CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

1016 Barriers to Care and 1-Year Mortality in Newly Diagnosed HIV+ Persons in South Africa Ingrid V. Bassett 1 ; Sharon M. Coleman 2 ; Janet Giddy 3 ; Laura M. Bogart 4 ; Christine E. Chaisson 2 ; Douglas Ross 5 ;Tessa Govender 3 ; Rochelle P.Walensky 1 ; Kenneth A. Freedberg 1 ; Elena Losina 6 1 Massachusetts General Hosp, Boston, MA, USA; 2 Boston Univ Sch of PH, Boston, MA, USA; 3 McCord Hosp, Durban, South Africa; 4 Boston Children’s Hosp, Harvard Med Sch, Boston, MA, USA; 5 Midlands Med Cntr, Pietermaritzburg, South Africa; 6 Brigham and Women’s Hosp, Harvard Med Sch, Boston, MA, USA Background: Despite increasing availability of antiretroviral therapy (ART), only a fraction of those newly diagnosed with HIV enter care promptly in South Africa, leading to premature mortality among those not linked to care. Our objective was to evaluate the impact of self-perceived barriers to receipt of health care at the time of HIV diagnosis on 1-year mortality among newly diagnosed HIV-infected individuals in South Africa. Methods: We surveyed adults (≥18y) prior to HIV testing at four sites (two hospital outpatient departments and two primary health clinics) in Durban from August 2010 to January 2013. HIV-infected participants were offered CD4 testing and underwent TB screening. We used Cox proportional hazards models to determine the association between the number of perceived barriers to care and time to death within one year of HIV diagnosis. Perceived barriers included: 1) service delivery (wait too long to see a provider, not treated with respect by clinic staff); 2) financial (could not afford medication or transport); 3) personal health perception (not sick enough or too sick); 4) logistical (could not get off of work, care responsibilities for others); and 5) structural (could not get to clinic due to hours or transport, did not know where to find care). We assessed deaths via phone calls to family members and confirmed these through the South African death registry. Results: Among 4,903 participants enrolled, 1,899 (39%) were HIV-infected and 521 (28%) were co-infected with TB. Mean age was 35 years (SD 10), 49%were female, and median CD4 count was 192/µl (IQR: 72-346/µl). 1,057 participants (56%) reported no barriers, 370 (20%) reported 1-3 barriers, and 460 (24%) reported >3 perceived barriers to care. By one year after enrollment, 250 (13%) of participants had died. Adjusting for age, sex, distance to clinic, TB status and baseline CD4 count, participants who identified 1-3 barriers (adjusted hazard ratio [aHR] 1.49, 95% CI 1.06, 2.08) and >3 barriers (aHR 1.81, 95% CI 1.35, 2.43) had higher risk of 1-year mortality compared to those without self- identified barriers. Conclusions: HIV-infected individuals in South Africa who reported multiple perceived barriers to medical care at the time of diagnosis were nearly twice as likely to die within one year. Targeted structural interventions such as extended clinic hours, travel vouchers, and more streamlined clinic operations may improve linkage to care and ART initiation for these patients. 1017 Retention of Clinically Stable ART Patients in a Rapid Model of Care in Haiti Colette Guiteau Moise 1 ; Clovy Bellot 1 ; Kelly A. Hennessey 2 ;Vanessa R. Rivera 3 ; Patrice Severe 1 ; Darley Aubin 1 ; Fabienne Homeus 1 ; Alix Saint-Vil 1 ; Serena P. Koenig 4 ; JeanW. Pape 1 1 GHESKIO, Port-au-Prince, Haiti; 2 Analysis Group, Boston, MA, USA; 3 Weill Cornell Med Coll, New York, NY, USA; 4 Brigham and Women’s Hosp, Harvard Med Sch, Boston, MA, USA Background: Long clinic waiting times are a major contributor to attrition among antiretroviral (ART) patients. We evaluated outcomes for clinically stable ART patients enrolled in a new rapid model of care, the “Rapid Pathway” (RP), in one ART clinic at the GHESKIO Center in Port-au-Prince, Haiti. Methods: Clinically stable patients who had received at least 6 months of ART were eligible for RP care. Once enrolled in RP, patients were scheduled for clinic appointments every two months. At each visit, patients were contacted one day in advance by a community health worker. Patients who were asymptomatic were eligible for RP care, and those with symptoms or poor adherence were referred to a physician for evaluation. Patients in RP received care by a nurse, who evaluated health status, dispensed ART and other medications, and completed a visit form