Hepatitis C Screening Among Persons Experiencing Homelessness in the Emergency Department: A Secondary Analysis of the Determining Effective Testing in Emergency Departments and Care Coordination on Treatment Outcomes (DETECT) for Hepatitis C (Hep C) Screening Trial.

Abstract

Emergency departments (EDs) are well positioned to screen for hepatitis C virus (HCV), particularly among persons experiencing homelessness who are at increased risk for HCV and are traditionally underserved in health care settings. We examined associations between homelessness and HCV test offer, acceptance, seropositivity, and viremia as part of ED screening. We performed a secondary analysis of data from Denver Health Medical Center (Denver, Colorado), one site in the DETECT Hep C Screening Trial, a multicenter randomized pragmatic clinical trial of nontargeted versus targeted screening in EDs. Stratified by the 2 screening strategies, we performed multivariable log binomial regression while adjusting for demographics, clinical severity, arrival method, payor, and, in targeted screening, HCV risk factors. A total of 67,223 visits were included, with 13.5% persons experiencing homelessness (n=4,539/33,626) in the nontargeted group and 13.7% persons experiencing homelessness (n=4,614/33,597) in the targeted group. Compared to housed patients, persons experiencing homelessness more commonly accepted testing (21.7% [n=895/4,122] versus 15.7% [n=4,044/25,752] in nontargeted and 30.2% [n=540/1,788] versus 23.2% [n=1,600/6,895] in targeted) and were more commonly HCV seropositive (14.2% [n=91/643] versus 2.6% [n=65/2,479] in nontargeted and 17.6% [n=76/432] versus 6.0% [n=72/1,198] in targeted). In multivariable regression, homelessness was not associated with test offer in either screening group. In nontargeted screening, homelessness was associated with test acceptance (adjusted risk ratio [aRR] 1.31, 95% confidence interval [CI] 1.22 to 1.41), but this association was not observed in targeted screening (aRR 1.05, 95% CI 0.97 to 1.14). In both nontargeted and targeted screening, homelessness was associated with HCV seropositivity (aRR 2.91 [95% CI 2.06 to 4.13] and aRR 2.07 [95% CI 1.46 to 2.93], respectively) and viremia (aRR 3.33 [95% CI 2.12 to 5.24] and aRR 3.70 [95% CI 2.14 to 6.39], respectively). Compared to housed individuals, persons experiencing homelessness were more likely to agree to nontargeted HCV screening and equally likely to agree to targeted HCV screening, suggesting that nontargeted HCV screening in the ED may better support HCV detection among persons experiencing homelessness. Furthermore, persons experiencing homelessness had higher risk of HCV seropositivity and viremia regardless of screening strategy and HCV risk factors.