Enterprise IT
Build, integrate, secure, operate, and support the digital environment reliably.
Institutional clinical informatics / The philosophy
Making institutional
intelligence actionable.
Healthcare technology requires technical intelligence, clinical-informatics intelligence, and deep expertise from the clinical domains. Governance makes their interaction routine rather than accidental.
A Philosophy of Institutional Clinical Informatics · Justin Halls, MD
This is my synthesis of established clinical-informatics, sociotechnical-systems, CDS-governance, and learning-health-system literature. It brings those ideas together into a practical philosophy for how an institution learns, decides, builds, and improves.
The model in two minutes
Three forms of intelligence.
A deliberate partnership.
Enterprise Clinical Informatics
Align information and technology with clinical work, evidence, people, and institutional goals.
Distributed clinical expertise
Bring the depth of knowledge needed to determine what is clinically appropriate, practical, and valuable in each domain.
These are complementary forms of expertise. None substitutes for the others.
How the institution acts
Governance is an operating system,
not an org chart.
It establishes where needs enter, whose expertise is required, who can decide, how priorities are set, how disagreements are handled, and who remains responsible for results.
Frontline experience must be able to reach the system. Decisions, constraints, and explanations must return to the people who raised the issue.
The five aims
What this should accomplish.
- 01
Increase signal while reducing noise
Expand useful CDS while reducing unnecessary clinician-facing interruption through better targeting, specificity, workflow integration, and lifecycle management.
- 02
Surface and navigate friction
Create trusted mechanisms to identify friction points, bring the appropriate expertise to them, give them accountable disposition, and explain constraints.
- 03
Shorten the distance from clinical need to institutional action
Make it easier for legitimate frontline needs to reach people with the authority and expertise to evaluate and act on them, with a visible path to disposition.
- 04
Measure after implementation
Deployment begins evaluation. Measure adoption, workflow effects, outcomes, unintended consequences, and continuing value. Modify or retire interventions when appropriate.
- 05
Measure value, not merely activity
Tickets, builds, alerts, meetings, and completed projects measure activity. Informatics should increasingly measure patient, clinician, operational, safety, quality, and institutional value.
The clearest exemplar
More useful CDS.
Less unnecessary interruption.
Clinical decision support makes the model concrete: clinical content, targeting, format, channel, and timing all require the right expertise. A portfolio can keep developing while poor-performing interventions are redesigned or retired.
The destination / A learning health system
“The purpose of informatics governance is not to control technology. It is to make institutional intelligence actionable.”