CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

temporal trends in cause-specific mortality over ~30 years among HIV-infected and uninfected PWID. Methods: Mortality was ascertained in the AIDS Linked to the IntraVenous Experience (ALIVE) cohort of PWID from 1988 to 2015 through linkage to the National Death Index. Deaths were classified into HIV and infectious related (HIV-deaths), chronic conditions (CC-deaths), and overdose and drug-related (drug-deaths). All-cause and cause-specific mortality rates were assessed for each calendar year. Figure displays HIV-deaths against the median CD4+, CC- deaths against the median age, and drug-deaths against the % of current IDUs within the cohort. Drug abuse information was further assessed through the Drug Abuse Screening Test (DAST-20). Results: Of 4,794 participants (25.8% HIV-infected) contributing 75,327 person-years (pys), there were 2,070 deaths. The median age increased from 34.6 in 1988 to 44.2 in 2015. A higher proportion of those who died were male, black, less educated, and HIV or hepatitis C infected. All-cause mortality increased from 15.4 per 1000-pys in 1988 to 31.4 per 1000-pys in 2001, with a subsequent plateau through 2015 (32 per 1000-pys). HIV-deaths peaked at 19.6 per 1000-pys in 1994 then continuously declined to 5.7 per 1000-pys in 2015. CC-deaths increased from 2.7 per 1000-pys in 1988 to 17.7 per 1000-pys in 2015. drug-deaths peaked at 11.1 per 1000-pys in 1998, had a nadir of 3.8 per 1000-pys in 2012, then most recently doubled to 7.9 per 1000-pys in 2015. Though active injection drug use declined overall in this period, moderate/ severe drug abuse based on the DAST increased after 2013, coincident with increased drug-related deaths. Conclusion: Besides the expected decline in HIV-deaths due to ART, the rising rates of chronic disease mortality demonstrate the critical importance of chronic disease management in this aging population. Though drug-related deaths initially declined with decreasing active injection, a recent resurgence in drug- related deaths is likely related to non-injection drug use and prescription opioids and critically requires prompt intervention for this vulnerable population.

were used to identify associations between clinic-level variables and time to death. Results: Among 6,608 participants, 292 (4.4%) died from 2011-2016; 1.06 deaths per 100 person-years. Deaths were among males (71.6%), blacks (81.8%), and mean age at death was 56.5 yrs. From death certificates, 25% of deaths were HIV-related, 13% cardiovascular disease related and 11% due to non-AIDS related cancers. Median time from HIV diagnosis to death was 14.3 years (IQR:9.1, 21.1). The mean number of days from last care encounter to death was 78 (IQR:30.5-188) with a median CD4 closest to death of 346 cells/ µl (IQR:159, 562). In separate multivariate analyses, an increased risk of death was observed among those in care at clinics with no ART monitoring (aHR 1.82; 95%CI:1.36, 2.45), no retention monitoring (aHR1.54; 95%CI: 1.15, 2.06), and those with lower clinic assessment scores (≤6) (aHR1.43; 95%CI: 1.08, 1.90). HIV care continuum outcomes among PWH who died vs. survived found that higher proportions of those who died were retained in care in the 6-months prior to death (p=0.0359), yet lower proportions were prescribed ART (p=0.0223) and virally suppressed (VL<200 copies/ml)(p<0.0001)(Figure). Conclusion: Our findings suggest that despite PWH being relatively well- engaged, comprehensive site-level services may improve quality of care, thereby mitigating poor outcomes and improving survival.

Poster Abstracts

893 MORTALITY IS HIGHER AMONG SUCCESSFULLY TREATED PWH COMPARED TO MATCHED CONTROLS Veronica Svedhem-Johansson 1 , Zaake de Coninck 1 , Laith Hussain-Alkhateeb 2 , Anna Mia Ekström 1 , Göran Bratt 3 , Magnus Gisslén 4 , Max Petzold 4 1 Karolinska Institute, Stockholm, Sweden, 2 Sahlgrenska University Hospital, Gothenburg, Sweden, 3 Södersjukhuset, Stockholm, Sweden, 4 Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden Background: There is an ongoing debate among researchers about whether the lifespan of successfully treated people living with HIV (PLHIV) is comparable to that of the general population. In this 15 years follow up cohort study we present a survival analysis of People With HIV (PWH) and HIV-negative persons in relation to socio-demographic, virus load, CD-4 count, ART and mortality data from the national Swedish HIV cohort, InfCareHIV. Methods: A total of 4,066 people living with HIV were matched against 8,072 HIV-negative controls according to age, gender, and region of birth. Furthermore the association between viral load and CD-4 level at diagnosis, treatment outcome and mortality was assessed over a 15-year period by Cox regression estimates to compare the overall crude and adjusted Hazard Ratios for mortality. Results: Mortality rates/100 PY, HIV-Positive n=275/4,066 1.13 (1.00-1.27) as compaired to HIV-Negative n=110/8,072, 0.22 (0.18-0.26)p <0.001. After a 15 year follow up period, successfully treated PLHIV were found to be 3 times more likely to die when compared to HIV-negative controls (HR 3.01, 95%, CI 2.05-4.44, p<0.001). The risk of mortality decreased from HR 6.02 after the first year of successful treatment. Only 11 of 58 patients in our cohort died from an AIDS-related condition. When AIDS-attributed mortality is excluded, successfully treated PWH are still 2.4 times more likely to die compared to HIV negative persons (HR 2.43, 95%, CI 1.61-3.65, p<0.001). Among the 275 that died, the initial VL was ≥30,000 c/ml among 171 of the patients (62.2%) and <30,000 c/ml among 83 of the patients (30.2%). Patients with VL≥30,000c/mL at HIV diagnosis were associated with a 1.74 (95% CI 1.34-2.26) greater hazard of death comparedto patients with VL< 30,000 at diagnosis (p<0.001).

892 INDIVIDUAL AND SITE-LEVEL FACTORS ASSOCIATED WITH RISK OF DEATH AMONG PEOPLE WITH HIV Amanda D. Castel , Lindsey Powers Happ, Anne K. Monroe, Maria Jaurretche, Arpi Terzian, Alan Greenberg The George Washington University, Washington, DC, USA Background: Although antiretroviral therapy allows persons with HIV (PWH) to live longer, healthier lives, many PWH may not be receiving comprehensive HIV care, resulting in shortened survival. To identify factors potentially associated with improved survival, we sought to compare individual, clinic-level, and care continuum patterns among those who died and those who survived in a cohort of PWH receiving care in Washington, DC. Methods: Participants of the DC Cohort, a longitudinal observational cohort of PWH in care, at 14 sites as of 12/31/2016 were examined. Clinic-level variables were assessed with a site survey which queried available services (e.g., hours, referrals, visit intervals, re-engagement services, subspecialty care). An overall clinic assessment score (range 0-9) was assigned to each site. Other clinic-level variables examined were systematic retention in care monitoring and routine review of medication pick up (ART monitoring). Care continuum outcomes were assessed for the 6-months prior to death. Univariate analyses were used to compare participants who died vs. survived. Cox proportional hazards models

CROI 2018 339