Data-informed Stepped Care (DiSC) to improve adolescent and young adult HIV care outcomes in Kenya: a cluster randomized trial

Introduction

Systematic use of data-driven tools to allocate care services based on needs, including differentiated care for stable individuals and intensive care for those with higher risk, may improve retention and viral suppression in adolescents and young adults living with HIV (AYLHIV).

Methods

This cluster randomised trial in western Kenya tested a data-informed stepped care intervention that assigned AYLHIV to four intensities of care according to need. AYLHIV at 12 intervention facilities underwent step assignment at each visit; those at lowest risk were offered differentiated models of service delivery (DSD), and those with risk factors more intensive services. AYLHIV at control sites received standard care. AYLHIV were followed for 12 months. Clinical and viral load data were abstracted from medical records. The primary outcome was the proportion of missed visits (defined as > 30 days late for scheduled visit). Secondary outcomes included loss to follow-up, viral non-suppression and assignment to DSD (multi-month refills or pharmacy fast-track visits). Mixed effects regression was clustered by individual and facility and adjusted for outcomes during the pre-enrolment period and baseline variables that differed by arm.

Results

Between April and July 2022, 1911 AYLHIV ages 10–24 were enrolled (control: 1016, intervention: 895, 1708.8 person-years). Median age was 17, and 1512 (79.5%) were in school. Characteristics were balanced by arm, except for a higher proportion coming to the clinic alone in control arm (68.5% vs. 61.1%, p = 0.04). At intervention facilities, using the DiSC tool, 574 (64.6%) AYLHIV were assigned to DSD, 122 (13.7%) to standard care, 100 (11.3%) to mental health and retention counselling, and 92 (10.4%) to intensive case management. Missed visits were 8.5% in intervention versus 8.3% in control (adjusted risk ratio [aRR]: 1.04, 95% CI: 0.89−1.20); viral non-suppression (7.7% vs. 9.7%, aRR 0.79 95% CI: 0.54−1.16) and antiretroviral therapy adherence (92.8% vs. 94.6%, aRR 0.98 95% CI: 0.94−1.02) were similar between arms. AYLHIV in the intervention arm received more fast-track visits (aRR 1.21, 95% CI: 1.01−1.46). Intervention facilities experienced fewer scheduled appointments compared to control (aRR: 0.95, 95% CI: 0.91−0.98, p = 0.004).

Conclusions

Overall, missed visits and non-suppression were infrequent (< 10%) and did not decrease with the DiSC intervention. The DiSC intervention resulted in increased assignment to differentiated services without increasing missed visits or viral non-suppression.