Physician-initiated rapid HIV testing in an urban emergency department: comparison of testing using a point-of-care versus a laboratory model.

Abstract

We compare the outcomes of 2 models of physician-initiated rapid HIV testing in an emergency department (ED). One-year retrospective cohort comparing 2 6-month models of physician-initiated rapid HIV testing, point-of-care versus laboratory. Patients aged 12 years or older and able to give verbal consent were eligible for physician-initiated rapid HIV testing if their treating physician believed testing was clinically indicated. During the point-of-care phase, nursing staff performed oral fluid testing. During the laboratory phase, the laboratory performed whole-blood testing. The proportion of potentially eligible patients who had physician-initiated rapid HIV testing ordered (order rate), proportion of ordered tests completed (test completion rate), and proportion of potentially eligible patients who completed testing (overall testing rate) during each phase were assessed. ED length of stay and testing times were also compared. For the point-of-care versus laboratory phase, respectively, there were 24,345 potentially eligible patients versus 26,363; order rate was 3.3% versus 2.4% (P<.001); test completion rate was 75.3% versus 86.8% (P<.001); overall testing rate was 2.5% versus 2.1% (P=.009). Eighteen (3.0%) of the point-of-care-tested patients and 15 (2.7%) of the laboratory-tested patients had reactive tests (P=0.02). The total testing time was greater in the laboratory phase (88 versus 66 minutes; P<.001); however, there was no significant difference in the length of stay between phases (6.2 versus 6.9 hours; P=.15). Relatively few ED patients undergo physician-initiated rapid HIV testing regardless of whether a point-of-care or laboratory approach is used. Differences exist in most outcome measures when point-of-care and laboratory models are compared, which should be considered when testing is implemented.