Background:  In 2006, the US Centers for Disease Control and Prevention (CDC) recommended opt-out HIV screening in health-care settings (all patients 13 to 64 years of age are notified that HIV testing will be performed as a routine part of their medical care unless they decline). We evaluated factors associated with acceptance of opt-out HIV screening and with not receiving testing at the Alameda County Medical Center (ACMC) Emergency Department (ED).

Methods: In August 2007, the ACMC ED implemented voluntary, opt-out rapid HIV screening. Data on demographics, acuteness of illness, eligibility for screening, acceptance and receipt of testing, and test results were collected from all ED patients. We used logistic regression to identify factors associated with acceptance of screening among medically stable 13- to 64-year-old patients and, of those accepting, with not being tested.

Results:  During August 2007 to March 2008, there were 49,187 ED visits by 30,418 unique patients. A total of 23,159 (76.1%) unique patients were eligible for opt-out screening, of whom 8472 (36.3%) accepted testing; 5275 (62.3%) patients who accepted screening were tested. On multivariable analysis, patients who were <30 years of age (odds ratio [OR] 1.6; 95% confidence interval [CI] 1.5 to 1.7), female (OR 1.1, 95%CI 1.0 to 1.1), Hispanic or non-Hispanic black (OR 1.3, 95%CI 1.2 to 1.3), unmarried (OR 1.3, 95%CI 1.2 to 1.4), and not acutely ill (OR 1.2, 95%CI 1.0 to 1.4) were more likely to accept screening. Factors independently associated with not receiving a test among those accepting it were age <30 years (OR 1.2, 95%CI 1.1 to 1.3) and acute illness (OR 2.3, 95%CI 1.8 to 3.1). Overall, there were 25 (0.5%) new HIV diagnoses among patients tested.

Conclusions:  Age, sex, race/ethnicity, marital status, and acuteness of presenting condition were associated with acceptance of opt-out HIV screening, though their effects were modest. However, not all patients who accepted screening received testing in this busy ED. Younger patients were more likely to accept testing, yet they were less likely to receive it. Acutely ill patients were less likely to accept and receive testing. Although the proportion of new HIV diagnoses was small, it was greater than the threshold HIV prevalence for conducting opt-out HIV testing (≥0.1%) recommended by CDC, and each HIV diagnosis is an opportunity to provide care and reduce transmission. Our findings may assist ED in developing effective approaches to increase acceptance and enhance receipt of opt-out HIV screening.