Introduction

Voluntary counselling and testing (VCT) for HIV is an important tool in the control of the HIV epidemic in developed countries, where the majority of individuals diagnosed with HIV present in this way. In the United Kingdom, such tests are readily available at genitourinary medicine (GUM) clinics, where the promotion of routine VCT for HIV is a strategic priority.¹ In spite of this, around 40–50% of HIV seropositive individuals who attend UK GUM clinics remain undiagnosed, and despite well publicised advances in the management of HIV this proportion did not change between 1996 and 1998.¹ People attend GUM clinics with a variety of sexual health problems and many may not have considered taking an HIV test before the subject is raised during the consultation. Whittall Street Clinic, a large urban GUM clinic in the West Midlands, United Kingdom, has a “universal offer” policy for HIV tests: clinic protocol requires doctors or health advisers seeing new patients to ask about and record risk factors for HIV infection, and offer information about HIV testing to all regardless of risk. This policy is audited with regular case note review, which gave rise to initial concerns about differences in HIV test uptake between different practitioners. This problem has been well described in antenatal clinics where whether patients accept an HIV test is significantly affected by the person offering the test.² We wished to investigate the extent of this effect in our setting. The local seroprevalence of HIV infection in GUM outpatients as assessed by anonymous unlinked testing at the time was 3.25% in homosexual or bisexual men, 0.21% in heterosexual men, and 0.13% in heterosexual women.¹Sex Transm Inf 2001;77:143–144

Subjects, methods, and results

We studied case records of 500 male and 500 female patients who attended the Whittall Street Clinic for the first time ever from 1 January 1998 to 10 March 1998. We recorded demographic details, standard risk factors for HIV infection, the patient's primary reason for attendance, and which of 22 different clinicians they saw. We excluded from subsequent analysis 142 patients (14%) whose stated reason for attendance was specifically to obtain an HIV test, and examined factors influencing uptake of an HIV test in the remaining 858 patients (the “opportunistic group”). The proportion of patients undergoing an HIV test in each group were compared with the χ² test, and odds ratios adjusted for age, sex, ethnic group, marital status, grade of doctor (part time or full time), presence of genital symptoms, and presence of risk factors for HIV infection were calculated by logistic regression analysis using spss 8.0.

Of 858 patients offered an opportunistic HIV test, 259 (30.2%, 95% confidence interval 27.1% to 33.4%) were tested. Three patients (1.2%, 0.2% to 3.3%) tested HIV seropositive, compared with just one (0.7%, 0.02% to 3.9%) of the 142 who had attended specifically for an HIV test. One patient found to be HIV positive through opportunistic testing had disclosed no identifiable risk factors. Individual doctors varied greatly in the proportion of patients they saw who actually took an HIV test (range 11% to 61%, χ² for trend 62.3, p<0.001). In a multivariate model, variables that significantly predicted lower uptake of HIV test were being married, female, aged 20 years old or less, having no identifiable risk factors for HIV, presenting with symptomatic sexual infection, and seeing a doctor who only worked part time in GUM (table 1). Patient ethnic group did not influence test uptake.

Discussion

Most new cases of HIV in the study period were diagnosed because patients decided to take an HIV test during their clinic attendance. This highlights the value of discussing HIV during the GUM consultation, even where no risk factors are immediately disclosed by the patient, or HIV testing is not the reason given for attendance. A contemporaneous British Co-operative Clinical Group case note review conducted in 53 UK GUM clinics found that of 3710 GUM attenders not directly requesting an HIV test, just 33% were offered a test and only 6.9% were actually tested.³ Our policy of a universal HIV test offer in Whittall Street Clinic therefore resulted in much greater HIV test uptake than in some UK GUM clinics.

There is no agreement within the specialty as to what proportion of GUM outpatients should be tested for HIV infection, but promotion of VCT remains a national strategic priority.¹ Normalising HIV testing undoubtedly increases uptake: a typical US public STD clinic reported that 79% of attenders were tested for HIV.⁴ GUM centres should strive to reduce barriers to taking an HIV test within the clinic by discussing and offering HIV tests to all, and minimising inconvenience for the patient should they decide to take a test. Dispensing with HIV pretest discussion for patients assessing themselves as low risk and offering all patients clear written material on HIV infection modestly increased HIV test uptake in one UK GUM clinic.⁵

We found a huge variation between individual doctors in HIV test uptake by their patients in spite of all working to the same protocol in a setting with good HIV awareness. Doctors who test a low proportion of their patients may not be “bad” doctors, in that they may better target testing to those at higher risk of HIV infection. Because of the low number of HIV seropositive individuals we were unable to determine which doctors were better at detecting HIV positive patients, as opposed to simply achieving an HIV test. Such wide variation within a single GUM clinic by staff who are highly aware of HIV highlights the difficulty faced in instituting broader HIV testing programmes in antenatal clinics and beyond.