Socio-Technical Evaluation of an EHR–PHR Interoperability Model for Tuberculosis Care Using SEIPS 3.0: A Pre-Deployment Mixed-Methods Study in Indonesia
Abstract
Technical exchange standards alone do not guarantee successful health information interoperability; adoption depends on work-system, governance, and organizational factors as well as data-exchange capability, yet socio-technical evaluations remain scarce in low- and middle-income countries (LMICs). This pre-deployment study evaluates a seven-component electronic health record (EHR)–personal health record (PHR) interoperability model for tuberculosis (TB) care, implemented as a six-layer Fast Healthcare Interoperability Resources (FHIR) R4 architecture aligned with Indonesia’s national SATUSEHAT exchange, using the Systems Engineering Initiative for Patient Safety (SEIPS) 3.0 framework. A formative–summative mixed-methods design triangulated four evidence streams: a multi-stakeholder contextual inquiry with 93 participants across three regions (60 TB patients, 7 family caregivers, and 26 facility, district, vendor, and national-level professionals); sandbox technical verification; expert judgment; and user/stakeholder evaluation including System Usability Scale (SUS) assessment. A component was rated as strong, adequate, or needing reinforcement only where at least two streams converged. Four components were strong, two adequate, and one in need of reinforcement; all overall technical thresholds were met, but two components required organizational reinforcement. Verification showed 98.9% FHIR conformance (184/186 assertions; two failures reported), zero critical security findings, and zero offline data loss. Field evidence indicated organizational gaps: all 14 facility officers ran a facility EHR alongside the national SITB register across four EHR products with no automatic synchronization, while only 3 of 41 patients had ever been told how their data are used and 29 of 42 wanted a say in data sharing. Mean SUS was 71.8 (95% CI 66.5–77.1). All technical testing used synthetic data; clinical outcomes are projected. The findings identify work-system alignment and patient-facing data governance as priorities for subsequent implementation in LMIC disease programs.