• Africa
• Information Tech
• Prostitution

This month we go to press amid continuing uncertainty over the future of a proposed anti-homosexuality bill in Uganda. Our publication this month of our Editor's choice study1 (see page 136) of men who have sex with men in Cameroun is timely, and highlights the vulnerability of men who have sex with men in the African region, and their lack of access to the prevention initiatives now so well established across the world—57% of participants reported unprotected anal intercourse in the past 6 months. The fear of arrest was described by a number of participants, and only just over half of the group had been exposed to prevention initiatives. In an accompanying editorial, McIntyre et al2 (see page 82) discuss the failure in prevention for men who have sex with men in Africa, and focus on the priorities for prevention that are now emerging from a growing and welcome literature.

Looking back in history through our archive, a small lesson in the value of legal toleration in disease prevention can be seen in a study that compared male homosexual new attendances in Newcastle's genitourinary clinic across the decade 1964–74.3 This period spanned the 1967 Homosexual Law Reform Act in England, which legalised homosexual acts between consenting adult men. The proportion of gonorrhoea cases in men remained constant, along with those that were homosexually acquired. However, the proportion of male patients acknowledging homosexual behaviour, and of homosexual men with gonorrhoea willing to cooperate with contact-tracing activity, increased. Before this, ‘the patient is unwilling to admit the source of his infection, both from natural reticence and from fear that his guilty secret might be passed on to the police’.4

The internet is now emerging as a setting for health care, currently less regulated than traditional relationships between patients, healthcare professionals and diagnostic services. Owens et al5 (see page 112) undertook a two-part investigation of what is offered by internet-based sexually transmitted infection testing services. They looked at what was offered at what price, and the processes for consent and return of results by US and international sites. The services were evaluated by attempts to contact them, and the accuracy of tests for gonorrhoea and chlamydial infection was then evaluated by ‘mystery shopper’ transactions. Both international and US-based services often had non-contactable or invalid contact information, and only two of 27 websites completed a survey. Failure to return a result was common, and the do-it-yourself kits were of poor quality. The authors point out the potential of the internet to contribute to health care and surveillance, and emphasise the need for concerted, strategic action to develop and accredit validation protocols.

An Indian study6 (see page 131) described the experiences of female sex workers (FSW), described through exit interviews in a region where a large-scale physician capacity-development programme had taken place. This did not appear to be better in the programme-linked clinics, despite in-service training. Although the general standard of care was thought to be adequate, partner treatment, speculum examination and HIV testing remained unsatisfactory, and provider attitudes towards FSW often remained negative and stigmatising. By contrast, Wilson et al7 (see page 117) demonstrate that the current frequency of sexually transmitted infection testing for FSW required by statute in Victoria, Australia, is not cost-effective for disease prevention.

The choice of input parameters is crucial for any model, and the paper by Burington et al8 (see page 84) creates a substantial challenge for the scientific community, in its claim that sexual behaviour surveys typically produce biased estimates of duration, due to statistical problems related to censoring, truncation, selection bias and correlated data. They provide a set of proposed solutions that will be of considerable interest both to users and producers of sexual behaviour data. Biased data relating to partnerships is also a major preoccupation of the network study of adolescent heterosexual ‘dyads’ by Yamazadi et al9 (see page 141), and shows that agreement on partnership type was poor, with a majority of adolescent couples not reciprocally nominating each other.

Another African study,10 (see page 148) like the Editor's choice with its origins in Cameroon, reports a high prevalence of unsafe sex among people living with HIV/AIDS, particularly within stable relationships. Inconsistent condom use was reported less frequently by individuals taking antiretroviral therapy. The authors emphasise the importance of developing counselling interventions for all stages of HIV care, especially those not yet receiving antiretroviral therapy.

Vernel-Pauillac et al11 (see page 106) report that a useful prediction of antibiotic susceptibility can be achieved by genotyping chromosomal genes, and that genotypes can be related to geographical origin. This could be useful in the surveillance of antibiotic resistance. Closer to the clinic, Augenbraun et al12 (see page 97) report a relationship between hepatitis C infection and biological false-positive syphilis testing.

By the time your journal falls on the doormat, podcasts on these and issues from this month's edition will be live on the website—do come and listen in, or join the debate on the blog.