Severity grader

Ranks every error by how much harm it could do.

In audits Layer 2 · Measure
Takes inError list
The agentSeverity grader
Hands backSeverity-weighted error rate

Checked against clinicians

Agreement with clinician severity ratings.

Where a person decides

Clinician adjudicates all major and critical.

Works on

All.

Delivered inside a Quality Audit today. Reviewed .

The problem it solves

A count of errors hides the point. A missed allergy and a missed greeting aren't the same, and teams need to fix the worst first.

How it works

01

Classify the error

The agent names the error type, such as omission, invented or flipped.

02

Judge potential harm

It estimates whether the error could change diagnosis or management if uncorrected.

03

Record whether it reached the patient

Errors caught before use are tracked separately.

An example, illustrative

Before

14 errors found.

With the agent

2 major (could change care), 9 minor, 3 trivial. Fix the 2 first.

What it needs from you

A severity scale agreed with your clinicians.

Its limits

Severity is a judgement. Clinicians audit a sample and the agent reports its agreement.

Quick answers

What does the severity grader do?

Classify errors by type, potential harm and reached-patient status.

How is the severity grader checked against clinicians?

Agreement with clinician severity ratings.

What counts as 'major'?

An error that could change diagnosis or management if nobody corrected it. We agree the scale with your clinicians.

Why separate errors that reached the patient?

Reviews catch many errors. The ones that got through measure the real risk.

What can the severity grader not do?

Clinician adjudicates all major and critical. No agent makes a clinical decision.

Use the severity grader on your product