We read with interest the article by Briggs et al.1and associated commentary by Bates2 suggesting a widening of HIV testing venues, to include termination of pregnancy services (TOP) and colposcopy services.

The Homerton Hospital, East London, UK has been performing opt-out HIV testing in the TOP unit since April 2008 and has an uptake of 60%.3 The HIV prevalence is 0.3%, higher than the genitourinary medicine or antenatal clinic. We initiated HIV testing in the colposcopy unit in September 2010 and report on the preliminary findings here.

All women attending the colposcopy unit of Homerton Hospital between 1 September 2010 and 28 February 2011 were offered opt-out HIV testing at the time of their colposcopy appointment. HIV testing was performed by point of care testing (Insti™ HIV Antibody Test, Pasante Healthcare, Hove, UK) and results were given immediately. Any woman with a reactive result was seen at the sexual health clinic on the same day for confirmatory serology and onward care. A total of 687 women were seen, of whom 136/687 (20%) had cervical intraepithelial neoplasia grade 2/3 (CIN 2/3).

The uptake of HIV testing is presented in Table 1. A total of 518/687 (75%) of the women accepted HIV testing and 14/687 women were already known to be HIV positive; 93/169 (55%) of the women who declined had previously had an HIV test, 39 (42%) of whom had tested in the preceding 3 months. Untested women were twice as likely to accept HIV testing than those who had previously been tested [odds ratio (OR) 2, 95% confidence interval (CI) 1.4–2.9].

One woman was newly diagnosed with HIV. Her cervical biopsy revealed CIN1. Her baseline CD4 was 280 cells/µl.

In total, 15/687 women were HIV positive, four of whom had CIN 2 or 3, five had CIN 1 and six had no CIN.

This study demonstrates a high uptake of HIV testing. Women who had never previously had an HIV test were significantly more likely to accept HIV testing in the colposcopy unit than women who had previously received an HIV-negative result.

The majority of women with HIV did not have high-grade CIN. This implies that if British HIV Association standards were adopted for HIV testing in colposcopy, rather than providing routine opt-out testing for everyone, we would have missed three quarters of the HIV diagnoses.

Since the widespread use of highly active antiretroviral therapy, two other London units have commenced HIV testing in colposcopy units, using either point-of-care testing4 or conventional serology.5 These demonstrate similar levels of uptake, with uptakes of 59–71%. However, 41% of women initially accepting HIV testing declined to wait for phlebotomy,5 suggesting that point-of-care testing increases the coverage.

It is feasible that areas with lower HIV prevalence may find this approach less efficacious and further studies should be done in areas with different prevalence. It may be effective for the clinician referring to colposcopy to perform the HIV test. Paradoxically, areas with lower HIV prevalence may have less well-established background HIV testing programmes and may benefit from an HIV testing service. Further research should investigate the relative efficacy of women having abnormal smears having an HIV test performed at the time of referral to colposcopy, with the approach of delivering an HIV test at the time of colposcopy.

This demonstrates that routine HIV testing embedded into colposcopy and TOP services is feasible and effective. The high uptake, especially among women who had never previously had an HIV test, suggests that this approach may target an at-risk population who might not attend for HIV testing in more traditional settings. When combined with the high positivity rate we concur with Bates and Briggs et al.'s conclusions of promoting more widespread HIV testing in TOP and colposcopy services.1 2