CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

1 University of California San Francisco, San Francisco, CA, USA, 2 Centre for Infectious Disease Research in Zambia, Lusaka, Zambia, 3 Africa Health Research Institute, Mtubatuba, South Africa, 4 Center for Infectious Disease Research in Zambia, Lusaka, Zambia, 5 Johns Hopkins University, Baltimore, MD, USA, 6 University of California Berkeley, Berkeley, CA, USA Background: Although differentiated service delivery (DSD) models offer a range of health systems innovations, their comparative desirability to patient populations, implementability and effectiveness remains unknown. We conducted a discrete choice experiment (DCE) to quantify model features most desired by patients to informmodel prioritization during scale-up in Zambia. Methods: We surveyed a random sample of HIV positive adults on ART at 12 clinics in Zambia and asked patients to choose between two hypothetical facilities which differed across six attributes: location of ART pick-up (clinic vs. community), frequency of ART pick-up (1 vs. 3 monthly), time spent waiting to pick up ART (1, 3 or 6 hrs), time spent waiting for a doctor (1, 2 or 5 hrs), type of adherence counselling (group vs. individual), and ability for a ‘buddy’ to collect ART. Each respondent answered one of two blocks of seven questions. We used mixed logit models to determine the degree of preference (i.e. preference weights - β) for each DSD feature, preference heterogeneity and willingness- to-trade. Results: Of 486 respondents, 59%were female and 85% resided in urban locations. Patients strongly preferred infrequent clinic visits (3 vs. 1 month visits: β=2.84; p <0.001) (Figure). Milder preferences were observed for reduced waiting time for ART (1 vs. 6 hrs.: β=-0.67; p<0.001) and reduced waiting time to see a doctor (1 vs. 3 hrs., β=-0.41; p=0.002), and facilities accommodating ‘buddy’ ART collection (β=0.84; p <0.001). In order to obtain 3 instead of 1 monthly refills, patients were willing to wait 6 hrs. for ART (vs. 1), wait 3 hrs. for a doctor (vs. 1), pick-up ART in the community instead of clinic, attend large group counselling, and forego a buddy system (β difference: 0.23; p=0.487). When stratified by residence, urban patients had a strong preference for collecting ART in at the health facility (β=1.32, p<0.001) whereas rural patients preferred drug pick-up in the community (β=-0.74, p=0.049). Conclusion: Patients in Zambia primarily want to attend health facilities infrequently, and this preference outweighs the desire for all other DSD features. Substantial preference heterogeneity was demonstrated by urban and rural participants, suggesting that Zambia should prioritize DSD models that remain facility-based but require infrequent contact, particularly in urban settings, with consideration of community based drug distribution for those more rural.

1053 REACHING TOWARD 90-90-90 AMONGST CORRECTIONAL FACILITY INMATES IN ZAMBIA Michael Herce 1 , Christopher Hoffmann 2 , Steph Topp 1 , Harry Hausler 3 , Helene Smith 1 , Lucy Chimoyi 4 , Candice Chetty-Makan 4 , Rachek Mukora 4 , Abraham Olivier 3 , Monde Muyoyeta 1 , Stewart Reid 1 , Salome Charalambous 4 , Katherine Fielding 5 1 Center for Infectious Disease Research in Zambia, Lusaka, Zambia, 2 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 3 TB/HIV Care Association, Cape Town, South Africa, 4 The Aurum Institute, Johannesburg, South Africa, 5 London School of Hygiene & Tropical Medicine, London, UK Background: Achieving the 90-90-90 goals among key populations is believed to be critical for HIV control. We sought to implement a universal test and treat (UTT) program in correctional facilities in Zambia and South Africa and measure success with the 90-90-90 goals in mind. Here we describe outcomes from Lusaka Central Correctional Facility in Zambia. Methods: We offered immediate ART to all inmates ≥18 years, with HIV regardless of CD4 or WHO stage who were expected to be incarcerated ≥30 days after ART initiation. We strengthened health services with personnel and training to make UTT feasible. We determined the corrections census on two days-a baseline day prior to UTT and an endline day 12 months after UTT initiation and 18 months after the baseline. We used the single day census to create virtual cross sections for HIV testing, ART initiation, and viral load suppression. The denominator for status was the prison census, the numerator included those in the census with HIV testing in the prior 12 months or known to be HIV-positive. The proportion on ART was assessed with a denominator of those known HIV-positive from the virtual cross-section and those HIV-positive and on ART as the numerator. Viral load suppression included the denominator of those known to be on ART and the numerator of those with a viral load <1000 c/mL. Results: On the baseline cross-section day there were 1,467 inmates in the facility. Of these, 857 (58.4%) knew their HIV status and 277 were HIV-positive (18.9%). Of those with known HIV, 188 (67.9%) were on ART. Viral loads were not routinely obtained prior to UTT. On the endline day, 1,370 inmates were in the facility and 1,263 (92.2%) had been tested or were already known positive. Of those, 647 (47.2%) were HIV-positive of whom 438 (67.7%) were on ART (a 2.3 fold increase in inmates on ART). Of those on ART, 85 had a viral load result; 68 (91.8%) having a viral load <1000 c/mL. Expected release within 30 days of HIV testing was noted as an important reason for not initiating ART among some HIV-positives. Conclusion: High levels of HIV testing and virologic suppression are feasible within correctional facilities. Although many more inmates were placed on ART, the second 90 goal was not reached possibly due to many inmates leaving the facility within 30 days of HIV testing. Justice involved populations should be included in efforts to achieve 90-90-90 goals and specific correctional facility programs are feasible. 1054 DIFFERENTIATED CARE PREFERENCES OF STABLE PATIENTS ON ART IN ZAMBIA Ingrid Eshun-Wilson 1 , Mpande Mukumbwa-Mwenechanya 2 , Hae-Young Kim 3 , Arianna L. Zanolini 4 , Chanda Mwamba 4 , David Dowdy 5 , Laura K. Beres 5 , Monika Roy 1 , Anjali Sharma 2 , Steph Topp 2 , Nancy Padian 6 , Izukanji Sikazwe 2 , Charles B. Holmes 2 , Carolyn Bolton Moore 4 , Elvin Geng 1

Poster Abstracts

1055 DURABLE VIRAL SUPPRESSION AMONG PEOPLE WITH ACUTE AND NONACUTE HIV IN NORTH CAROLINA Nicole Dzialowy Adams 1 , Brad Wheeler 1 , Anna Cope 2 , Victoria L. Mobley 1 , John Barnhart 1 , Joann D. Kuruc 3 , Cynthia L. Gay 3 , Lindsey M. Filiatreau 3 , Rhonda Ashby 1 , Erika Samoff 1 1 North Carolina Division of Public Health, Raleigh, NC, USA, 2 CDC, Atlanta, GA, USA, 3 University of North Carolina at Chapel Hill, Chapel Hill, NC, USA Background: North Carolina (NC) has had statewide screening for acute HIV infection (AHI) since 2002. The program involves a coordinated effort between NC Disease Intervention Specialists (DIS) who locate and interview people diagnosed with AHI within 72 hours of case notification and HIV providers who expedite their care appointments. Non-AHI cases take DIS approximately two weeks to locate and interview. We assessed whether prioritizing the linkage

CROI 2019 414