Background:  In the US, 20% of HIV infections are undiagnosed. In 2006 the CDC recommended expanding routine HIV screening to all healthcare settings, including Emergency Departments (ED). Rapid opt-out testing in the ED may help identify new cases but low seroprevalence rates have been found in ED programs across the US. We present results from a widespread, routine HIV screening program in 2 urban ED in southeastern US.

Methods:  A prospective cohort study was conducted at Duke University Hospital (DUH) ED and Durham Regional Hospital (DRH) ED in Durham, North Carolina, between December 2008, and August 2011. Patients were offered free HIV testing using an oral swab (OraQuick Rapid Antibody Test). Counselors approached patients in the ED who were ≥18 years old, mentally competent, not known to be HIV⁺, and not tested in the last 6 months by our program. Demographics, prior testing history, and presenting symptoms were recorded on all patients who agreed to participate. Preliminary positive results were confirmed through Western blot analysis, and patients were linked directly to care with Duke University’s Infectious Diseases Clinic.

Results:  As of August 1, 2011, 4586 patients were approached for testing at the 2 hospitals (3642 DUH; 944 DRH). Acceptance rates were excellent, ranging from 64.0% at DUH (2329/3642 accepting testing) to 69.7% at DRH (658/944). At both sites, the majority of patients approached were female (60% DUH; 63% DRH), African American (55% DUH; 70% DRH), with a prior history of HIV testing (66% DUH; 83% DRH). The overall median age was 38 at DUH and 34 at DRH (range 18 to 90 years old):  7 new positives were identified (5 DUH; 2 DRH), resulting in an overall positivity rate of 0.23% (0.21% DUH; 0.30% DRH). The mean age of the 7 newly diagnosed patients was 34.3 years. Among them, 6 of 7 (86%) had clinical symptoms suggesting late-stage infection; 4 of 7 (57%) presented with CD4 counts <200 cells/mm³. The majority were successfully linked to care with 5 of 7 (71%) currently receiving treatment at the Infectious Diseases Clinic.

Conclusions:  Routine ED-based HIV screening is feasible with high acceptance rates and successful linkages to care. Of the 7 newly diagnosed patients, 6 presented with HIV-related symptoms and could have been identified for testing based on their clinical presentation. Given the low seroprevalence, changing to a diagnostic testing strategy may be a better use of resources and increase the yield in identifying positives.