CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

Facility-level factors influencing LTFU in the pre-ART and ART periods

*p<0.05; **p<0.001; Adjusted for male gender, age at enrollment (pre-ART period), age at ART initiation (ART period), WHO stage, year of enrollment and education level. HR=Hazard Ratio, CI=Confidence Interval Conclusions: Higher LTFU rates were identified in the pre- versus post-ART period. Facility-level factors associated with LTFU both before and after ART initiation included the level of care of the facility and availability and timeliness of labs. Our findings have implications for the development of facility-based strategies to improve retention in pre- and ART care. This can help to improve the proportion of patients initiating ART who achieve viral suppression. 1073 Patient Retention in HIV Care Is Related to Point of Diagnosis inWestern Kenya Becky L. Genberg 1 ; Hana Lee 1 ; Fatma Some 2 ; Joseph Hogan 1 ; Paula Braitstein 3 1 Brown University, Providence, RI, US; 2 Moi University School of Medicine, Eldoret, Kenya; 3 Indiana University, School of Medicine, Indianapolis, IN, US Background: Home-based counseling and testing (HBCT) successfully diagnoses HIV earlier than provider-initiated (PITC) and voluntary counseling and testing (VCT). However, it is unknown whether patients entering care from HBCT have better retention compared with patients entering care from other testing modalities. The objective of this study was to determine the impact of point of diagnosis (i.e., HIV testing program) on retention among patients in western Kenya. Methods: AMPATH (Academic Model Providing Access to Healthcare), a partnership between Moi University, Moi Teaching and Referral Hospital and a consortium of North American institutions, has provided HIV care to over 160,000 individuals in western Kenya since 2001. This retrospective analysis included all individuals >13 years enrolled in care between January 2008-September 2013, with data available on point of diagnosis. Lost-to-follow-up (LTFU) was defined as no clinical contact for at least 90 days following a missed scheduled return, without any information on vital status. Survival analysis methods using Cox regression were employed to estimate the impact of point of diagnosis on LTFU and mortality while adjusting for likely confounders. LTFU analysis was limited to those in care > 90 days. Censoring included death and administrative censoring. Results: The full sample included 19,425 individuals, of whom 64%were female, with median age of 38 years and median CD4 count at baseline of 225 cells/ml 3 . The incidence of LTFU was 7.3, 9.3, and 11.2 per 100 person-years for HBCT, VCT, and PITC, respectively, over a median of 2 years in care. Cox regression modeling adjusting for age (<30, 30-45, >45), sex, marital status, disclosure of HIV, electricity/water in the home, number of people in the household, BMI (<25, 25-30, >30), CD4 count and WHO stage at baseline revealed that patients entering care from PITC (adjusted hazard ratio (AHR)=1.24, 95% CI: 1.03,1.50) were more likely to be LTFU compared to HBCT. Conclusions: Patients who enrolled in care following diagnosis in HBCT and VCT have similar rates of retention, however those entering from PITC were more likely to become 1 EmoryUniversitySchoolofMedicine,Atlanta,GA,US; 2 EmoryUniversityRollinsSchoolofPublicHealth,Atlanta,GA,US; 3 EmoryUniversityRollinsSchoolofPublicHealth,Atlanta,GA,US; 4 McCordHospital,Durban,SouthAfrica Background: In resource-limited settings, provision of HIV care and antiretroviral therapy (ART) is shifting from specialized programs to the local sector. Decentralized care is critical for long-term programmatic sustainability. This extra step can be disruptive for patients who have established care with providers at the referring clinic especially if experiencing treatment complications. Real-world experience of this process is lacking in this vulnerable population. We examined the rate of re-linkage for patients receiving ART after a clinic closure, focusing on those with virologic failure (VF). We hypothesized that fewer patients overall would establish care at the down-referral clinic (re-linkage) than previously reported. Those with history of VF and poor adherence would be at greater risk of failure of down-referral. Methods: We conducted secondary analyses on a study of predictors of virologic failure in South Africa. We examined individual-level factors’ (survey and chart review) relation to re-linkage. In the parent study, after 5 months of first-line ART cases had VL ≥ 1000 copies/mL and controls VL ≤ 1000 copies/mL. The primary outcome was re-linkage, defined as self-reported attendance to the new clinic. We used two-sample tests to assess the statistical independence of select patient-level variables and failure to re-link. Variables that achieved statistical significance in the univariate analysis along with important epidemiologic variables were included in a multivariate logistic regression model. Results: The study cohort consists of 458 patients, 158 cases and 300 controls. Table 1 shows study population characteristics. 436 (95%) patients re-linked to care. In the univariate analysis patient satisfaction with original clinic, case-control status, duration of ART, and adherence were significant. In the multivariate analyses, not being pleased with clinic (OR 3.24, 95% CI 1.19–8.82), shorter ART duration (OR 1.04, 95% CI 1.00–1.07), and poor adherence (OR 3.85, 95% CI 1.20–12.36) remained significant in their association with failing to re-link to care at the down-referral site. LTFU. Additional efforts to track and retain patients entering care from PITC may be warranted. 1074 Successful Down-Referral Even Among Patients With Virologic Failure in South Africa Jonathan Colasanti 1 ; Darius McDaniel 2 ; Brent Johnson 2 ; Henry Sunpath 4 ; Carlos del Rio 1 ;Vincent C. Marconi 1

Poster Abstracts

627

CROI 2015