Abstract
BACKGROUND: India's free antiretroviral therapy (ART) programme was initiated in 2004. People living with HIV who were registered with ART centres (ARTC) were initiated on ART based on the CD4 count cutoffs as per prevailing guidelines. The others with higher counts remained on six-monthly follow up. We estimated retention rates among people living with HIV receiving ART in the programme and their determinants during 2012-2017. METHODS: In this retrospective cohort study, the records of people living with HIV aged ≥15 years, registered between April 2012 and March 2017 (reference period) in 81 of 396 ARTC across 33 Indian states were reviewed. 'People living with HIV not on ART' were defined as all those who were registered but not eligible for ART initiation or not started ART through the reference period. 'People living with HIV on ART' were those who were already on ART or initiated on ART as per prevailing guidelines. Relevant data from the clinic records were extracted and analysed for 'Not on ART' and 'On ART' groups separately using life-table method, Cox proportional hazards model to estimate retention probability and potential determinants. FINDINGS: Of 154,154 registered people living with HIV, 82.3% received ART ('on ART') during 2012-2017. Proportion retained was lower among 'not on ART' vs 'on ART' people living with HIV and was statistically significant (71.1% vs 88.9%, p < 0.001). Five-year retention probability was 57% for 'not on ART' and 81% for 'on-ART' people living with HIV (p < 0.001). The incidence of cases who were lost to follow up was 12.9 and 4.3/100 person-years among 'not on ART' & 'on ART' people living with HIV, respectively. Determinants of becoming lost to follow up (Adjusted HR, 95% CI) included 'being in not on ART' (Adjusted HR: 2.95, 95% CI: 2.85-3.05) 'being male' (1.08, 1.05-1.11); 'having CD4 count 351-500 cells/mm(3)' at registration (1.21, 1.16-1.26); and 'having tuberculosis' (1.15, 1.10-1.19). INTERPRETATION: New programmatic strategies for improving retention of people living with HIV in care may benefit by focussing on males, younger ages (15-29 years), CD4 counts during registration, history of or new TB diagnoses and early intervention within the first year. FUNDING: This study was supported with NOA# SAMS/NACP/IE-ART/NARI/2017/09 dated March 27, 2017, awarded by Strategic Alliance Management Services Pvt. Ltd. (SAMS) with funding support from the Global Fund.