CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

HIV care and treatment outcomes and whether depression mediates these relationships. Methods: We analyzed cross-sectional data from 436 women living with HIV enrolled in the Women’s Adherence and Visit Engagement (WAVE) sub-study of the Women’s Interagency HIV Study (WIHS), conducted in San Francisco, CA, Atlanta, GA, Birmingham, AL and Jackson, MS. The exposure was experienced poverty stigma, measured using 4 items from the Perceived Stigma of Poverty Scale. Outcomes were viral suppression, CD4 ≥ 350 cells/mm3, self-reported ≥ 95% adherence, and no missed HIV care visits in the past 6 months. The mediator was depression, measured by the 20-item Center for Epidemiological Studies Depression Scale. Multivariable logistic regression models were adjusted for income, age, race/ethnicity, education, non-prescribed drug use, and months taking ART. We tested whether the association of poverty stigma with the outcomes was mediated by depression scores, using indirect effects analysis with bootstrapping. Results: Each unit increase in mean experienced poverty stigma score was associated with lower adjusted odds (aOR) of viral suppression (aOR:0.79, 95% CI:0.64, 0.98), having a CD4 count ≥ 350 cells/mm3 (aOR:0.69, 95% CI: 0.53, 0.89), ≥ 95% ART adherence (aOR 0.72, 95% CI: 0.55, 0.93), and no missed HIV care visits (aOR:0.71, 95% CI:0.53, 0.95). Depression significantly mediated the negative relationship between experienced poverty stigma and having a CD4 count ≥ 350 cells/mm3 (indirect effect: -0.09, 95% CI: -0.16, -0.04; direct effect: -0.27, 95% CI: -0.31, 0.05), as well as experienced poverty stigma and ≥ 95% ART adherence (indirect effect: -0.11, 95% CI: -0.18, -0.04; direct effect: -0.17, 95% CI: -0.26, 0.04). Conclusion: Experienced poverty stigma was associated with worse HIV health outcomes, even after adjusting for income, and depression was a significant pathway for some of these relationships. Longitudinal research should assess these relationships over time. Findings support interventions and policies that seek to both reduce poverty stigma and address depression among people living with HIV. 899 RISK FACTORS FOR INCREASED HOSPITAL LENGTH OF STAY AMONG PWH, 2014-2015 Julia Fleming 1 , Stephen Berry 1 , W. C. Mathews 2 , Judith Aberg 3 , Laura W. Cheever 4 , Richard D. Moore 1 , Kelly Gebo 1 , for the HIV Research Network 1 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 2 University of California San Diego, San Diego, CA, USA, 3 Icahn School of Medicine at Mt Sinai, New York, NY, USA, 4 HRSA HIV/AIDS Bureau, Rockville, MD, USA Background: Length of stay (LOS) is an important indicator of hospital efficiency and severity of illness but can vary by geographic region. The objective of this study was to evaluate inpatient LOS among PWH by diagnostic category and to identify factors associated with increased LOS. Methods: Hospitalization data from 2014-2015 was obtained on all adults receiving longitudinal HIV care at 14 geographically diverse sites in the HIV Research Network. Modified clinical classification software from the AHRQ assigned primary ICD-9 codes into mutually exclusive diagnostic categories. Patient-specific mean LOS was used to calculate mean and median LOS per diagnostic category. Multivariate negative binomial regression analysis was used to evaluate factors associated with LOS. Results: Of 20,608 patients followed, 3196 patients were hospitalized over 4704 person-years of active outpatient care. Study subjects had a median age of 50 (IQR 43 – 58), were predominately male (67.6%) and black (50.9%), had CD4 > 200 (72.8%), and were HIV-virally suppressed (65.8%). Health care coverage was Medicaid (46%), Medicare (11.9%), private insurance (9.4%), and uninsured (12.5%). Median LOS was 5 days (IQR 3-8); mean LOS was 6.8 days (SD 9.3). Mean LOS was longest for AIDS-defining illness (ADI) (9.3 days), non-AIDS defining infections (7.4 days), and pulmonary (7.3 days). In multivariate analysis, mean LOS for ADI was significantly longer than non-ADI (aIRR vs. non-ADI, 1.16 [1.02,

