CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

903 LIFE EXPECTANCY IN KEY POPULATIONS OF ADULTS WITH HIV IN THE US AND CANADA Keri N. Althoff 1 , Richard D. Moore 1 , Michael John Gill 2 , Michael A. Horberg 3 , Amy C. Justice 4 , Mari Kitahata 5 , Marina Klein 6 , Fidel A. Desir 1 , W. C. Mathews 7 , Angel Mayor 8 , Peter F. Rebeiro 9 , Michael J. Silverberg 10 , Stephen J. Gange 1 , Robert S. Hogg 11 1 Johns Hopkins University, Baltimore, MD, USA, 2 University of Calgary, Calgary, AB, Canada, 3 Kaiser Permanente Mid-Atlantic States, Rockville, MD, USA, 4 VA Connecticut Healthcare System, West Haven, CT, USA, 5 University of Washington, Seattle, WA, USA, 6 McGill University, Montreal, QC, Canada, 7 University of California San Diego, San Diego, CA, USA, 8 Universidad Central del Caribe, Bayamon, Puerto Rico, 9 Vanderbilt University, Nashville, TN, USA, 10 Kaiser Permanente Northern California, Oakland, CA, USA, 11 Simon Fraser University, Vancouver, BC, Canada Background: ART has improved the life expectancy (LE) of those with HIV. However, there is less information on LE changes in certain key populations living with HIV in the US, including Black men who have sex with men (MSM), Black women, Hispanics, people with a history of injecting drugs (PWID), and individuals co-infected with hepatitis C (HCV) in the more recent treatment era. Methods: Using data from NA-ACCORD, we estimated LE after ART initiation in three time-periods (2005-07, 2008-11, 2012-15). Cohorts in the NA-ACCORD have previously demonstrated good ascertainment of deaths using active and passive methods. Standard abridged life table methodology was used to estimate LE for Black MSM (vs. white MSM), Black women (vs. white women), Hispanic (vs. non-Hispanic) adults, PWID (vs. those who do not) and HCV+ (vs. HCV-) adults with HIV. Results: Among 55,858 ART initiators who contributed 248,931 person-years and 3,123 deaths between 2005 to 2015, the LE increased in both men and women (Figure). LE increased in Black MSM over time, but was consistently lower than non-Black MSM in all time periods. Black women had a similar LE compared to white women in 2005-07, but had a greater LE in 2012-15. Hispanic adults had a greater LE compared to non-Hispanic adults in all time periods. Although PWID had little increase in LE, non-PWID saw an increase in LE from 2008-11 to 2012-15. Similarly, adults with HCV saw little increase in LE and did not exceed 50 years in any time period; however those without HCV had consistent increases in LE across time. Conclusion: Although LE has improved since 2005 for most of these underserved populations, disparities persist. Factors that may be influencing the higher LE in Black compared to white women may be related to forces that are decreasing survival in whites in the US. The increased LE in Hispanics vs. non-Hispanics is reflective of what has been shown in the general US population. Adults with HIV and HCV saw very little (<2 years) increase in LE; further analysis reducing the influence of PWID are underway. Our LE results may be overestimated if NA-ACCORD cohorts under-represent those who are less likely to remain in care after ART initiation; estimates from 2012-15 are projections that are influenced by data truncated prior to 31 Dec 2015. The impact of the “treat all” era and HCV direct acting agents may further increase LE, but may not narrow disparities in key populations if maintaining HIV RNA suppression is playing a significant role in these disparities.

902 POST-HOSPITAL MORTALITY AND READMISSION AMONG HIV-INFECTED ADULTS IN SOUTH AFRICA Chris Hoffmann 1 , Minja Milovanovic 2 , Deanne Kerrigan 1 , Neil A. Martinson 2 , Ebrahim Variava 2 1 Johns Hopkins University, Baltimore, MD, USA, 2 University of the Witwatersrand, Johannesburg, South Africa Background: Despite the scale-up of the antiretroviral (ART) program in South Africa, HIV continues to cause substantial mortality. Hospitalization presents an opportunity to stabilize and engage individuals in ongoing care. Among a cohort of HIV-infected adults we describe health care use and survival following an index hospital admission. Methods: From April to June 2016, 121 hospitalized HIV-infected adults were enrolled, by convenience sampling, in a prospective cohort and followed to six months post-discharge. We sought to describe characteristics associated admission and mortality using multivariate logistic regression. Results: The median age of cohort members was 40 years (interquartile range (IQR) 31, 50), 56%were women, the median CD4 count was 260 (IQR: 113, 464), 11 had a discharge diagnosis of TB. The median duration of hospitalization among survivors was 5.5 days (IQR: 3, 9). 96 had a prior HIV diagnosis of which 73 (74%) reported being on ART, and 70% reported having at least one HIV clinic visit in the prior 6 months. A routine follow-up was scheduled for 98z%, this was within 2 weeks of discharge for 51%; 92% of visits were scheduled at the hospital specialty clinics. Among a sample of 15 participants who did attend clinic, the median time of a clinic visit was 4.3 hours and the median cost was 2% of the monthly household income. After discharge, 83 (68%) participants reported a clinic visit, 45% to a local primary care clinic, the median time after discharge to follow-up clinic was 5 weeks (IQR 3, 8). By 6 months following the index discharge, 43 (36%) participants had been readmitted and 31 (26%) died. Older age (OR 3.1, p=0.05, >51 versus 17-35 years), failure to attend clinic (OR 4.7, p=0.003), reporting “skipping going to clinic because it is hard to get to” (OR 2.9, p=0.03), and longer length of stay were all associated with death or readmission. Conclusion: HIV-infected adults admitted to hospital remain at high risk of death and readmission after discharge despite access to ART. While most participants had a post-discharge clinic visit, this occurred later and at a different location than scheduled. Notably, self-reports of difficulty getting to the clinic and not attending a clinic visit were both associated with hospital re-admission and mortality. Our results suggest that improving post-discharge outcomes may require strategies that improve access, acceptability, and use of timely follow-up care after discharge.

Poster Abstracts

CROI 2018 344