Male circumcision for prevention of homosexual acquisition of HIV in men
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To assess the evidence of an interventional effect of male circumcision for preventing acquisition of HIV‐1 and HIV‐2 by men through sex with men
Background
Sub‐Saharan Africa is the region most heavily affected by HIV, accounting for two thirds of all people living with HIV globally in 2007. Nearly two million people were newly infected with the virus in sub‐Saharan Africa in 2007, bringing the number of people currently living with HIV in the region to 22 million [UNAIDS 2008]. The HIV epidemics in countries of the region are reported to be predominantly heterosexual, and are contrasted with the epidemics in much of the western world and Eastern Europe where injection drug use and homosexual transmission predominate. Consequently, much of the published HIV research assumes heterosexual spread and suggested preventive measures have largely been directed towards this mode of HIV transmission.
Publications suggesting a partially protective effect of male circumcision, the surgical removal of the foreskin of the penis, on the risk of human immunodeficiency virus (HIV) acquisition in men began to appear in the mid‐1980s [Fink 1986]. Thereafter, two decades of observational data indicated that circumcised heterosexual men have lower levels of HIV infection than uncircumcised men [Fink 1986; Moses 1990; O'Farrell 2000; Weiss 2000; Auvert 2001; Drain 2004; Baeten 2005]. Throughout the world, HIV prevalence is generally lower in populations that traditionally practise male circumcision than in populations where most men are not circumcised [Drain 2004]. However, it was unclear whether this observation may be the result of a biological effect of male circumcision, or due to cultural or social factors that occur in parallel with high levels of male circumcision. Consequently, three randomised controlled trials were set up to examine the impact of male circumcision on HIV acquisition in heterosexual men in South Africa, Kenya, and Uganda [Auvert 2005; Bailey 2007; Gray 2007].
The Orange Farm trial, South Africa, which enrolled 3274 uncircumcised men aged 18 to 24 years showed a 61% protection against HIV acquisition [Auvert 2005]. The trial in Kisumu, Kenya, of 2784 HIV‐negative men showed a 53% reduction of HIV acquisition in circumcised men relative to uncircumcised men [Bailey 2007]. The trial of 4996 HIV‐negative men in Rakai, Uganda, showed that HIV acquisition was reduced by 48% in circumcised men [Gray 2007]. The trials involved adult, HIV‐negative heterosexual male volunteers assigned at random to either undergo circumcision performed by trained medical professionals in a clinic setting or wait until after the end of the trial to be circumcised.
Siegfried and colleagues have conducted a Cochrane review of the effect of male circumcision in prevention of heterosexual acquisition of HIV in men [Siegfried 2003]. However, no Cochrane review to date has reviewed the currently available evidence on the association between male circumcision and homosexual acquisition of HIV infection in men. We are thus undertaking this Cochrane review to determine the association between male circumcision and HIV acquisition in men who have sex with men. Significant differences exist between the sexual practices of men who have sex with women and those of men who have sex with men to justify the proposed Cochrane review. For example, prospective cohort studies have identified unprotected receptive anal intercourse as a practice with the greatest risk of HIV infection in men who have sex with men [Koblin 2006]; suggesting that the circumcision status of the insertive partner may be an important variable influencing risk of HIV acquisition by the receptive partner [Fankem 2008]. The latter is also of interest to women who engage in unprotected intercourse with circumcised men.
Objectives
To assess the evidence of an interventional effect of male circumcision for preventing acquisition of HIV‐1 and HIV‐2 by men through sex with men
Methods
Criteria for considering studies for this review
Types of studies
We will preferentially look for randomised or quasi‐randomized controlled trials. If we find no randomized controlled trials (RCTs), we will include data from cohort, case‐control, and cross‐sectional studies in the review.