897 ED VISITS AND HOSPITALIZATIONS AS OPPORTUNITIES TO IMPROVE HIV CARE ENGAGEMENT Tigran Avoundjian , Matthew R. Golden, Meena Ramchandani, Julia C. Dombrowski University of Washington, Seattle, WA, USA Background: Many health departments around the United States use HIV surveillance data to identify poorly engaged persons living with HIV (PLWH) and direct HIV care relinkage activities. The effectiveness of these Data to Care (D2C) programs has been hindered by difficulty contacting individuals who appear to be out of HIV care. Identifying opportunities, such as emergency department (ED) and inpatient (IP) admissions, to interact with poorly engaged PLWH is crucial to improving their success. In this study, we describe the characteristics, ED/IP utilization, and viral load status of PLWH seen at a UWMedicine ED/IP. Methods: We used UWMedicine’s clinical data repository and Public Health Seattle and King County’s HIV surveillance database to identify all PLWH residing in King County who had an ED/IP admission at a UWMedicine facility – one of the largest ED providers in King County– in 2017. Using HIV laboratory reporting data, we determined the HIV viral load status of patients at the beginning and end of 2017 and immediately prior to each ED/IP admissions. We compared the demographic characteristics and viral load status at the beginning and end of 2017 of patients who had at least one ED/IP admission while unsuppressed (i.e., viral load > 200) to those who had no ED/IP admissions while unsuppressed (i.e., viral load < 200). Results: In 2017, 831 PLWH had 1841 ED/IP admissions at a UWMedicine facility. Of these, 189 (23%) had at least one ED/IP admission while virally unsuppressed. Of the 189 unsuppressed patients, 134 (71%) were unsuppressed at the beginning of 2017, and 114 (60%) were unsuppressed at the end of 2017. Of the 642 patients who were suppressed during their ED/IP admissions, 47 (7%) were unsuppressed at the beginning of 2017, and 23 (4%) were unsuppressed at the end of 2017. Unsuppressed patients were younger (mean age: 42 vs 47 years) and more likely to report injection drug use compared to suppressed patients (40% vs 28%; p<0.01). Unsuppressed patients were more likely to have 3 or more ED/IP admissions compared to suppressed patients (39% vs 18%; p<0.01). Conclusion: In 2017, about 25% of PLWH who had an ED/IP admission had at least one visit while unsuppressed and 60% of unsuppressed patients remained unsuppressed at the end of 2017. ED/IP admissions provide an opportunity to interact with PLWH who experience sustained poor engagement in care. Interventions that leverage partnerships with emergency departments are needed to improve the HIV care outcomes of this population. 898 POVERTY STIGMA AND HIV TREATMENT OUTCOMES AMONG WOMEN LIVING WITH HIV IN THE US Anna M. Leddy 1 , Janet M. Turan 2 , Mallory Johnson 1 , Torsten B. Neilands 1 , Mirjam- Colette Kempf 2 , Deborah Konkle-Parker 3 , Gina Wingood 4 , Phyllis Tien 1 , Tracey Wilson 5 , Carmen H. Logie 6 , Sheri Weiser 1 , Bulent Turan 2 1 University of California San Francisco, San Francisco, CA, USA, 2 University of Alabama at Birmingham, Birmingham, AL, USA, 3 University of Mississippi Medical Center, Jackson, MS, USA, 4 Columbia University, New York, NY, USA, 5 SUNY Downstate Medical Center, Brooklyn, NY, USA, 6 University of Toronto, Toronto, ON, Canada Background: Individuals with low income experience worse HIV treatment outcomes. Although such disparities may be due to financial barriers in accessing HIV care, stigma related to poverty may be another mechanism linking socioeconomic disadvantage to poor HIV-related health outcomes. We examined whether experienced poverty stigma was associated with worse

Poster Abstracts

CROI 2019 350