Healthcare systems increasingly resemble fragmented ecosystems in which patients must be routed across autonomous organizations, heterogeneous institutional logics, and regulated yet loosely coupled digital infrastructures. In these settings, digital triage tools have become prominent entry points for navigating acute care. However, prior research evaluates symptom checkers and triage systems predominantly in terms of diagnostic and triage accuracy, offering limited guidance on how to design artifacts that reliably translate medically plausible assessments into actionable navigation under level-specific capacity and resource constraints. This omission is consequential because a correct urgency classification creates limited coordination value if it cannot be enacted within the concrete infrastructure of a given care level. This study reconceptualizes digital triage as an ecosystem orchestration problem and develops prescriptive design knowledge for cross-level patient navigation. We report an eight-year Action Design Research study (2017–2025) conducted with a German digital health provider in which a Healthcare Ecosystem Orchestration (HEO) artifact was developed across three building–intervention–evaluation cycles. The artifact evolved from a monolithic symptom checker into an orchestration artifact and yields four interdependent design principles. We evaluate the utility of the layered instantiation in a naturalistic, ex post assessment using 48,516 real-world navigation cases (September–December 2025) across supra-regional, regional, and local entry contexts. Urgency distributions differ strongly across levels, indicating that uniform triage logic is structurally misaligned with ecosystem-level variation and that effective digital navigation in fragmented healthcare ecosystems profits from configurable orchestration mechanisms that mediate between patient, clinical, and resource logics across levels.