (n = 210) reported having a circumcised partner (31%, n = 381 were uncertain about their partners’ MC status). MC knowledge Overall, knowledge of MC and its health benefits was low; out of eight MC items, 57% (n = 1555) provided 0–2 correct answers, 32% (n = 863) 3–5 answers and 12% (n = 326) 6–8 answers, with women being less informed. Similarly, only 38% of men (n = 380) and 25% of women (n = 429) knew that MC could prevent HIV. Information was obtained from friends ⁄ relatives (males = 38%, n = 144, females = 27%, n = 114) and ⁄ or testing and counselling service providers (males = 29%, n = 109, females = 23%, n = 98). Newspapers (21%, n = 79 and 12%, n = 50), television (16%, n = 62 and 15%, n = 62), and health workers (18%, n = 69 and 19%, n = 83) were additional sources of information (data not shown). Attitudes towards adult and infant MC Attitudes towards MC were relatively positive. If it could prevent HIV, 52% of men (n = 399) reported that they would undergo MC and 58% of women (n = 840) would like their partners circumcised. Main reasons for unwillingness were: disbelief that MC protects against HIV (50%, n = 186 and 39%, n = 238 respectively), cultural issues (22%, n = 83 and 16%, n = 99) and fear of pain and ⁄ or adverse effects (26%, n = 96 and 20%, n = 123) (data not shown).
ª 2011 Blackwell Publishing Ltd
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
Table 1 Socio-demographic characteristics and main outcome variables by gender (%) Sex Age
Ethnic group Education
Religion Marital status
Occupation
Circumcised (oneself ⁄ the partner)
Knowledge about MC*
Knowledge that MC is an HIV prevention method Willingness to undergo MC (oneself ⁄ the partner) Willingness to have one’s son undergo MC
Subcategory
Men (n = 999)
Women (n = 1747)
Total (n = 2746)
Men Women 18–19 years 20–24 years 25–29 years 30–34 years 35–39 years 40–44 years Mean Other Shona Primary or less Form 1–3 Form 4 Above form 4 Non-Christian Christian Married Previously married Never married Subsistence farming Commercial farming Formally employed Other employment Not employed
– – 20.2 27.1 18.4 12.3 12.2 9.7 27.2 4.1 95.9 11.5 21.9 50.9 15.7 26.6 73.4 45.1 10.4 44.5 13.2 14.9 16.2 22.1 33.5
– – 14.4 26.6 20.3 14.9 13.6 10.3 28.1 4.6 95.4 22.5 29.5 40.2 7.9 27.7 72.4 66.6 19.2 14.1 17.2 25.6 8.3 10.3 38.6
36.4 63.6 16.5 26.8 19.6 14.0 13.1 10.1 27.8 4.4 95.6 18.5 26.8 44.1 10.7 27.3 72.7 58.8 16.0 25.2 15.7 21.7 11.2 14.6 36.8
No Yes Don’t know Total Low Medium High No Yes
79.7 20.3 – (n = 999) 46.5 37.9 15.6 61.9 38.1
51.5 17.2 31.3 (n = 1219) 62.5 27.8 9.7 75.4 24.6
56.7 31.5 11.9 70.5 29.5
No Yes
48.4 51.6
42.2 57.8
44.3 55.7
No Yes
42.3 57.7
39.7 60.3
40.6 59.4
– – –
MC, male circumcision. *MC knowledge index (0–8) was created by adding answers (1 = yes, 0 = no ⁄ don’t know) to items: MC (i) protects against STIs, (ii) protects against penile cancer, (iii) protects against uterine cancer, (iv) improves hygiene, (v) reduces fertility in men (recoded), (vi) reduces male sexual pleasure (recoded), (vii) reduces female sexual pleasure (recoded), (viii) does not benefit women (recoded); the index was categorized into low (0–2), medium (3–5), and high (6–8) knowledge. Excludes the cases of men who have undergone male circumcision and of women whose partners are circumcised.
Acceptability of neonatal circumcision was high, with 58% of men (n = 576) and 60% of women (n = 1053) reporting willingness to have their sons circumcised if it
ª 2011 Blackwell Publishing Ltd
protected them against HIV. However, 15% of men (n = 62) and 12% of women (n = 86) thought their sons should decide for themselves when they became older (data not shown).
