Background: To identify the estimated one-third of HIV+ patients (pts) in the U.S. who are undiagnosed, CDC guidelines urge routine HIV Counseling, Testing and Referral (HIV CTR) in hospitals with an HIV prevalence of 1%. This is rarely done. Our objectives were: 1) to establish a routine, voluntary HIV CTR program called “Think HIV”; 2) to identify and refer to care undiagnosed HIV+ pts; and 3) to determine the prevalence of undiagnosed infection in Massachusetts (MA) urgent care centers (UCCs).

Methods: We instituted Think HIV in the UCCs of 4 MA hospitals with the highest rates of HIV reporting. Think HIV offered voluntary, confidential HIV CTR using the Orasure HIV-1 Western Blot Kit (Epitope, Inc., Beaverton, OR) to all pts in these UCCs. Follow-up and referral of HIV+ pts was established. The Think HIV program results were compared to HIV test results done in the same period statewide (by self or physician referral) at anonymous or confidential HIV CTR DPH sites.

Results: From 1/02–8/02, Think HIV offered HIV CTR to 5,893 people of whom 1,853 were tested (acceptance rate 31.4%). At the time of this analysis, 1,674 individuals had completed results. In the same period, 24,452 tests were done at DPH sites statewide. Think HIV identified 37 new HIV diagnoses compared to 240 statewide (HIV prevalence 2.2% vs 1.0%, p < 0.0001). We indexed pts into those requesting HIV testing for documentation (e.g., pregnancy, insurance), self-described high HIV risk (e.g., HIV⁺ partner, IDU), or self-described low HIV risk. Think HIV tested 1,343 people (80.2%) vs 3,124 (12.8%) statewide who had self-described low HIV risk and found 31 vs 28 HIV infections, respectively (HIV prevalence 2.3% vs 0.9%, p < 0.0001). In Think HIV, 733 (43.8%) vs 10,275 (42.0%) statewide identified their only HIV risk as having a single heterosexual partner in the last 3 years, of whom 12 and 92 were HIV⁺, respectively (HIV prevalence 1.6% vs 0.9%, p = 0.0285). Among those who report a negative HIV test in the previous year, Think HIV had an incidence of 0.20 per person month vs 0.08 per person month statewide, p = 0.0745.

Conclusions: Think HIV, an urgent care center-based routine voluntary HIV CTR program, had a significantly higher yield than regular self-referral testing. Routine HIV testing will lead to identification of HIV-infected persons who do not describe themselves as high risk for infection. Further efforts to expand such programs nationally are essential.