Before the introduction of the universal offer, all nurses received training on HIV consent and counselling by sexual health advisors. Clinical data was prospectively recorded in an Access database, including patient demographics, travel history and HIV testing outcomes. From May 2009, this included acceptance or decline of the HIV test, reasons for declining and any HIV test result. This database was established for clinical audit and service evaluation on 26 August 2008. HIV tests requested prior to phase 1 were identified using

the Trust’s patient results database. The introduction of universal laboratory and POCT testing in our Trust was approved by the ethics Epigenetic Reader Domain inhibitor committee as being in line with the UK 2008 guidelines, and therefore service development rather than research. Comparisons of testing rates between phases, and between patient groups, were made in Epi-info v 3.5.3 (Centers for Disease Control and Prevention, Atlanta, USA) and Stata v 10.0 (StataCorp LP, Texas, USA) using either Yates-corrected or single table χ2 tests with the appropriate number of degrees of freedom. During phase 0 and phase 1, consenting patients had a venous ethylenediaminetetraacetic acid (EDTA) sample sent to the virology laboratory for initial analysis in a 4th generation HIV test. This was an Abbott HIV Ag/Ab Combo CMEIA performed on the Abbott Architect i2000SR platform (Abbott Diagnostics

Ltd, Maidenhead, England). Samples showing reactivity in the screening test received supplementary testing in a

Tyrosine Kinase Inhibitor Library mw second 4th generation assay (VIDAS® HIV Duo Ultra, BioMerieux SA, Marcy l’Etoile, France) and an antibody-only immunoassay (Orgenics Immunocomb, Clomifene Orgenics Ltd, Yavne, Israel) that permitted identification of the infection as HIV-1 or HIV-2. All patients received information on the time taken to receive a result (average 48–72 h) and how they would be informed of the result. Phase 2 started on introduction of POCT (INSTITM HIV Rapid Antibody Test; Pasante Healthcare, West Sussex, UK). This test is a visually read qualitative immunoassay able to detect antibodies to HIV-1 and HIV-2 in a finger-prick blood sample. Results are read within 60 seconds. All the clinical nurse specialists who provided this test in phase 2 undertook a 1-hour training session on the use of the test. Reactive results were confirmed according to the same laboratory protocol described above using a separate EDTA sample. There were a total of 4965 visits to the emergency open-access clinic between 26 August 2008 and 31 December 2010: 1342 in phase 0 (26 August 2008 to 31 April 2009), 792 in phase 1 (1 May 2009 to 20 September 2009), and 2831 in phase 2 (21 September 2009 to 31 December 2010). The acceptance rates of testing for HIV and the associated prevalence of newly diagnosed HIV infections are shown in Figure 1. Testing rates increased significantly across the three phases (χ2 test for trend 823.