Behavioural and demographic correlates of undiagnosed HIV infection in a MSM sample recruited in 13 European cities.

BACKGROUND: Reducing the number of people with undiagnosed HIV infection is a major goal of HIV control and prevention efforts in Europe and elsewhere. We analysed data from a large multi-city European bio-behavioural survey conducted among Men who have Sex with Men (MSM) for previously undiagnosed HIV infections, and aimed to characterise undiagnosed MSM who test less frequently than recommended. METHODS: Data on sexual behaviours and social characteristics of MSM with undiagnosed HIV infection from Sialon II, a bio-behavioural cross-sectional survey conducted in 13 European cities in 2013/2014, were compared with HIV-negative MSM. Based on reported HIV-testing patterns, we distinguished two subgroups: MSM with a negative HIV test result within 12 months prior to the study, i.e. undiagnosed incident infection, and HIV positive MSM with unknown onset of infection. Bivariate and multivariate associations of explanatory variables were analysed. Distinct multivariate multi-level random-intercept models were estimated for the entire group and both subgroups. RESULTS: Among 497 participants with HIV-reactive specimens, 234 (47.1%) were classified as previously diagnosed, 106 (21.3%) as incident, and 58 (11.7%) as unknown onset based on self-reported status and testing history. MSM with incident HIV infection were twice as likely (odds ratio (OR) = 2.22, 95% confidence interval (95%CI): 1.17-4.21) to have used recreational substances during their last anal sex encounter and four times more likely (OR = 3.94, 95%CI: 2.14-7.27) not to discuss their HIV status with the last anal sex partner(s). MSM with unknown onset of HIV infection were 3.6 times more likely (OR = 3.61, 95%CI: 1.74-7.50) to report testing for a sexually transmitted infection (STI) during the last 12 months. CONCLUSIONS: Approximately one third of the study participants who are living with HIV were unaware of their infection. Almost two-third (65%) of those with undiagnosed HIV appeared to have acquired the infection recently, emphasizing a need for more frequent testing. Men with the identified behavioural characteristics could be considered as primary target group for HIV Pre-Exposure Prophylaxis (PrEP) to avoid HIV infection. The increased odds of those with unknown onset of HIV infection to have had an STI test in the past year strongly suggests a lost opportunity to offer HIV testing.

HIV serostatus knowledge and serostatus disclosure with the most recent anal intercourse partner in a European MSM sample recruited in 13 cities: results from the Sialon-II study.

BACKGROUND: Knowledge of HIV status can be important in reducing the risk of HIV exposure. In a European sample of men-who-have-sex-with-men (MSM), we aimed to identify factors associated with HIV serostatus disclosure to the most recent anal intercourse (AI) partner. We also aimed to describe the impact of HIV serostatus disclosure on HIV exposure risks. METHODS: During 2013 and 2014, 4901 participants were recruited for the bio-behavioural Sialon-II study in 13 European cities. Behavioural data were collected with a self-administered paper questionnaire. Biological specimens were tested for HIV antibodies. Factors associated with HIV serostatus disclosure with the most recent AI partner were examined using bivariate and multilevel multivariate logistic regression analysis. We also describe the role of serostatus disclosure for HIV exposure of the most recent AI partner. RESULTS: Thirty-five percent (n = 1450) of the study participants reported mutual serostatus disclosure with their most recent AI partner or disclosed having HIV to their partner. Most of these disclosures occurred between steady partners (74%, n = 1077). In addition to the type of partner and HIV diagnosis status, other factors positively associated with HIV serostatus disclosure in the multilevel multivariate logistic regression model were recent testing, no condom use, and outness regarding sexual orientation. Disclosure rates were lowest in three south-eastern European cities. Following condom use (51%, n = 2099), HIV serostatus disclosure (20%, n = 807) was the second most common prevention approach with the most recent AI partner, usually resulting in serosorting. A potential HIV exposure risk for the partner was reported by 26% (111/432) of HIV antibody positive study participants. In 18% (20/111) of exposure episodes, an incorrect HIV serostatus was unknowingly communicated. Partner exposures were equally distributed between steady and non-steady partners. CONCLUSIONS: The probability of HIV exposure through condomless AI is substantially lower after serostatus disclosure compared to non-disclosure. Incorrect knowledge of one's HIV status contributes to a large proportion of HIV exposures amongst European MSM. Maintaining or improving condom use for anal intercourse with non-steady partners, frequent testing to update HIV serostatus awareness, and increased serostatus disclosure particularly between steady partners are confirmed as key aspects for reducing HIV exposures amongst European MSM.