HIV screening programs in US emergency departments: a cross-site comparison of structure, process, and outcomes

We present findings from a multisite evaluation that systematically compares HIV screening programs in 6 emergency departments (EDs).

From 2007 to 2008, we collected previous-year data on structural factors, process attributes, testing outcomes, and cost-effectiveness from 6 ED HIV testing programs operating for 6 months or longer. We administered questionnaires to program directors, conducted site visits, and interviewed key informants.

HIV care providers (n=3 sites), emergency physicians (n=2), or health departments (n=1) initiated the testing programs. ED leadership and providers helped design and implement the programs (n=5), and emergency physicians or administrators provided daily oversight (n=5). Testing strategies included targeted (patients selected from at-risk populations; n=2), nontargeted (patients selected without regard to risk or intention of testing all; n=3), and universal (all patients selected; n=1) screening. Testing was conducted by supplemental staff (n=4) and existing hospital staff (n=2). ED testing programs were funded by grants (n=3), city HIV prevention/care budgets (n=2), or the hospital (n=1). The median percentage of census tested was 4.7% (range 2.1% to 8.4%). The median rate of preliminary positive test results was 1.2% (range 1.0% to 7.3%). The median confirmed new HIV diagnosis rate was 0.9% (range 0.8% to 6.4%). The median linkage to care rate was 92.0% (range 50% to 100%). The median cost per patient receiving a new diagnosis and linked to care was $10,200 (range $3,400 to $12,300).

Although structure and process of screening programs varied across EDs, outcomes were similar, which suggests that with current ED environments, testing methods, and resources available, the capacity and structure to increase testing in EDs has limits. These ED HIV screening programs were cost-effective according to standard thresholds.