AI-generated research brief — verify at source
Person-Centered Care Intervention Boosts HIV Care Reengagement and Halves Repeat Treatment Interruptions
A large stepped-wedge cluster randomized trial found that a multi-component person-centered care intervention raised 12-month return rates to HIV care from 67.7% to 72.3% (about 16% higher likelihood of return; aHR 1.16, 95% CI 1.12–1.20) and cut the rate of repeat treatment interruptions among those who returned by approximately half (aHR 0.50, 95% CI 0.45–0.55). Overall, the proportion of clients remaining in care 12 months after a treatment interruption rose from 51.5% to 60.1%, representing a meaningful improvement in sustained retention at scale.
What Was Studied
This analysis investigated whether a behavioral intervention targeting healthcare worker attitudes and the quality of client-provider interactions could improve two specific retention outcomes: the likelihood of returning to HIV care after a treatment interruption, and the likelihood of avoiding further interruptions once care was resumed. The hypothesis was that friendlier, more supportive healthcare worker behavior lowers the psychological and practical barriers that discourage clients from re-engaging after a missed visit.
How It Was Studied
The analysis drew on data from a stepped-wedge cluster randomized trial conducted across 24 clinics in Zambia between August 2019 and November 2021. Clinics were allocated into four groups and crossed over sequentially from routine HIV care (control) to the person-centered care (PCC) intervention every six months. The PCC intervention consisted of three coordinated components: structured training and coaching for healthcare workers on person-centered practices, systematic collection and feedback of client experience data through exit interviews, and small facility-level incentive payments. The analytic sample comprised 128,910 individuals living with HIV who became more than 30 days late for a scheduled visit — 69,671 under control conditions and 59,239 under the intervention. Multistate methods and Cox proportional hazards and Poisson models were used to estimate outcomes while accounting for the stepped-wedge design.
What Was Observed
- Return to care after treatment interruption: At 12 months, 72.3% of clients in the intervention group had returned to care compared with 67.7% in the control group — about 16% higher likelihood of return (aHR 1.16, 95% CI 1.12–1.20; p < 0.001). The crude risk difference was +4.6 percentage points (CI 4.1–5.2).
- Repeat treatment interruptions among returners: Among clients who did return, 44.3% in the intervention group experienced a subsequent interruption versus 55.6% in the control group — approximately half the rate of re-interruption (aHR 0.50, 95% CI 0.45–0.55; p < 0.001), a risk difference of −11.4 percentage points.
- Sustained presence in care after return: The proportion of returners remaining in care at 12 months after their initial return rose from 73.4% to 82.7% — about 14% higher probability of sustained engagement (aRR 1.14, 95% CI 1.05–1.25; p = 0.002).
- Overall in-care prevalence at 12 months post-interruption: Combining all pathways, 60.1% of intervention-exposed clients were in care 12 months after a treatment interruption versus 51.5% under control — approximately 19% higher overall retention (aRR 1.19, 95% CI 1.05–1.35; p = 0.008), a risk difference of +8.7 percentage points.
Why This Matters
Treatment interruptions are a recognized cycle in HIV care: poor client-provider interactions contribute to disengagement, and fear of judgment upon return deepens avoidance. This trial provides evidence that addressing the interpersonal quality of care — specifically healthcare worker behavior and responsiveness — can meaningfully interrupt that cycle, improving both the probability of return and the durability of re-engagement. The consistency of effects across multiple sensitivity analyses strengthens confidence that the intervention, not secular trends alone, drove the observed improvements.
How to Read This Result
While effect sizes are large and directionally consistent across analyses, the findings derive from a post-hoc analysis of a stepped-wedge trial, which limits causal inference and introduces potential time-related confounding; the results are therefore best interpreted as strongly suggestive rather than definitively causal.
Limitations
The analysis was not pre-specified — it was conducted post-hoc — which raises the possibility of outcome selection bias and limits its standing as a primary confirmatory test. The stepped-wedge design, while well-suited to phased rollouts, introduces challenges including secular trends over the study period and potential contamination between control and intervention phases, which the authors acknowledge may have influenced observed effect estimates despite statistical adjustments.