Executive Decision Pack

SAMPLE-05 · Radiology AI triage assistant (hospital)
Generated 2026-10-31 14:52 UTC · Owner: Dr. Elena Novak · Sponsor: Chief Medical Information Officer

DECISION
APPROVE WITH CONDITIONS
Trust Score: 59/100 · ARA residual: Medium · CALIBRE band: High (GII 57.50) · Evidence coverage: 80%

Scenario

A tertiary hospital wants to deploy an AI system that reads incoming CT scans and prioritises the reading queue for radiologists, flagging suspected acute findings (stroke, haemorrhage, pulmonary embolism) for immediate attention. The AI does not diagnose; it re-orders the queue. Model is FDA-cleared but hospital will validate against local population.

Decision narrative

Initiative SAMPLE-05 (Radiology triage AI) is assessed at ARA overall-residual Medium, CALIBRE band High (GII 57.50/100). The advisory (queue-ordering) nature of the system, and the fact that radiologists retain full diagnostic authority, keeps residual risk manageable.

Approve with conditions. Do not proceed to production until the four conditions are met. The dominant residual risk is automation bias, which no software control can fully eliminate; the training and continuous-monitoring controls are the primary mitigations.

Top risks driving the decision

  1. R-CL-01 False-negative deprioritisation
  2. R-CL-02 Automation bias
  3. R-RG-01 TGA post-market surveillance obligations

Conditions for approval

  1. Local-population validation complete with sensitivity meeting acceptance criteria
  2. Automation-bias training delivered to all reading radiologists
  3. Continuous performance monitoring live before go-live
  4. TGA post-market vigilance process documented and tested

Evidence attached

None yet attached.

Remaining evidence gaps

None outstanding.

Actions and owners

IDActionOwner rolePriorityStatus
SAMPLE-05-A-01Complete Local-population validation (500 scans)Head of RadiologyHighOpen
SAMPLE-05-A-02Deliver Radiologist automation-bias trainingHead of RadiologyHighOpen
SAMPLE-05-A-03Establish Continuous performance monitoringAI Product OwnerHighOpen
SAMPLE-05-A-04Document TGA post-market vigilance processRegulatory AffairsHighOpen
Approved by (name and role):

____________________________
Signature:

____________________________
Date:

____________________________

Audit trail: full machine-readable finding history, model IDs, timestamps, and evidence citations are stored in the Governance Fabric under initiative SAMPLE-05. Regenerate this pack at any time from the initiative page.