591
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
Table 2 Main factors associated with knowledge about male circumcision* (n = 2744) Linear regression Independent variable Sex Female Male Age Education Primary or less Form 1 to 3 Form 4 Above form 4 Marital status Never married Married Previously married Occupation Not employed Subsistence farming Commercial farming Formally employed Other employment Programme exposure No Yes Ever had sex No Yes HIV knowledgeà Low Medium High Reported HIV status Not tested ⁄ not reported Negative Positive
Mean
b
95% CI
2.06 2.86 –
0.00 0.94 0.02
0.74 to 1.13 0.00 to 0.03
2.10 2.19 2.44 2.77
0.00 0.06 0.18 0.34
)0.20 to 0.31 )0.06 to 0.42 )0.02 to 0.70
1.93 2.50 2.46
0.00 0.49 0.38
0.19 to 0.78 0.04 to 0.73
2.07 2.57 2.27 2.97 2.51
0.00 0.16 0.04 0.47 0.08
)0.10 to 0.42 )0.19 to 0.27 0.15 to 0.79 )0.18 to 0.35
2.09 2.67
0.00 0.25
0.02 to 0.47
1.64 2.50
0.00 0.42
0.13 to 0.72
1.63 2.32 2.60
0.00 0.57 0.79
0.27 to 0.87 0.48 to 1.11
2.24 2.47 2.62
0.00 0.09 0.12
)0.09 to 0.27 )0.24 to 0.48
P-value 0.000 0.013 0.204
0.005
0.062
0.031
0.005
0.000
0.551
b, unstandardized coefficient, CI, confidence interval, b = 0.00 indicates the reference category for each categorical variable. *MC knowledge index (0–8) (see categorisation under Table 1). Programme exposure is 1 (yes) if the respondent had either: (i) been to a meeting where a trained behaviour change facilitator (BCF) talked about HIV or behaviour change, (ii) participated in the programme’s behaviour change ‘Love and Respect’ course, (iii) been visited by a BCF in his ⁄ her home in the last year, or (iv) listened to the ‘Love carefully’ radio programme in the last year. àHIV knowledge index (0–6) was created by adding answers (1 = yes, 0 = no ⁄ don’t know) to items: (i) If you look carefully, you can know if someone has HIV (recoded), (ii) Using condoms can prevent you from being infected by HIV, (iii) A person who looks strong and healthy can have HIV, (iv) A mother can transmit HIV through breastfeeding, (v) You can get HIV if you share utensils with someone who is infected (recoded), (vi) If a mosquito bites you it can infect you with HIV (recoded); the index was categorised into low (0–2), medium (3– 4), and high knowledge (5–6). In addition to the variables listed in the table, also controlled for district (only the linear regression); At the bivariate level, the associations with religion, marital status and HIV knowledge were not statistically significant.
Factors associated with MC Sexually active men were more likely to be circumcised than non-sexually active men (25%, 95% confidence interval (CI) 22–28% vs. 6%, 95% CI 3–10%) (data not shown). Circumcised men were more likely to be married 592
(29%, n = 129) and divorced ⁄ separated ⁄ widowed (24%, n = 25 vs. 11% of unmarried men, n = 49). Circumcised men had better MC knowledge than uncircumcised ones (P < 0.001); 29% of men with high knowledge were circumcised, compared to 23% with medium (and 16% with low) knowledge.
