The analysis of the entry questionnaire data from this MSM-friendly Geneva-based VCT facility from 2005 to 2009 has shown an increase in the number of clients and of HIV tests performed, presumably due in part to changes in the organisation of the consultation (acceptance of non-MSM clients, change in business hours, and no need for appointment). The prevalence of new HIV-positive tests among all tests performed was about 2% each year. Clients were a population at particular risk of HIV/STI that had previous experience with HIV testing. The three main reasons for testing were having been exposed to a sexual risk, wanting a routine test, intending to stop using condom in the context of a new steady relationship.
Checkpoint’s clientele increased rapidly since its opening; the number of clients more than doubled between 2005 (n = 245) and 2009 (n = 574). The proportion of new positive HIV tests among all tests performed varied from 1.9% in 2009 to 3.0% in 2007, which is high and similar to the proportion found in the other MSM-specific VCT facilities in Switzerland (3.4% in 2006 in Zurich) . In Amsterdam Checkpoint in 2002–03, the prevalence among MSM was 5.2% and the average prevalence 2.8% .
Checkpoint has a good detection rate. In 2009, 28% (9/32) (2005: 10%, 2006: 8.7%, 2007: 11.1%, 2008: 12.1%) of all new HIV cases among MSM living in the canton of Geneva were detected at Checkpoint , though a survey conducted among MSM in 2009 in Switzerland (national monitoring) found that most MSM in Geneva had their last test performed by their physician (43.3%) or at other non-MSM-specific VCT facilities of the canton (11.7%) or elsewhere (25.8%). Only 17.5% of MSM reported that they had their last test at Checkpoint .
Roughly 90% of MSM tested at Checkpoint had been tested prior to the visit, with a lifetime median number of three prior tests.
The median number of partners in the previous 12 months with whom AI had been practiced (between three and four over the years) and the stable and high proportion (about 30%) of MSM Checkpoint clients reporting UAI with one or more partners of different/unknown HIV status in the previous 12 months suggest that Checkpoint is able to attract clients at high risk of HIV infection. These figures were lower (median of two AI partners and 25% having UAI with partner of different/unknown HIV status) among the Geneva respondents to the national MSM behavioural survey conducted in 2009 . However, we presently underestimate the real proportion of people concerned by this risk exposure (low estimate) because we calculated the proportion without excluding clients who had no anal or no sexual relationship in the previous 12 months and who gave no answer to this question. The capacity to attract high-risk populations was also demonstrated for the similar MSM-specific facility in Zurich .
The variety of testing situations encountered in an MSM-specific VCT facility like Checkpoint is illustrated by the distribution of reasons for testing. The three major reasons, accounting for approximately 85% of all reasons, were having been exposed to a sexual risk of HIV transmission, “routine” testing, and wanting to ascertain HIV status in the context of a new sexual relationship. As expected, clients mentioning sexual risk as the main reason for testing were those who had a higher proportion of UAI with a partner of different/unknown HIV status; they were also younger, reported more frequent use of alcohol or drugs before or during sex, and were more likely to report recently feeling sad/depressed. These characteristics have been reported in other studies .
Clients who reported “routine testing” as main reason for testing had the lowest proportion of UAI with a partner of different/unknown HIV status, which contrasts with previous results in studies on this topic [28–30]. This may be due to our logic algorithm. However, a non-negligible proportion of ”routine” clients (12.5%) reported UAI with a partner of different/unknown HIV status, and their median number of AI partners was quite high. “Routine” testers are mainly those who come in each year for testing even if they report having taken no sexual risk; they may be in a steady relationship, have not been tested for a while, or want to be reassured (H. Crevoisier, nurse in charge of the facility, personal communication). This population probably deserves special attention and counselling that includes confrontation with their level of risk and assessment of possible psychological suffering.
Clients mentioning the ascertainment of HIV status in the context of a new relationship, including the possibility of stopping condom use, as the main reason for testing also require specific attention and counselling. This group is the youngest and reports the fewest number of partners, but also significant risks and antecedents of STI.
The procedure used at Checkpoint, i.e. answering a computer questionnaire on risks and reasons for testing before the consultation, allows simplification and reinforcement of the individualisation of the counselling, which is especially necessary in the population displaying a wide variety of risk management and risk reduction strategies [31–34]. The “Chat 2009” survey , a retrospective qualitative study focusing on the context of new HIV cases in Switzerland, also showed that people are infected in very different ways, which deserves a more specific, personalized approach to risk taking, especially for MSM. Pre-test and post-test counselling may play an important role in helping people better manage the situations in which risks are taken.
The quality of VCT provided at Checkpoint was confirmed in a 2009 national behavioural survey of MSM in Switzerland: among mentioned facilities where MSM go for HIV testing, 79.2% of those having done their last HIV test at Checkpoint Geneva mentioned having received pre-test counselling and 87.5% post- test counselling, the highest rates among all facilities; corresponding figures were 31.5% and 22.2% for test made at a GP's office .
Our study has limitations: clients coming back after at least 6 months after a previous visit are considered new clients and complete a new entry questionnaire. Consequently, in the analysis of reasons for testing, as the data from several years are pooled together, the population of reference is the total number of testing situations and not the persons tested since some clients may be included several times.
Another limitation is related to the fact that we conducted secondary analysis of already existing data. The questionnaire used in Checkpoint was designed for counselling purposes and not primarily for research purposes, therefore some questions were not very detailed: for example only one item was used to identify depressive symptoms or psychoactive drugs use; another example is the lack of definition of what constitutes a relationship with a steady partner, left to client's appreciation. However, as literature has shown that the perceived qualification of the relationship is relevant for behavioural adjustment [36, 37], we feel that the information collected is valid.