CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Results: Of 2,517 adult patients initiated on ART during the study period, 1,622 (65%) were female, 1,203 (48%) from dispensaries/health centers (HC), 1,082 (44%) initiated D4T-based regimens. Median age at ART initiation was 35.1 years, (interquartile range [IQR]; 28.8–42.8), and median CD4 count at initiation was 174 cells m/L, (IQR; 78-258). Thirty-three percent of patients (839) were non-adherent at least once. Non-adherent patients were more likely to be from county referral (38.1%) or national hospitals (30.7%) versus dispensaries/HC (27%, p<.001); to have been initiated on TDF (32%) or D4T-based regimens (36%) compared to AZT-based regimen (28%, p=0.008). Factors associated with repeated non-adherence were being; males (Odds Ratio [OR] 1.4; 95% confidence interval [CI] 1.2-1.6, p|<0.004), from national facilities, OR 3.7: 95% CI 2.7-4.9, p<.001, from county referral facilities, OR 1.9; 95% CI 1.6-2.1, p<.001, and initiated on D4T-based regimen, OR 1.3; 95% CI 1.1-1.5, p<.001. Once categorized non-adherent, patients were more likely to have repeated instances of non-adherence, (OR 2.7, 95% CI 2.4-3.1, p<.001). Conclusion: While patient retention is important to ensure adherence to treatment, substantially high rates of repeat non-adherence to clinic appointments continue to be observed. National and county referral hospitals had high non-adherence rates, possibly suggesting high patient load affecting the patient-caregiver relationship and reduced quality patient management. Males may need more targeted intervention to improve adherence.

1 Reed College, Portland, OR, USA, 2 Population Services International, Washington, DC, USA, 3 PSI/Ethiopia, Addis Ababa, Ethiopia, 4 Office of Evaluation Sciences, Washington, DC, USA Background: Though female sex workers (FSW) in Ethiopia are disproportionally impacted by HIV, they face numerous barriers to accessing and remaining in HIV care services. PSI/Ethiopia provides HIV care services through a network of FSW-friendly drop-in-centers (DICs), and tested a “praise message” intervention to improve ART adherence and retention in care among FSW living with HIV. Methods: FSW newly diagnosed with HIV were randomized to standard of care (SoC) or a “praise message” arm (PM). The PM arm received “praise messages” (short, positive calls from the DIC nurse thanking them for investing in their health) 24 hours and 2 weeks after each completed ART appointment. Praise messages were provided for those in the PM arm through 6 months on ART. Outcomes of interest included ART adherence and retention in care at 1, 3, and 6 months after enrollment. Data were collected using the project’s routine monitoring and information system, and all study activities were carried out by existing DIC staff. Data were analyzed using OLS regression in STATA 15.0. Results: We recruited 866 participants. Of these, 436 (50.3%) were randomized to the PM arm, and 430 (49.7%) were randomized to SoC. Participants were recruited from 25 DICs, with a median age range of 25-29 years old. Age did not vary significantly between the study arms (mean age SoC: 29.43; PM: 29.59, p-value=0.793). Of the 735 respondents with completed follow-up by September 3, 2018, overall one-month retention in care was 76.1% and ART adherence was 74.5%. Preliminary data analysis found that 1 month retention did not differ significantly by study arm (SoC: 76.4%; PM: 77.4%, p>0.05). Similarly ART adherence at 1 month did not differ significantly by study arm (SoC: 74.8%; PM: 75.6%, p>0.05). Conclusion: The intervention did not improve retention in care or ART adherence among FSW living with HIV in our study, providing only a statistically insignificant 1.0 percentage point increase in retention and 0.8 percentage point increase in ART adherence at 1 month. While the intervention did not have an impact on the primary outcomes of interest, this study demonstrates the feasibility of conducting rigorous randomized evaluations of important health outcomes in the context of routine service delivery. With the continued scale-up of electronic, client-based record management systems, routine data should increasingly be leveraged to facilitate low-cost research under operational conditions.

Poster Abstracts

1042 REPEAT NONADHERENCE TO CLINIC APPOINTMENTS AMONG HIV- INFECTED ADULTS ON ART IN KENYA Jacques Muthusi , Tai Ho Chen, Kenneth Masamaro, Peter W. Young, Emily C. Zielinski-Gutierrez US CDC Nairobi, Nairobi, Kenya Background: Since the early 2000s, Kenya has scaled-up antiretroviral therapy (ART) for HIV-infected persons. Patient adherence to medication is key to avoid drug resistance, treatment failure, and death among HIV-infected patients. Occurrences of non-adherence to clinic appointments could be an objective proxy for non-adherence to treatment. We investigated factors influencing repeat non-adherence among HIV-infected adults newly initiated on ART. Methods: We conducted a retrospective, national survey of adult patients, aged 15 years and above, who initiated ART from October 2003-September 2013 in Kenya. Using clinic appointments data, patients were considered non- adherent if they missed a scheduled appointment by >90 days. We used Chi- square statistics to compare patient characteristics by non-adherence status. We used generalized estimating equations to investigate factors associated with repeated non-adherence. All analyses were weighted per study design.

1043 ATTRITION ALONG THE CARE CASCADE IN SOUTH AFRICAN EMERGENCY DEPARTMENTS Bhakti Hansoti 1 , Aditi Rao 1 , George Mwinnyaa 1 , John Black 2 , Reinaldo Fernandez 1 , Elizabeth Hahn 1 , Oliver Laeyendecker 1 , Roshen Maharaj 2 , Pamela Mda 2 , Jernelle Miller 1 , Andrew D. Redd 1 , Steven J. Reynolds 1 , David Stead 2 , Thomas C. Quinn 1 , for the WISE research group 1 Johns Hopkins University, Baltimore, MD, USA, 2 Walter Sisulu University, Mthatha, South Africa Background: The Eastern Cape region of South Africa is known for high prevalence of HIV worldwide. HIV service delivery in this region is challenged

CROI 2019 409