ª 2011 Blackwell Publishing Ltd
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
Table 3 Main factors associated with knowledge that male circumcision is an HIV prevention method (n = 2744)
Logistic regression Independent variable Sex Female Male Age Ethnic group Shona Other Education Primary or less Form 1 to 3 Form 4 Above form 4 Occupation Not employed Subsistence farming Commercial farming Formally employed Other employment Programme exposure* No Yes Ever had sex No Yes Reported HIV status Not tested ⁄ not reported Negative Positive
%
AOR
95% CI
24.6 38.1 –
1.00 1.91 1.01
1.59–2.30 1.00–1.03
29.0 41.0
1.00 1.51
1.01–2.24
27.9 26.5 28.9 40.8
1.00 0.88 0.88 1.35
0.68–1.15 0.69–1.13 0.95–1.92
26.0 34.5 25.3 39.9 31.1
1.00 1.31 0.93 1.39 1.05
1.01–1.70 0.73–1.19 1.02–1.91 0.81–1.38
25.2 34.7
1.00 1.48
1.16–1.87
24.7 30.5
1.00 1.14
0.88–1.49
26.7 32.7 33.1
1.00 1.37 1.30
1.14–1.64 0.91–1.86
P-value 0.000 0.026 0.042
0.030
0.038
0.001
0.325
0.003
AOR, adjusted odds ratio; CI, confidence interval. *Programme exposure (see description under Table 2). In addition to the variables listed in the table, also controlled for district (only the logistic regression); At the bivariate level, the associations with religion, marital status and HIV knowledge were not statistically significant.
Factors associated with MC knowledge Participants registered higher ‘MC knowledge index’ scores if they were men (unstandardised coefficient (b) = 0.94, 95% CI 0.74–1.13), older (b = 0.02, 95% CI 0.00–0.03), married (b = 0.49, 95%CI 0.19–0.78) or previously married (b = 0.38, 95% CI 0.04–0.73), sexually active (b = 0.42, 95% CI 0.13–0.72), formally employed (b = 0.47, 95% CI 0.15–0.79), exposed to NBCP (b = 0.25, 95% CI 0.02–0.47) and had good HIV knowledge (high knowledge b = 0.79, 95% CI 0.48–1.11, medium b = 0.57, 95% CI 0.27–0.87) (Table 2). Similarly, sex, age, education, occupation and programme exposure were significantly associated with knowledge that MC prevents HIV (P < 0.05) (Table 3). In addition, non-Shona participants (adjusted odds ratio (AOR) = 1.51, 95% CI 1.01–2.24) and those who knew their HIV status (negative AOR = 1.37, 95% CI 1.14–1.64,
ª 2011 Blackwell Publishing Ltd
positive AOR = 1.30, 95% CI 0.91–1.86) were more likely to know that MC prevents HIV. Factors associated with willingness for adult MC Men were more willing to undergo MC if they had sound MC knowledge (high knowledge AOR = 3.73, 95% CI 2.24–6.21) (Table 4). Women were more likely to favour adult MC if they were informed about its health benefits (high knowledge AOR = 2.97, 95% CI 1.87–4.70) and had been exposed to NBCP’s activities (AOR = 1.32, 95% CI 1.00–1.47). Factors associated with willingness for infant MC Men were more likely to favour infant MC if they had either low or high education (1st–3rd form AOR = 0.52, 593
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
Table 4 Logistic regression predicting willingness to undergo circumcision (oneself ⁄ the partner) if it prevented HIV*
Independent variable Programme exposure No Yes Knowledge about MCà Low Medium High Education Primary or less Form 1–3 Form 4 Above form 4 Marital status Never married Married Previously married Ever had sex No Yes HIV knowledge§ Low Medium High Reported HIV status Not tested ⁄ not reported Negative Positive
Men (n = 773)
Women (n = 1453)
Logistic regression
Logistic regression
%
AOR
95% CI
P-value
– – 43.5 53.1 75.9
1.00 1.46 3.73
1.05–2.03 2.24–6.21
0.000
63.5 44.7 50.6 55.6
1.00 0.44 0.50 0.60
0.25–0.78 0.30–0.84 0.32–1.10
51.4 48.6 64.9
1.00 0.66 1.20
0.45–0.96 0.68–2.12
45.8 53.8
1.00 1.40
0.94–2.08
46.9 48.7 58.0
1.00 0.98 1.31
0.61–1.57 0.78–2.20
50.1 52.3 75.0
1.00 1.02 2.77
0.73–1.41 1.02–7.56
0.033
%
AOR
95% CI
55.1 61.4
1.00 1.32
1.00–1.74
54.9 58.1 79.2
1.00 1.18 2.97
0.92–1.52 1.87–4.70
57.0 51.7 61.0 66.1
1.00 0.76 1.12 1.32
0.56–1.04 0.84–1.49 0.83–2.09
0.029
– – –
0.101
– –
0.234
– – –
0.136
– – –
P value 0.047
0.000
0.015
AOR, adjusted odds ratio; CI, confidence interval; MC, male circumcision. *Excludes the cases of men who have undergone male circumcision and of women whose partners are circumcised. Programme exposure (see description under Table 2). àMC knowledge index (0–8) (see categorisation under Table 1). §HIV knowledge index (0–6) (see categorisation under Table 2). In addition to the variables listed in the table, also controlled for district (only the logistic regression); At the bivariate level, the associations with religion, marital status and HIV knowledge were not statistically significant.
