Transforming routine health data use in LMICs through modular, AI-supported automation: insights from Zimbabwe

Abstract Health information systems (HIS) in Low- and Medium-Income Countries (LMICs) are often hindered by fragmented data flows, manual reporting processes and limited analytical capacity. These challenges compromise data quality, divert critical resources from patient care, delay reporting and limits the use of routine data for programme improvement. This descriptive case study documented the design, co-creation and rollout of the Organization for Public Health Interventions and Development Modular Data Intelligence Platform (OMDIP) across 15 districts in Zimbabwe. System performance and user experience were assessed through routine metrics, dashboards, supervision reports and user feedback collected between January 2023 and June 2024. The reporting of the intervention was guided by selected domains of the WHO mHealth Evidence Reporting and Assessment checklist. Development of the OMDIP began in May 2023. Additional modules were added: ReportAID AI enabled module for narrative synthesis, Data Diagnostic Module for error detection, the Data Analytics Platform for visual dashboards and the Data Export Request Listener for automated submission ready reports. These modules integrated with District Health Information System 2 and EHRs. Across 335 facilities supporting 345 000 clients, timely report submission improved from 27% to 100%, data-cleaning time decreased from 10.2 to 2.9 days, report preparation time dropped from 7 to under 2 days, and critical data errors were eliminated. OMDIP enhanced efficiency, quality and use of routine health data in Zimbabwe. Integrated with national systems and aligned with WHO digital health frameworks, it demonstrates a scalable model for strengthening data-driven decision-making and health system performance in LMICs.