Repeat cross-sectional survey of sexual behaviour and partnerships of HIV-positive MSM attending a HIV clinic in London, UK: implications for HIV prevention among MSM

Abstract

BackgroundIn the UK in 2010, recent HIV diagnoses increased despite 80% of diagnosed HIV-positive men who have sex with men (MSM) being on antiretroviral therapy (ART) with an undetectable viral load (UVL). We examined the prevalence of non-concordant unprotected anal sex (ncUAI) in 2010 and 2000 by partnership type among HIV-positive MSM and by factors associated with ncUAI. MethodsA piloted survey questionnaire was used for data collection. In 2000, 411 men, and in 2010, 423 men, participated in the survey. Data on recent plasma viral load (detectable [DVL] or UVL); ART status (treatment naive or receiving ART); unprotected anal sex (UAI) with a boyfriend (main partner) and casual partners in the past 6 and 12 months, respectively, and most recent casual partner (only 2010); partners' HIV status (unknown or negative, or positive); attitudes towards ART for HIV prevention; and ART adherence were collected. ncUAI was defined as UAI with unknown or negative HIV status partner or partners. Odds ratios (ORs) for 2010 compared with 2000 for ncUAI with casual partners and boyfriend were derived by multiple logistic regression (MLR) adjusted for age and DVL or UVL. MLR was done to examine factors associated with ncUAI with a boyfriend and the most recent casual partner in 2010. This analysis was restricted to men reporting a HIV-negative or unknown status partner. FindingsCompared with 2000, 2010 respondents were more likely to be on ART with UVL (324 of 428 [76%] vs 163 of 411 [40%]) and less likely to be on ART with DVL (26 [6%] vs 128 [31%]) or treatment naive (80 [19%] vs 120 [29%]; overall p<0·0001). They were older (age ≥45 years 197 [46%] vs 90 [22%]; p<0·0001) and more likely to report a HIV-positive boyfriend (88 of 221 [40%] vs 51 of 203 [25%]; p=0·001). ncUAI with a boyfriend did not change (38 of 212 [18%] vs 20 of 178 [11%]; p=0·63). ncUAI with casual partners increased (98 of 288 [34%] vs 51 of 300 [17%]; OR 2·79, 95% CI 1·93–4·04; p<0·0001). In 2010, 59 (17%) of 344 men had missed at least one ART dose in the past 2 weeks. One in five men believed that being on ART with an UVL eliminates risk of HIV transmission during UAI. Duration of partnership, recreational drug use during sex, and belief that UVL eliminates the risk of HIV transmission were all independently associated with ncUAI with a boyfriend. Disclosure of HIV status and recreational drug use during sex were independently associated with ncUAI with the most recent casual partner. Viral load status was not associated with ncUAI. InterpretationPrevalence of ncUAI with casual partners has increased and with a boyfriend remains high. Men engaged in ncUAI irrespective of viral load status, suggesting that early initiation of ART could reduce the risk of HIV transmission. Behavioural interventions addressing recreational drug use, attitudes towards ART for HIV prevention, and adherence to ART should be integrated with routine clinical care. About 11% of eligible patients declined to participate. Nevertheless, age and HIV-related health indicators of the participants are similar to the Survey of Prevalent HIV Infections Diagnosed participants. The association between attitudes towards HIV transmission risk reduction strategies and sexual behaviour with primary partner might be due to post-hoc rationalisation of risky sexual behaviour, the prevalence of which might be overestimated and should be interpreted with caution. FundingOverseas Research Student Awards, University College London; Brighton and Sussex Medical School; Charitable Trustees Fund, University College London; and the Medical Research Council. None of the authors have received any funding from a pharmaceutical company or other agency for writing this article or for this project.