Currently submitted to: JMIR Formative Research
Date Submitted: Apr 30, 2026
Open Peer Review Period: Apr 30, 2026 - Jun 25, 2026
(closed for review but you can still tweet)
NOTE: This is an unreviewed Preprint
Warning: This is a unreviewed preprint (What is a preprint?). Readers are warned that the document has not been peer-reviewed by expert/patient reviewers or an academic editor, may contain misleading claims, and is likely to undergo changes before final publication, if accepted, or may have been rejected/withdrawn (a note "no longer under consideration" will appear above).
Peer review me: Readers with interest and expertise are encouraged to sign up as peer-reviewer, if the paper is within an open peer-review period (in this case, a "Peer Review Me" button to sign up as reviewer is displayed above). All preprints currently open for review are listed here. Outside of the formal open peer-review period we encourage you to tweet about the preprint.
Citation: Please cite this preprint only for review purposes or for grant applications and CVs (if you are the author).
Final version: If our system detects a final peer-reviewed "version of record" (VoR) published in any journal, a link to that VoR will appear below. Readers are then encourage to cite the VoR instead of this preprint.
Settings: If you are the author, you can login and change the preprint display settings, but the preprint URL/DOI is supposed to be stable and citable, so it should not be removed once posted.
Submit: To post your own preprint, simply submit to any JMIR journal, and choose the appropriate settings to expose your submitted version as preprint.
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Artificial Intelligence for Dynamic Risk Assessment and Outcome Optimization (AIRO): A Governance-Driven Framework for Machine Learning in Cardiac Surgery
ABSTRACT
Background:
Contemporary cardiac surgical risk assessment is largely based on static regression-derived models that do not account for evolving perioperative data or integration within clinical workflows. Although machine learning approaches offer improved capacity to model complex relationships, most existing applications remain retrospective, outcome-specific, and disconnected from real-world clinical implementation.
Objective:
To describe the development of the Artificial Intelligence for dynamic Risk assessment and Outcome optimization (AIRO) platform, a clinician-led, governance-driven framework designed to support dynamic, explainable risk assessment and clinical decision support across the perioperative continuum.
Methods:
AIRO was developed using a formative research approach integrating retrospective model development with prospective implementation design within a learning health system framework. Baseline predictive modelling (AIRO 1.0) was conducted using a prospectively maintained cardiac surgery registry comprising approximately 30,000 patients and over 13 million data elements. A standardized modelling pipeline was established to support development of multiple outcome-specific models. The AIRO conceptual framework was structured around three layers: population-level risk stratification, individual-level risk explanation, and dynamic outcome optimization. For clinical deployment (AIRO 2.0), a system architecture was designed to enable real-time model inference within the electronic health record using interoperable data standards (FHIR), supported by a dedicated integration layer and governance framework. An implementation and evaluation strategy was defined to assess feasibility, usability, and clinician adoption.
Results:
The AIRO platform was developed as an integrated system combining standardized model development, a three-layer conceptual framework, and a scalable architecture for electronic health record integration. Baseline models have been developed for multiple clinically relevant outcomes, including surgical site infection, postoperative delirium, and discharge disposition. A workflow-integrated deployment strategy was defined, intended to enabling near–real-time risk estimation and delivery of explainable outputs within the clinical environment. Governance structures were established supporting data security, auditability, and model oversight. The platform is designed to support prospective deployment and evaluation within routine perioperative care.
Conclusions:
AIRO represents a conceptual and implementation-focused approach to integrating explainable machine learning into perioperative clinical care. By combining standardized modelling, EHR-integrated deployment, and a governance-driven framework, the platform is designed to support clinically meaningful risk assessment and decision support across the care continuum. Further work is required to evaluate its impact on clinical practice, clinician adoption, and patient outcomes.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.