Background Endemic HIV transmission in men who have sex with men (MSM) is a major concern in England. Since 2008, a new national anonymised genitourinary medicine (GUM) clinic reporting system provides follow-up data on clinic attendees.

Objective To calculate HIV seroconversion rates and identify predictors of acquisition in MSM clinic attendees to inform the development of further HIV prevention initiatives.

Methods National cohort of MSM who tested HIV negative at a GUM clinic in England in 2009 and had a follow-up test within 1 year were included in these analyses. HIV seroconversion rates (per 100 person-years (py)) with 95% CI were calculated by subgroups and risk markers. HR with 95% CI are reported for significant (p<0.05) predictors of HIV seroconversion identified using Cox regression analyses. Population attributable risk was calculated to estimate the importance of each predictor for HIV infection.

Results Among the 15 500 men who attended in 2009, there were 277 seroconversions, giving an overall incidence of 2.7/100 py (95% CI 2.4 to 3.1). Incidence was higher among MSM aged 35–49 years (3.4/100 py), of black ethnicity (4.1/100 py) and with a previous gonorrhoea or chlamydia infection (8.6/100 py and 9/100 py, respectively). In multivariable analysis, risk of acquiring HIV was higher among MSM with a previous gonorrhoea (HR: 2.4, 95% CI 1.4 to 4.1) or chlamydia infection (HR: 3.0, 95% CI 2.0 to 4.7) or who received treatment as a STI contact (HR: 1.8, 95% CI 1.1 to 2.9). Age predicted HIV acquisition in 30% of new infections and clinical risk markers from the previous year another 10%.

Conclusions Annual HIV incidence among MSM re-attending GUM clinics is very high at almost 3%. None of the clinical risk factors were important predictors of HIV acquisition. Therefore more discriminatory behavioural information is required to identify MSM at higher risk of HIV and facilitate better triaging of HIV prevention measures in GUM clinics.