CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

Methods: Two HIV rapid tests were performed at each visit; infections were confirmed with an HIV RNA assay. Samples were analyzed retrospectively at the HPTN Laboratory Center using two 4th generation assays; a qualitative HIV RNA assay; a viral load assay; Western blot; an HIV-1/HIV-2 discriminatory assay; HIV genotyping; next generation sequencing (NGS); and antiretroviral drug (ARV) testing. Results: Twelve subjects acquired HIV infection. In 4 cases, at least 1 rapid test was non-reactive at the seroconversion visit. Six subjects were not randomized (3 had acute infection at enrollment, 2 were infected during the DOT phase, 1 was not randomized due to pregnancy and was infected months later). One subject had acute infection at the end of the DOT phase but was not diagnosed for 3-4 months after randomization because HIV rapid tests were non-reactive; resistance testing failed in this case because HIV RNA levels were persistently low/undetectable. Five subjects were infected after randomization. In 4 cases, ARV test results were consistent with infrequent drug use (event-driven arm) or poor adherence (daily or time-driven arm); one woman in the time-driven arm had ARV levels prior to infection indicating PrEP use with 4+ doses/week. She developed HIV resistance consistent with TDF/FTC use (K65R +M184I). Mutations associated with TDF/FTC use were also detected after 4 weekly DOT doses in two subjects who had acute infection at enrollment (one K65R; one M184I). In 2/3 cases, resistance was only detected using NGS. Conclusion: All but one infection in this study was acquired when subjects were not on PrEP, were in the once-weekly DOT phase, or had ARV levels indicating infrequent dosing. Early infections were often missed by HIV rapid testing, but were detected using a sensitive HIV RNA assay and 4th generation tests. One subject who continued PrEP with undiagnosed infection had persistent low/undetectable HIV RNA. Resistance was detected in 2 subjects who received only 1 TDF/FTC dose/week. 956 A PRAGMATIC RANDOMIZED CLINICAL TRIAL OF RAPID HIV SCREENING IN EMERGENCY DEPARTMENTS Jason Haukoos 1 , Michael S. Lyons 2 , Douglas A. White 3 , Emily Hopkins 1 , Meggan Bucossi 1 , Sarah Pfeil 3 , Andrew Ruffner 2 , Danielle Signer 4 , Richard Rothman 4 , for the HIVTESTED Trial Investigators 1 Denver Hlth Med Cntr, Denver, CO, USA, 2 Univ of Cincinnati Coll of Med, Cincinnati, OH, USA, 3 Highland Hosp, Oakland, CA, USA, 4 Johns Hopkins Univ, Baltimore, MD, USA Background: HIV screening in emergency departments (EDs) is a major focus of national HIV prevention efforts. Screening strategies, including risk-based (targeted) and non-risk-based (nontargeted), have been endorsed by the CDC and the USPSTF; however, little is known about their comparative effectiveness. The objective of this study was to evaluate targeted and nontargeted rapid opt-out HIV screening strategies when fully-integrated into EDs with the principal hypothesis that targeted screening using the Denver HIV Risk Score (DHRS) would be superior to nontargeted screening for identifying newly-diagnosed HIV infections. Methods: Design: Prospective pragmatic randomized clinical trial. Setting: Four urban EDs in Baltimore, MD, Cincinnati, OH, Denver, CO, and Oakland, CA with an annual combined census of 300,000 visits. Population: Patients were eligible if: (1) ≥ 16 years of age; (2) not critically ill or mentally altered; and (3) not already HIV diagnosed; or (4) with an anticipated length of stay < 30 minutes. Interventions: Consecutive patients underwent concealed balanced randomization to 1 of 3 arms: (a) nontargeted HIV screening; (b) enhanced targeted HIV screening, using the DHRS as a validated HIV risk prediction tool; or (c) traditional targeted HIV screening, using conventional risk behaviors as defined by the CDC. 4th-generation HIV testing was performed based on results of screening and opt-out consent. Outcomes: Newly-diagnosed HIV infection and disease stage. Analyses: Intention-to-treat using risk ratios (RRs) with 95% confidence intervals (CIs) and patient visit as the unit of analysis. Results: During the 22-month study period, 76,561 patient visits were randomized with outstanding balance (Table). Of these, 25,469 underwent nontargeted screening, 25,453 enhanced targeted screening, and 25,639 traditional targeted screening, with a total of 14,405 completed HIV tests and 25 (0.2%) confirmed new HIV infections. Of the 25 new diagnoses, only 6 (24%) were AIDS defined and 1 (4%) was acute. The RR between the combined targeted screening strategies vs nontargeted screening and new HIV diagnoses was 0.5 (95% CI: 0.2–1.2), and the RR between enhanced targeted screening vs nontargeted screening was 0.7 (95% CI: 0.3–1.6). Conclusion: Among ED patients, targeted rapid opt-out HIV screening was not superior to nontargeted rapid opt-out HIV screening. All three strategies identified comparable numbers of newly diagnosed HIV-infected patients, although the overall prevalence of disease was low.

Poster and Themed Discussion Abstracts

957

OPT-OUT HIV/HCV TESTING AMONG JAIL INMATES Carolina De La Flor 1 , Esmaeil Porsa 2 , Ank E. Nijhawan 1 1 Univ of Texas Southwestern Med Cntr, Dallas, TX, USA, 2 Parkland Hlth and Hosp Systems, Dallas, TX, USA

Background: Incarceration provides an opportunity to provide HIV/HCV screening in high-risk and hard-to-reach individuals. The Centers for Disease Control recommends routine opt-out HIV testing in jails and prisons, however only 19% of prisons and 35% of jails offer this service. HCV testing is recommended for those born between 1945-1965 (“baby boomer” cohort) and with risk factors such us injection drug use, and incarceration. The aim of our study is to describe the results of an opt-out combined HIV and HCV testing program in a criminal justice setting.

CROI 2017 414