Effect of Routine Opt-Out HIV Screening on Emergency Department Operational Metrics: Results From the Pragmatic Randomized HIV TESTED Trial

Abstract

Study objective: To assess the effect of routine opt-out HIV screening on emergency department (ED) patient throughput.

Methods: We performed a multicenter, prospective, pragmatic 3-arm, parallel-group, randomized clinical trial to evaluate the effectiveness of 3 triage-driven, opt-out HIV screening strategies-nontargeted, enhanced targeted, and traditional targeted screening-when fully integrated into usual ED practices. As the 3 screening approaches were performed concurrently, the trial was nested in an equivalent time samples quasi-experiment to allow for the evaluation of HIV screening on ED operational metrics. Hierarchical multivariable linear and log binomial regression were performed to estimate associations between opt-out HIV screening and patient wait time, length of stay, boarding time, door-to-door time, and left before treatment was complete.

Results: From January 19, 2014, through July 7, 2016, a total of 377,392 patient visits were included, 125,743 (33.3%) of which were included from intervention periods. Among these visits, 65,093 (51.8%) were randomized into the HIV TESTED (HIV Testing Using Enhanced Screening Techniques in Emergency Departments) trial with 39,720 (61.0%) ultimately offered opt-out HIV testing. Of those offered HIV testing, 17,485 (44.0%) did not decline testing and 13,314 (33.5%) tests were completed. HIV screening was associated with an increase, on average, in wait time of 1.4 (95% confidence interval [CI] 0.3 to 2.6) minutes when adjusted for mode of arrival, acuity, overall crowding, time of day, and day of the week. HIV screening was negatively associated with length of stay for admitted patients (-18 [95% CI -30 to -4] minutes) and door-to-door time (-18 [95% CI -24 to -6] minutes), whereas HIV screening was not associated with length of stay for nonadmitted patients (-6 [95% CI -12 to 6] minutes) and boarding time (-6 [95% CI -42 to 24] minutes). Finally, HIV screening was associated with an increased prevalence of left before treatment was complete (prevalence ratio: 1.10, 95% CI 1.06 to 1.15) when adjusted for the same covariates.

Conclusions: When fully integrated into clinical ED care, routine opt-out HIV screening was associated with small but nonclinically meaningful increases in wait time and left before treatment was complete and no significant change in length of stay, boarding time, or door-to-door time.

Keywords: Clinical operations; HIV screening; Nontargeted screening; Targeted screening.