Types of participants
Men who have sex with men (including transgenders)
Types of interventions
Male circumcision (i.e. surgical removal of the foreskin of the penis) as determined by direct observation, self report, or partner‐report
Types of outcome measures
Primary outcome: HIV‐1 or HIV‐2 infection (incidence or prevalence) determined by a laboratory result.
Secondary outcome: Any adverse events associated with circumcision reported in the studies.
Search methods for identification of studies
See: Cochrane HIV/AIDS Review Group search strategy
We will attempt to identify all relevant trials regardless of language or publication status. We will search MEDLINE, EMBASE, the Cochrane Database of Systematic Reviews, and the York Database of Abstracts of Reviews of Effectiveness (DARE) for previous reviews of male circumcision for prevention of HIV acquisition in men published by by June 2009. We will obtain reports of relevant studies (i.e. those which assess the association between male circumcision and HIV acquisition in men who have sex with men) which are referred to in these reviews.
We will then carry out an exhaustive search of the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, AIDSearch and Gateway for studies published by June 2009, which assess the association between male circumcision and HIV acquisition in men who have sex with men.
We plan to use the following search strategy for MEDLINE and MBASE.
MEDLINE:
1: ("hiv 1"[MeSH Terms] OR "hiv 2"[MeSH Terms]) OR HIV
2: "hiv infections"[MeSH Terms]
3: circumcision OR circumcis* OR uncircumcis*
4: #1 OR #2 {i.e. All HIV terms}
5: #4 AND #3 {i.e. Combining All HIV‐terms AND All Circumcision terms}
EMBASE:
1: Circumcision/
2: circumcis$.tw.
3: uncircumcis$.tw.
4: or/1‐3
5: exp Human Immunodeficiency Virus/
6: exp Human Immunodeficiency Virus Infection/
7: hiv.tw.
8: aids.tw.
9: acquired immune deficiency.tw.
10: human immunodeficien$ virus.tw.
11: acquired immun$ deficiency.tw.
12: or/5‐11
13: 4 and 12
Hand searches of the reference lists of all pertinent studies found will also be undertaken. Agencies, organisations and experts in the field of male circumcision will be contacted to locate any additional studies.
Data collection and analysis
The search for studies will be performed with the assistance of the Cochrane HIV/AIDS Group. We will screen the titles, abstracts and descriptor terms from the electronic searches to create a pool of eligible studies. All identified citations will be independently inspected by each of the four review authors to establish the possible relevance of the article to the review. Full articles will be obtained for all abstracts considered relevant. Studies will be reviewed for relevance based on study design, types of participants, exposures, and outcome measures.
The methodological quality of included trials will be evaluated by the randomisation sequence generation, allocation sequence concealment, blinding of outcome assessors, and completeness and reporting of outcome data. The quality assessment will be conducted using a standardadised quality assessment form.
Data will be extracted independently by each of the four review authors using a standardadised data abstraction form. Abstracted information will include:
-
Study details: location and setting, population demographics and risk characteristics, whether participants self‐classified themselves as homoesxual or it was the researchers who classified the participantsy as homoesxual, study design, time period, population size, and attrition rate.
-
Outcome details: HIV incidence or prevalence, degree of compliance with male circumcision (where applicable), adverse effects, and types of laboratory tests used to confirm HIV diagnosis
We will attempt to obtain missing data by contacting the correspondence authors of eligible or included studies. Disagreement amongst the review authors on the eligibility and quality of trials or extracted data will be resolved by consensus.
Statistical analyses will be undertaken using RevMan 5. We will express each study result as an odds ratio (OR) with its 95% confidence intervals (CI), examine statistical heterogeneity between studies using the chi‐square test of homogeneity, and describe the statistical heterogeneity bewteen study results using the Higgins I2 statistic [Higgins 2003]. In the absence of significant statistical heterogeneity between studies, we will combine the results using a fixed‐effects method; otherwise we will use the random‐effects method. We plan to explore the cause of any significant statistical heterogeneity by subgroup analysis (with subgroups defined by recorded study characteristics)
