CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

HIV care, along with predictors thereof, could inform the design of tailored interventions for improving HIV care engagement. We used visit data from the eight-site CFAR Network of Integrated Clinical Systems (CNICS) to examine patterns and predictors of HIV care attendance over a ten-year period. Methods: We conducted a retrospective cohort study of all adults newly entering CNICS between January 1, 2005 and December 31, 2015 (N=18,160), following them longitudinally until death, ten years, or March 22, 2018. Our outcome was HIV primary care visit attendance (yes/no) in each six-month interval after CNICS entry. We used group-based trajectory modeling to: 1) identify a set of longitudinal HIV care patterns followed from the time of CNICS entry, and 2) examine associations between each pattern and race/ethnicity, age at entry, and transmission risk group. We tested models with 2-7 trajectory groups and selected the final model based on the Bayesian Information Criterion. Results: We identified five distinct HIV care trajectories (Figure): ~32% of patients had consistently high care attendance over time (>75% probability of attendance in each interval); ~23% exhibited a rapid decline within two years to a sustained, low probability (<5%) of attendance; ~16% showed a very slow decline in attendance; ~17% had an intermediate rate of decline; and ~12% showed a slowly fluctuating pattern that started with a decrease but shifted to an increase starting ~three years after entry. Older age at entry was protective against all sub-optimal trajectories (with the “consistently high” pattern as referent): odds ratios per five-year age increase ranged from 0.79 (95% confidence interval: 0.77-0.81) for the “slow fluctuation” group to 0.86 (0.84- 0.88) for the “intermediate decline” group. Race/ethnicity and transmission risk group had mixed associations with care patterns. Conclusion: Most new CNICS entrants exhibited sub-optimal HIV care trajectories, but there was wide variation in the longitudinal pathways followed. By identifying heterogeneous care engagement patterns and predictors thereof, this analytical approach allows improved understanding of HIV care engagement over time for designing tailored interventions and refined models of the HIV care continuum.

Institute. The participants were randomized 1:1 to receive either CfLU or standard of care (SoC-clinic visits only). In English or 2 local languages, the CfLU arm received daily/weekly pill reminder calls or SMS messages, visit reminders, health information advice and symptom reporting. At 6, 12 and 24 months of follow up QOL assessments (HIV Medical Outcomes Survey, MOS-HIV including physical health score [PHS] and mental health score [MHS]) were done using Likert-type scale with difference in differences analysis and analysis of covariance (ANCOVA). Qualitative and tool use data also collected. Data from 6m are presented here. Results: Between August 2016 and February 2018 across 2 sites, 1031 PLHIV accessing care were screened and 600 enrolled on the study (n=300/site). Sixty-nine percent were female and median age was 32 (IQR25-40). Eight four participants were ART naïve, remaining ART experienced. At baseline, 97% chose IVR over SMS. There was no difference in arms for education level, marital & employment status, previous TB or alcohol use. 277 in each arm attended at 6m. There is no statistical observed difference in mean percentage score of MOS-HIV, MHS and PHS at baseline and 6m between CfLU and SoC arms. In those starting first line ART or switching to second line, there was a significant improvement in PHS (ANCOVA 4.01, p=0.048). There was no significant difference between CfLU versus the SoC in the proportion of patients with viral load <50 copies at 6m (21% vs 18%: p-value=0.372). Conclusion: This is the first RCT for PLHIV on ART incorporating options for IVR and SMS options; strong preference was shown for IVR over SMS. In this mixed group of patients, there was no statistical effect of CfLU observed on QOL at 6m. Within this study, a higher than expected baseline QOL and virological suppression was encountered for both sites which may have affected results.

Poster Abstracts

1038 HIV CARE CONTINUUM CHANGES AMONG PWID AND MSM IN INDIA: A TALE OF 2 KEY POPULATIONS Gregory M. Lucas 1 , Sunil S. Solomon 1 , Allison M. McFall 1 , Aylur K. Srikrishnan 2 , Canjeevaram K. Vasudevan 2 , Santhanam Anand 2 , Vinita Verma 3 , Kuldeep Sachdeva 3 , David D. Celentano 1 , Suniti Solomon 2 , Shruti H. Mehta 1 1 Johns Hopkins University, Baltimore, MD, USA, 2 YR Gaitonde Center for AIDS Research and Education, Chennai, India, 3 National AIDS Control Organisation, New Delhi, India Background: Key populations account for the substantial majority of people living with HIV outside of sub-Saharan Africa. Advancing the HIV care continuum among key populations is necessary, particularly in low- and middle-income countries, to reach the UNAIDS HIV treatment goals. We present longitudinal changes in the HIV care continuum among PWID and MSM across multiple cities in India. Methods: This is a secondary analysis of data collected in a cluster-randomized trial of an integrated care intervention to improve HIV outcomes among key populations in India. The study included 12 PWID sites and 10 MSM sites. We conducted baseline (2013) and follow-up (2017) respondent-driven sampling surveys of ~1000 participants in each of the 22 sites. We tested participants for HIV and measured HIV RNA and CD4 cell counts in HIV-positive participants. We used sampling-weighted estimates and linear regression to compare baseline and site-level changes in HIV continuum outcomes in PWID and MSM sites, controlling for outcome prevalence at baseline and study arm assignment. Results: In the baseline survey, we recruited 11,993 PWID (2,544 HIV-positive) and 9,997 MSM (1,086 HIV-positive) participants. In the follow-up survey, approximately 4 years later, we recruited 11,721 PWID (2517 HIV-positive) and 10,005 MSM (1763 HIV-positive) participants. The intervention was not significantly associated with changes in care continuum outcomes. At baseline,

1037 CALL FOR LIFE UGANDA TM: AN RCT USING INTERACTIVE VOICE RESPONSE FOR PLHIV ON ART Rosalind M. Parkes-Ratanshi , Maria S. Nabaggala, Agnes N. Bwanika, Mohammed Lamorde, Rachel King, Noela Owarwo, Eva A. Laker Odongpiny, Richard Orama, Barbara Castelnuovo, Agnes Kiragga Infectious Disease Institute, Kampala, Uganda Background: The WHO recommends use of mobile phone health technologies (mHealth) to support adherence in HIV. Studies on text messages show promise but with limited rigorous evaluations. The Call for Life UgandaTM (CfLU) study is a randomized controlled trial (RCT) using an interactive voice response (IVR) calls system designed to support PLHIV on ART. The primary study objective was to determine the effect of CfLU on quality of life (QOL) of people living with HIV (PLHIV) in Uganda. Methods: MOTECH software-based Connect for LifeTM (Janssen, Johnson & Johnson) was adapted for Ugandan setting, with the Infectious Diseases

CROI 2019 407