95% CI 0.32–0.85, 4th form AOR = 0.70, 95% CI 0.44– 1.10, above 4th form AOR = 0.95, 95% CI 0.54–1.69, reference = primary education), good MC knowledge (AOR = 6.01, 95% CI 3.61–10.01), and good HIV knowledge (high knowledge AOR = 1.60, 95% CI 1.00– 2.57, medium AOR = 1.63, 95% CI 1.06–2.50) (Table 5). Women were more likely to favour infant MC if they had high MC knowledge (AOR = 2.78, 95% CI 1.87–4.11) and were exposed to NBCP (AOR = 1.32, 95% CI 1.03– 1.70). There was a strong association between attitudes towards adult and neonatal MC; 94% of men (n = 374) who indicated own willingness and 90% of women (n = 752) who favoured adult MC also expressed willingness for infant MC (data not shown). 594
Discussion This study showed low reported MC prevalence. Additionally, knowledge of MC and its health benefits was poor. Despite the poor knowledge, attitudes towards adult MC were favourable. Of note, almost 60% of women with uncircumcised partners indicated willingness for their counterparts to undergo MC. A population-based survey conducted by Population Services International-Zimbabwe also found 62% of men willing to undergo MC (PSIZimbabwe 2006). In both instances, acceptability was higher than the 45% reported previously for Zimbabwe (Halperin et al. 2005), suggesting a gradual increase in MC acceptability.
ª 2011 Blackwell Publishing Ltd
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
Given the low knowledge of MC’s benefits, extensive awareness campaigns are required in order to realise the ambitious target to circumcise 1.3 million Zimbabwean men in the next 5–7 years. Awareness campaigns will need to employ several strategies. While men and urban populations are exposed to mass media (e.g. newspapers, radio), the same is not true for rural populations and women (Zimbabwe Central Statistical Office 2007). Therefore, the use of participatory approaches (e.g. community mobilisations, road shows) would be more appropriate for reaching the latter.
Since women and younger men had less MC knowledge, they need to be specifically targeted. It is likely to be important to promote MC among young men before they become sexually active. Additionally, as mothers and partners, women are likely to have considerable influence, even if covert (Westercamp & Bailey 2007). In this study, women’s attitudes towards infant MC were favourable. This is encouraging since UNAIDS and WHO recommend that, in addition to adult MC, high HIV prevalence countries consider neonatal MC as a longer term HIV prevention strategy (WHO 2007). Moreover, although public health benefits will take longer to realise, infant MC
Table 5 Logistic regression predicting willingness to have one’s son circumcised if it prevented HIV
Independent variable Programme exposure* No Yes Knowledge about MC Low Medium High Education Primary or less Form 1 to 3 Form 4 Above form 4 Occupation Not employed Subsistence farming Commercial farming Formally employed Other employment Ever had sex No Yes HIV knowledgeà Low Medium High Reported HIV status Not tested ⁄ not reported Negative Positive
Men (n = 998)
Women (n = 1746)
Logistic regression
Logistic regression
%
AOR
95% CI
P-value
– – 47.4 58.7 85.9
1.00 1.47 6.01
1.10–1.96 3.61–10.01
0.000
63.2 47.5 58.7 65.4
1.00 0.52 0.70 0.95
0.32–0.85 0.44–1.10 0.54–1.69
53.4 66.7 59.7 63.0 53.6
1.00 1.39 1.03 1.06 0.95
0.88–2.19 0.68–1.58 0.67–1.66 0.66–1.38
50.9 59.8
1.00 1.12
0.81–1.55
41.9 58.4 62.2
1.00 1.63 1.60
1.06–2.50 1.00–2.57
55.2 60.8 70.7
1.00 1.07 1.62
0.80–1.44 0.75–3.50
%
AOR
95% CI
58.0 63.2
1.00 1.32
1.03–1.70
56.7 61.9 78.8
1.00 1.29 2.78
1.03–1.61 1.87–4.11
0.023
– – – –
0.622
– – – – –
0.507
– –
0.082
– – –
0.459
– – –
P-value 0.029
0.000
AOR, adjusted odds ratio; CI, confidence interval; MC, male circumcision. *Programme exposure (see description under Table 2). MC knowledge index (0–8) (see categorisation under Table 1). à HIV knowledge index (0–6) (see categorisation under Table 2). In addition to the variables listed in the table, also controlled for district (only the logistic regression); At the bivariate level, the associations with religion, marital status and HIV knowledge were not statistically significant.
