Late diagnosis of HIV remains a problem in the UK and it has been estimated that 43.6% of men who have sex with men (MSM) have never tested for HIV (28.1% of MSM resident in Brighton and Hove primary care trust (n = 309)).1 2 The Department of Health set up pilot projects in 2004 to implement and evaluate rapid HIV testing in community based organisations (CBOs) in a number of UK sites targeting high-risk groups. In Brighton, the pilot targeted MSM and we set out to determine whether MSM using fasTest differed from those attending genitourinary medicine clinics (GUM) in their demographics, sexual behaviour, prior use of sexually transmitted infection (STI)/HIV testing, HIV prevalence, rates of referral to HIV services and stage of HIV infection at diagnosis.

METHODS

Brighton fasTest is a weekly evening session staffed by GUM professionals in a CBO premise (Terence Higgins Trust). The Abbott Determine HIV1&2 kit (Abbott Laboratories, Abbott Park, Illinois, USA) was used giving results within an hour. Men positive for HIV underwent immediate confirmatory blood sampling and were fast-tracked to GUM for results and referral.

A questionnaire assessing demographics, previous use of GUM/HIV testing and sexual behaviour was completed by all men using the Brighton fasTest for a prospective evaluation of all fasTest pilots.3 A simplified questionnaire was completed by MSM accessing GUM over the same time period. Follow-up and baseline investigation (including STI screening and Serologic Testing Algorithm for Determining Recent HIV Seroconversion (STARHS) test)4 in GUM/HIV services were monitored for all men testing HIV positive. Men were included in the analysis if they had a gay or bisexual identity or reported sex with a man in the last year, tested for HIV at this visit and received a result. Men with a previous HIV diagnosis were excluded.

Data were collected from November 2004 to March 2006 and analysed using SPSS version 15.0. Categorical variables were compared with χ² and continuous variables with one way analysis of variance (ANOVA). Significant differences between sites were explored using logistic regression for multivariate analysis.

RESULTS

CBO premises performed 338 HIV tests in 316 clients over 14 months of evaluation. Altogether, 294 questionnaires were completed (response rate 93.4%). Of these, 280 men (95%) met inclusion criteria. In GUM, 271 men completed the questionnaire (response rate 12.6%) of which 187 met inclusion criteria. Reasons for exclusion were incomplete questionnaires, no evidence of sex with men or not testing for HIV (in GUM).

Altogether, 3.2% of fasTest respondents and 9.1% of GUM respondents were diagnosed HIV positive (table 1). All positive fasTest clients were referred to GUM, confirmed to have HIV and linked into HIV services. Baseline HIV parameters, including those suggesting recent infection,4 and rates of other STIs were similar between groups (additional STIs in 4/9 fasTest clients and 10/17 GUM clients).

FasTest clients were younger and less likely to test positive for HIV (table 1).

GUM clients were 1.74 (95% CI 1.064 to 2.845) times more likely to have tested for HIV previously than were fasTest clients. However, previously tested men were older than those never tested and GUM clients were older than fasTest clients. In a logistic regression controlling for age, GUM clients were not significantly more or less likely to have previously tested than fasTest clients.

GUM clients were 3.01 (95% CI 1.312 to 6.908) times more likely to receive a positive result than fasTest clients. Receiving a positive result was not associated with having previously tested for HIV or with age.

DISCUSSION

Brighton fasTest successfully tested high-risk MSM with a substantial minority being previously untested (as in GUM). The men attending Brighton fasTest were younger than those attending GUM. MSM in both sites reported high rates of risky sexual behaviours. The HIV prevalence was lower at fasTest than GUM, which may be explained by MSM attending GUM as a result of partner notification or referred by clinicians suspecting HIV infection. This contrasts with data from the USA but service provision there differs greatly to that in the UK.6 All those with HIV were successfully followed up, which compares favourably to other studies.3 7 Those identified with HIV at fasTest were not diagnosed earlier than those in GUM.

The strengths of this study are the use of well-validated questionnaires and the ability to link to HIV follow-up data. There was a low response rate in GUM, but the demographics of those participating were not significantly different to other attendees; we have no reason to believe participants were selected in any way and the response rate was over 50% in the latter period of recruitment.

Brighton fasTest increases choice for MSM wanting to test for HIV and fits well with moves to deliver STI care in the community.8 However, fasTest cannot offer tests for other STIs—a concern given the high rates observed in the HIV positive men. The estimated cost of a test at fasTest is similar to that of a STI screen (HIV/syphilis/chlamydia/gonorrhoea/hepatitis) under the national GUM tariff.3 9 Commissioners must balance these issues of choice, value and quality when reviewing service configurations.

In conclusion, HIV testing in a CBO is feasible, reaches high-risk MSM and can increase capacity, access and choice for this population as an addition to GUM services.