in the electronic medical record. Patients spent a median of 29 minutes from arrival to discharge from clinic, including the dispensing of ART. Results: From June 1, 2014 to August 31, 2014, 1,799 eligible patients initiated RP care. Of these, 950 (53%) were women and 632 (35%) lived on <$US 125 per year. In the 6 months prior to RP initiation, patients were a mean of 12 (SD: 23) and a median of 2 (IQR: 0, 12) days late to scheduled clinic visits and ART re-fills. We evaluated 12-month retention in care, defined as having at least one visit during the period from 9 to 12 months after the date of RP eligibility; timeliness of visits; and ART adherence, defined as medication possession ratio (number of pills dispensed/number of pills that would have been dispensed with perfect adherence over the follow-up period). 1,663 patients (92%) were retained in care for 12 months; retention was 96% among patients with timely adherence in the 6 months prior to enrollment in RP. The mean adherence over the study period was 89% (SD: 17) and 42% of patients missed no more than 5 days of ART during the study period. The median CD4 T-cell count increased from 447 at baseline to 514 at 1 year. With multivariable analysis, significant predictors of 12-month retention were greater time on ART prior to RP enrollment (OR 1.13; 95% CI: 1.06,1.21) timeliness of visits in the 6 months prior to RP enrollment (OR 1.01; Conclusions: Clinically stable ART patients had outstanding retention with expedited, nurse-led care in Haiti. This programmay also serve as a model for other resource-poor settings. 1018 Evaluating Appointment Patterns to Improve Sustainability of HIV Treatment in Zambia Monika Roy 1 ; Charles Holmes 2 ; Izukanji Sikazwe 2 ;Thea Savory 2 ; Mwanza Mwanza 2 ; Carolyn Moore 2 ; Kafula Mulenga 2 ; Nancy L. Czaicki 3 ; Nancy Padian 3 ; Elvin H. Geng 1 1 Univ of California San Francisco, San Francisco, CA, USA; 2 Cntr for Infectious Disease Rsr in Zambia, Lusaka, Zambia; 3 Univ of California Berkeley, Berkeley, CA, USA Background: Prevailing facility-based models for HIV treatment in high-prevalence settings require substantial contact with clinically stable individuals, thus incurring high costs for patients (e.g., transport, lost wages) as well as systems (e.g., staffing, provider burn out). De-intensification of services through emerging community-based models for clinically stable patients is needed to sustain and expand treatment capacity. We examined patient appointment histories in a network of clinics in Zambia to characterize visit patterns and quantify the potential reductions in visit burden obtainable through community-based models. Methods: We evaluated a cohort of HIV-infected adults on ART who made at least one clinic visit between March 1, 2013 and February 28, 2015 at 61 clinics in Zambia. Appointment dates and clinical data were obtained from the electronic medical record system used in routine clinical care. We evaluated time between scheduled appointments and stratified this interval by time on ART and WHO stage and CD4 count at enrollment. We quantified the proportion of visits among patients on ART for greater than six months and with a most recent CD4 count > 500/µl or > 350/µl. Results: Overall, 150,213 patients, observed over a maximum of nine years, made 3,450,172 visits after ART initiation. The median appointment interval was 31 days in the first 6 months on ART and 61 days subsequently, even after five or more years on treatment. Median appointment interval did not differ by enrollment WHO stage nor CD4 count. In this cohort, 868,841 of all visits (25%) occurred among patients on ART for greater than six months and who had a most recent CD4 count >500 cells/µl; while 1,507,835 visits (44%) were made by patients on ART for more than six months and who had a most recent CD4 count > 350 cells/µl. Visits made by patients with CD4 count > 500 cells/µl increased over time from 5% (2971/58,059) in 2005 to 32% (205,389/639,799) in 2014. Conclusions: Current facility-based care applies a one-size-fits-all appointment interval not tailored to clinical changes or time on treatment. Between one-quarter to nearly one-half of all visits occurred in patients who are likely clinically stable. Evolution from facility-based care towards a differentiated system that includes community-based models is urgently needed to enhance efficiency and sustainability.

Poster Abstracts

434

CROI 2016