ª 2011 Blackwell Publishing Ltd
595
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
is more cost-saving than adult circumcision (Binagwaho et al. 2010). Fear of adverse effects was reported as a barrier to MC acceptability, which corroborates findings from other regional studies (Ngalande et al. 2006; Lukobo & Bailey 2007). Sustained acceptability and uptake will therefore require performance of MC with minimal adverse effects to increase people’s confidence in the surgical procedure (Westercamp & Bailey 2007). Of note, 75% of men who reported being HIV positive were willing to undergo MC, compared to 52% of those who reported the contrary. This suggests a misconception about MC’s ability to either eliminate HIV or reduce the risk of male-to-female transmission, which needs both exploration and clarification. Community-based interventions (e.g. NBCP) can play a dual role of promoting MC’s advantages and dispelling associated myths. This study has several strengths. Firstly, data were collected from a large, representative sample of 18- to 44-year olds. Secondly, the survey response rate (87%) was high and similar to that for the most recent Zimbabwe Demographic and Health Survey (86%) (Zimbabwe Central Statistical Office 2007). Thirdly, issues discussed here are recent, which offers an opportunity for comparative analyses with previous findings. Lastly, since neonatal MC is generally unavailable in sub-Saharan Africa, there are concerns around its acceptability (Plank et al. 2009). These data add to the growing evidence on neonatal MC acceptability in this region. There were, however, some limitations to this study. Firstly, as collected data were restricted to 18–44 years olds, in-laws and grandparents were likely underrepresented, as they are older. However, they are key decisionmakers on issues surrounding newborns (Gelfand 1965). A study conducted in India, where extended family members’ opinions around infant-related issues are also revered, found that other family members were as important as the child’s parents with respect to neonatal circumcision decisions (Madhivanan et al. 2008). The extended family’s role around neonatal MC decision-making is being explored, with the aim to eventually target in-laws and grandparents with information about MC’s health benefits. Secondly, the only two indigenous, formal terms for MC used in the questionnaire, kuchecheudzwa and kudzingiswa, are uncommon. However, a questionnaire pretesting exercise revealed that a description included to clarify these terms – ‘meaning that the foreskin of your private part (penis) has been removed’ – was well-understood by participants. Lastly, self-reported MC was not validated by clinical examinations. MC prevalence findings should therefore be interpreted with caution; one study demonstrated unreliability of reported MC when 34% of uncir596
cumcised men reported being circumcised (Lilienfield & Graham 1958). In conclusion, this study suggests that increasing people’s knowledge about MC’s health benefits is likely to positively affect attitudes towards the procedure. Programmes such as the NBCP should use a variety of communication tools to improve knowledge about MC to a wide audience. The relatively high MC acceptability suggests an enabling environment for the scale-up programme. Attaining high national MC targets will likely be important in furthering the HIV decline in Zimbabwe. References Auvert B, Taljaard D, Lagarde E, Sobngwi-Tambekou J, Sitta R & Puren A (2005) Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial. PLoS Medicine 2, 1112–1122. Bailey RC, Moses S, Parker CB et al. (2007) Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet 