Toward Healthcare System Digital Mission Engineering: A Multilevel Health Informatics Framework for AI Integration, Demonstrated Through Cardiovascular Risk Management

Abstract

Artificial intelligence increasingly operates within interconnected healthcare information systems, but governance remains model-centered and weakly tied to institutional mission. Healthcare System Digital Mission Engineering (HSDME) adapts digital mission engineering to treat the mission, rather than an application or model, as the primary system of interest. The framework links three architectural levels through six stages, fourteen tasks, and a bidirectional governance and quality-assurance thread. Two nested cardiovascular cases illustrate system-level adoption of the 2026 dyslipidemia guideline and encounter-level reuse of the resulting architecture. Structured author-led walkthroughs assessed specification completeness, consistency, traceability, and reuse; deployment and clinical effectiveness were outside their scope. An illustrative cascade analysis compared Type A (no AI-enabled functions under the HSDME rubric), Type B (pathway-level AI), and Type C (AI-enabled functions in the shared pipeline or governance layer under common system-level controls). With stipulated parameters, greater integration raised decision yield and closed-loop evidence return while increasing human-review burden; no configuration dominated. HSDME contributes a mission-derived multilevel architecture, the three-type AI-integration typology, differentiated evidence routing, and a governed organizational boundary with typed information provenance. It positions health informatics as a coordinating discipline for mission, computable knowledge, interoperable pipelines, and accountable decision support.