Deciding how to target tests to populations is a major theme this month. The balance between speed of diagnosis, sensitivity and specificity, predictive value, and treatment rates is ever-changing. In this month's Editor's Choice, Lusk et al1 (see page 227) present Australian data suggesting that Trichomonas vaginalis may be substantially underdiagnosed with existing policies for the discretionary use of wet mounts. As screening strategies for cervical dysplasia evolve—for example, the age at which cervical screening starts in England is now 25 years—we can expect to see changes in the epidemiology of this parasite, which has been well controlled in recent decades. What are the implications for testing patterns? The trade-off between rapid diagnosis and sensitivity in the diagnosis of syphilis is nicely demonstrated by Mishra et al2 (see page 193) in an evaluation of a rapid test. In a setting in which many women did not return for test results, an appropriate treatment rate of 68% was achieved with this imperfect test, by contrast with only 48% among women having laboratory serological testing. In another paper, Johnson et al3 (see page 217) describe patterns of coverage and positivity in England's National Chlamydia Screening Programme, demonstrating a higher yield of positives in healthcare settings, and particularly low positivity among university students. All these studies are important, but they need to be complemented by high quality economic studies.

The debate over who should be given which Human Papilloma Virus (HPV) vaccine looks set to continue. Last year we published a paper by Fairley et al4 showing the decline of genital warts in Australia following the introduction of quadrivalent vaccine, a paper that was acclaimed as worldwide paper of the year at this year's British Association for Sexual Health and HIV conference (see the STI conference blog for details of other papers of the year chosen by Steve Taylor). Pirotta et al5 have followed on from the epidemiology with an estimate of the healthcare costs of genital warts in the same country (see page 181), estimating these at A$14 million per year. Again, more information on cost-effectiveness is needed, which will need to recognise the differences between settings. What are we to make of data presented by Müller et al6 (see page 175) showing an association between HIV positivity and multiple HPV types? Even in resource-rich settings, men have poor awareness of HPV, as reported by Reiter et al7 (see page 241). Some problems, however, may not be related to HPV—Nasca et al8 report being unable to detect HPV in biopsies of erythroplasia of Queyrat (see page 199).

Clinicians will also be interested in a genotypic study that explores the relationship between APOE alleles and HSV (see page 202).9 Mucosal shedding was not related to APOE genotype, but self-reported clinical oral lesions were.

The potential of men who have sex with both men and women in Bangalore to act as a bridge for the transmission of HIV is interestingly reported by Phillips et al10 (see page 187). The authors acknowledge the complexity of gender identity in India, and recruited their sample in various sites where men seek various kinds of sex with men. Of 357 men, 196 had ever had sex with a woman, of whom 146 had done so in the past year. The authors conclude by emphasising the range of homosexual and bisexual behaviours and its potential importance in preventing HIV transmission.

Many other interesting papers are worth a read. For epidemiologists, we particularly recommend a modelling paper on Periodic Presumptive Treatment11 (see page 163) along with its commentary12 (see page 161), and an important paper extending uncertainty analysis approaches to STI prevalence data from multiple sources (see page 169).13