Assess whether a risk model is calibrated for this population (d9d8ae)

Situation

A model vendor claiming AUC that the local validation cannot reproduce put unit-level complication variation table in front of pediatric protocol committee chair in a children's hospital rewriting a sepsis protocol. This Healthcare / Outcomes Review close is a risk model is calibrated from unit-level complication variation table, and the live options are Proceed under protocol, Pause the pathway, Escalate safety review.

Decision

Pediatric protocol committee chair in a children's hospital rewriting a sepsis protocol must choose Proceed under protocol / Pause the pathway / Escalate safety review / Hold using unit-level complication variation table after a model vendor claiming AUC that the local validation cannot reproduce.

Hypotheses to test

• Unit-level complication variation table reads as Proceed under protocol once a model vendor claiming AUC that the local validation cannot reproduce is lined up to the same Healthcare population.
• Unit-level complication variation table is closer to Pause the pathway after a model vendor claiming AUC that the local validation cannot reproduce; Proceed under protocol would over-claim this Outcomes Review extract.
• Escalate safety review is still live in unit-level complication variation table for pediatric protocol committee chair in a children's hospital rewriting a sepsis protocol.
• Unit-level complication variation table is missing the fact pediatric protocol committee chair needs after a model vendor claiming AUC that the local validation cannot reproduce; stop this Healthcare close.

Analysis required

• Separate a documented exception from an OCR-relevant gap in a children's hospital rewriting a sepsis protocol.
• Validate clinical-outcome evidence and protocol steps in unit-level complication variation table after a model vendor claiming AUC that the local validation cannot reproduce.
• Trace access logs and outputs to the rule pediatric protocol committee chair must apply.
• For this Healthcare Outcomes Review file, read unit-level complication variation table against a model vendor claiming AUC that the local validation cannot reproduce and write the one fact that would move a risk model is calibrated for pediatric protocol committee chair.

Recommendation