CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

transported to other EMR systems allowing for the development of electronic care cascades and dashboards.

1030 ARE THEY REALLY LOST? MULTI-CENTER TRACING STUDY IN ART PROGRAMS IN SOUTHERN AFRICA

Benedikt Christ 1 , Kathrin Zürcher 1 , Frédérique Chammartin 1 , Josephine Muhairwe 2 , Laura Jefferys 3 , Janneke van Dijk 4 , Kombatende Sikombe 5 , Monique van Lettow 6 , Cleophas Chimbetete 7 , Sam J. Phiri 8 , Matthias Egger 1 , Marie Ballif 1 1 Institute of Social and Preventive Medicine, Bern, Switzerland, 2 SolidarMed, Maseru, Lesotho, 3 SolidarMed, Pemba, Mozambique, 4 SolidarMed, Masvingo, Zimbabwe, 5 Centre for Infectious Disease Research in Zambia, Lusaka, Zambia, 6 Dignitas International, Zomba, Malawi, 7 Newlands Clinic, Harare, Zimbabwe, 8 Lighthouse Trust Clinic, Lilongwe, Malawi Background: Low retention on antiretroviral therapy (ART) is a threat to the UNAIDS 90-90-90 targets. We studied outcomes of people living with HIV (PLHIV) on ART but lost to follow-up (LTFU) in Southern Africa. Methods: We traced patients defined as LTFU (>90 days after a missed visit) using a common protocol in 6 ART programs of the International epidemiology Databases to Evaluate AIDS (IeDEA): Malawi (2 sites), Zimbabwe (2 sites), Lesotho and Mozambique. We randomly sampled PLHIV lost at each site, stratifying for age, sex and time on ART. Tracing consisted of text messages (one attempt), phone calls (max. 3 attempts) and/or home visits (max. 3 attempts). We used descriptive statistics and univariate logistic regressions to assess predictors for mortality. Results: We included 1564 patients LTFU: 435 in Lesotho, 381 in Malawi, 408 in Mozambique and 340 in Zimbabwe. Median age at tracing was 35 years (interquartile range [IQR]: 26-46), 57%were female and 81% from rural clinics. Last median CD4 count was 392 cells/µl (IQR: 226-594, available for 741 [47%] PLHIV), median time on ART was 33 months (IQR: 21-47). Checking patients’ files clarified vital status in 272 (17%) cases, without need for tracing. No file was found in 183 (12%) cases. Among 1109 patients traced, 369 (33%) were found after a mean of 1.4 attempts (range 1-5); 11% of patients were traced by phone calls, 71% by home visits and 17% by both. Text messages were only used for <1%. The remaining 67%were either not found (250; 34%) or their status was obtained from other informants (490; 66%; Fig. 1). Overall, 922 (59%) PLHIV were alive, 207 (13%) had died and in 435 (28%) cases, vital status remained unclear. Among those alive, 225 (24%) had never missed a visit or returned to care at the same clinic, 368 (40%) had transferred to another clinic (218 silently), 233 (25%) stopped taking ART and there are no details available for 97 (11%). Predictors for mortality were age ≤15 (odds ratio [OR] 1.9, 95% CI 1.2-3.1) and >50 (OR 3.4, 95% CI 2.2-5.1) compared to 26-50 years, LTFU for >1 year compared to ≤1 year (OR 2.7, 95% CI 1.3-5.7), WHO stages 3&4 compared to stages 1&2 (OR 3.4, 95% CI 2.2-5.1), and last CD4 count <200 compared to ≥200 cells/µL (OR 2.1, 95% CI 1.2-3.8). Conclusion: Most PLHIV defined as LTFU were found alive and in care. Tracing remains necessary in most instances but needs improvement to locate all PLHIV lost. Better ways to inform health systems and novel approaches to follow up PLHIV are needed in the treat-all era.

1031 HEALTH DEPARTMENT RANDOMIZED TRIAL TO RE-ENGAGE OUT-OF-CARE HIV INFECTED PERSONS Robyn N. Fanfair 1 , George Khalil 1 , Nasima Camp 1 , Kathleen Brady 2 , Alfred DeMaria 3 , Merceditas Villanueva 4 , Liisa Randall 3 , Heidi Jenkins 5 , Crystal Lucas 2 , Frederick Altice 4 , Anthony Gerard 2 , Nina Kishore 3 , Tiffany Williams 1 , Taraz Samandari 1 , Paul J. Weidle 1 1 CDC, Atlanta, GA, USA, 2 Philadelphia Department of Public Health, Philadelphia, PA, USA, 3 Massachusetts Department of Public Health, Boston, MA, USA, 4 Yale University, New Haven, CT, USA, 5 Connecticut Department of Public Health, Hartford, CT, USA Background: Over a quarter of persons living with HIV in the United States do not receive care, and most transmissions of HIV come from persons known to be infected but not in care. We implemented a data-to-care model using health departments and local clinics to identify out-of-care (OOC) HIV-infected individuals with the objective of increasing the number of such persons re- engaged, retained in medical care, and achieving viral load suppression. Methods: Criteria for inclusion were age ≥18, and in care at a trial clinic during a 12-month eligibility period followed by no evidence of care in ≥ 6 months (i.e., no visit or labs). OOC was determined by HIV surveillance and clinic data from three jurisdictions: Connecticut (CT), Massachusetts (MA) and Philadelphia (PHI). All patients deemed OOC were randomized to receive standard engagement in care (SOC) services from the trial clinic or an active public health field services intervention. Re-engagement in care was defined as linking to a trial clinic within 90 days of randomization, as determined by HIV surveillance data. Each jurisdiction was analyzed separately as interventions and services varied by health department. Chi-square tests were performed and a p-value <0.05 was considered statistically significant. Results: Between 8/16/2016 and 7/31/2018, a total of 533 (CT), 591 (MA), and 609 (PHI) OOC HIV-infected persons were enrolled and had ≥ 90 days since date of randomization. Among all sites 64%-76%were born male, 38%-66% were non-Hispanic black, 55%-69%were aged ≥ 40 years, and 44%-62% were diagnosed with HIV ≥ 10 years. In CT, 222 (41.7%) re-engaged in care ≤ 90 days [118 (46.3%) in intervention vs 104 (37.4%) in SOC, P=0.038]; in MA, 285 (48.2%) re-engaged in care ≤ 90 days [153 (51.2%) in intervention vs 132 (44.9%) in SOC, P=0.108]; and in PHI 306 (50.2%) re-engaged in care ≤ 90 days [181 (58.6%) in intervention vs 125 (41.7%) in SOC, P<0.0001]. The median times to re-engagement in care for intervention vs SOC arms were: 37 and 48 days (p=0.011) in CT, 38 and 42 days (0.329) in MA and 29 and 45 days (p<0.001) in PHI, respectively. Conclusion: This randomized controlled trial showed that a collaborative data-to-care model and field services intervention increased the proportion of persons re-engaged in care in two jurisdictions and decreased the time to re-engagement in all three. Health department interventions can improve re- engagement in care among HIV-infected persons who are out of care.

Poster Abstracts

CROI 2019 404