Integrating clinical decision support and mobile health for differentiated HIV service delivery in Lesotho (VITAL): a cluster-randomised non-inferiority trial

Background: The World Health Organization conditionally endorses digital interventions to strengthen health systems, while acknowledging limited evidence on their benefits and harms. We assessed the effectiveness of digital health-supported differentiated service delivery on HIV treatment outcomes in rural southern Africa. Methods: The VITAL pragmatic, open-label, parallel-group, non-inferiority, cluster-randomised controlled clinical trial enrolled adults with HIV taking antiretroviral therapy (ART) at 18 rural, nurse-led clinics in Lesotho. Clinics were randomised to receive either digital documentation and clinical decision support for providers, along with mobile health for participants-comprising individualised automated text messaging and telemedicine-and preference-based multi-month dispensing of ART (VITAL group), or only digital documentation for providers (enhanced standard of care [eSOC] group). The primary endpoint was engagement in care with documented viral suppression (<50 copies per mL) at 24 months (window: 16-28 months). The adjusted odds ratio (aOR) was estimated in the modified intention-to-treat (mITT) population with a non-inferiority margin of 0.8. Safety endpoints were all-cause mortality, tuberculosis diagnoses, and disengagement from care. The trial was registered with ClinicalTrials.gov (NCT04527874) and the trial status is completed. Findings: Between October 14, 2020 and March 30, 2022, 5809 participants were enrolled, of whom 5770 were included in the mITT analysis (3401 in the VITAL group and 2369 in the eSOC group). The primary endpoint was reached by 2649 (77.9%) in the VITAL group and by 1759 (74.3%) in the eSOC group (aOR 1.18 [95% CI 0.95-1.46]). All-cause mortality and tuberculosis diagnosis were similar between groups (80 [2.4%] in the VITAL group and 53 [2.2%] in the eSOC group, adjusted hazard ratio 1.10 [0.78 to 1.58]; 15 [0.4%] in the VITAL group and 14 [0.6%] in the eSOC group, aOR 0.70 [0.30-1.63]). Disengagement from care was lower in the VITAL group (156 [4.6%] in the VITAL group and 167 [7.1%] in the eSOC group; aOR 0.67 [0.48-0.93]). Interpretation: Digital health-supported differentiated service delivery maintained levels of viral suppression and engagement in routine rural HIV care without increasing adverse outcomes. Although superiority was not demonstrated, the findings support the safe integration of digital health tools and multi-month dispensing of ART into HIV care. Mid-trial changes in national ART guidelines may have attenuated differences between groups. Funding: Moritz Straus-Foundation, Swiss National Science Foundation.