369, 643–656. Binagwaho A, Pegurri E, Muita J & Bertozzi S (2010) Male circumcision at different ages in Rwanda: a cost-effectiveness study. PLoS Medicine 7, 1–10. Gelfand M (1965) African Background: The Traditional Culture of the Shona. Juta, Cape Town. Gray RH, Kigozi G, Serwadda D et al. (2007) Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet 369, 657–666. Halperin DT, Fritz K, Mcfarland W & Woelk G (2005) Acceptability of adult male circumcision for sexually transmitted disease and HIV prevention in Zimbabwe. Sexually Transmitted Diseases 32, 238–239. Langhaug LF, Cheung YB, Pascoe SJS et al. (2010a). How you ask really matters: randomised comparison of four sexual behaviour questionnaire delivery modes in Zimbabwean youth. Sexually Transmitted Infections [Epub ahead of print October 13, doi: 10.1136/sti.2009.037374]. Langhaug LF, Sherr L & Cowan FM (2010b) How to improve the validity of sexual behaviour reporting: systematic review of questionnaire delivery modes in developing countries. Tropical Medicine & International Health 15, 362–381. Lilienfield AM & Graham S (1958) Validity of determining circumcision status by questionnaire as related to epidemiologic studies of cancer of the cervix. Journal of the National Cancer Institute 21, 713–720. Lukobo MD & Bailey RC (2007) Acceptability of male circumcision for prevention of HIV infection in Zambia. AIDS Care 47, 1–7. Madhivanan P, Krupp K, Chandrasekaran V, Karat SC, Reingold AL & Klausner JD (2008) Acceptability of male circumcision among mothers with male children in Mysore, India. AIDS 22, 983–988. Ministry of Health and Child Welfare [MOHCW] (2010) Strategy for Safe Medical Male Circumcision Scale Up to Support
ª 2011 Blackwell Publishing Ltd
Tropical Medicine and International Health
volume 16 no 5 pp 589–597 may 2011
W. Mavhu et al. Male circumcision in rural Zimbabwe
Comprehensive HIV Prevention in Zimbabwe. MOHCW, Harare. National Aids Council [NAC] & UNFPA (2009) National Behaviour Change Strategy Baseline Survey 2007 ⁄ 08. NAC & UNFPA, Harare. Ngalande R, Levy J, Kapondo C & Bailey RC (2006) Acceptability of male circumcision for prevention of HIV infection in Malawi. AIDS and Behavior 10, 377–385. Plank RM, Makhema J, Kebaabetswe P et al. (2010) Acceptability of infant male circumcision as part of HIV prevention and male reproductive health efforts in Gaborone, Botswana, and surrounding areas. AIDS and Behavior 14, 1198–1202. Population Services International-Zimbabwe [PSI-Zimbabwe] (2006) Tracking Results Continuously National Survey. PSIZimbabwe, Harare.
Stirling M, Rees H, Kasedde S & Hankins C (2008) Introduction: addressing the vulnerability of young women and girls to stop the HIV epidemic in southern Africa. AIDS 22(Suppl. 4), S1–S3. United States Agency for International Development [USAID] (2009) The Potential Cost and Impact of Expanding Male Circumcision in Zimbabwe. USAID, Washington, DC. Westercamp N & Bailey RC (2007) Acceptability of male circumcision for prevention of HIV ⁄ AIDS in sub-Saharan Africa: a review. AIDS and Behavior 11, 341–355. WHO (2007) New Data on Male Circumcision and HIV Prevention: Policy and Programme Implications. WHO, Geneva. http://www.who.int/hiv/pub/malecircumcision/en (accessed on 11 October 2010). Zimbabwe Central Statistical Office (2007) Zimbabwe Demographic and Health Survey (ZDHS) 2005–06. Macro International, Calverton.
Corresponding Author Webster Mavhu, Zimbabwe AIDS Prevention Project, Department of Community Medicine, University of Zimbabwe, 92 Prince Edward Rd, Harare, Zimbabwe. Tel.: +263 712 432 215; Fax: +263 4 707 291; E-mail:
[email protected]
ª 2011 Blackwell Publishing Ltd
597