Completeness / omission
Clinically relevant content missing from output; fact-list enumeration then check.
What completeness / omission means
Clinically relevant content missing from output; fact-list enumeration then check.
Why it matters. Missing content is the hardest error to see, because the output still reads well.
A summary of a visit never mentions the patient's anticoagulant, so the next clinician plans a procedure without knowing.
How to measure it
Define the unit
Decide whether you count per sentence, per note, per patient or per call.
Sample on purpose
Pick cases that cover specialties, languages and rare events, not only easy ones.
Judge with clinicians
Clinicians label a sample. Graders are then checked against those labels.
Report the margin
Show an interval, and the grader's own error rate beside the score.
A single average hides rare, serious errors and weaker subgroups.
A number with a clear unit, a stated sample, a margin and the grader's agreement with clinicians.
Report the unit (per sentence, per note, per patient), the sampling method, the severity scale, and how often the automated grader agreed with clinicians compared with how often clinicians agreed with each other.
Reviewed by the Precision research team.
How strong is the evidence
E2 · Preprint, benchmark page or vendor-run result. Backed by numbers from a preprint, a benchmark page or a vendor-run test. Treat as strong but not settled. At least one source is an unreviewed preprint.
What the sources say
Largest HealthBench axis (completeness 39%, accuracy 33%); omission judges near chance (arXiv 2608.31016) (unreviewed).
- Clinician judgement. Needs a clinician to judge or label, because the answer depends on clinical meaning.
- Automatable. Can be scored by software, including an AI grader that is spot-checked against clinicians.
- arXiv 2608.31016arxiv.org
Found something out of date? Write to hello@precision.example.
Agents that measure this
Quick answers
What is completeness / omission in healthcare AI?
Clinically relevant content missing from output; fact-list enumeration then check. It sits in the every product type group.
Why does completeness / omission matter?
Missing content is the hardest error to see, because the output still reads well. A summary of a visit never mentions the patient's anticoagulant, so the next clinician plans a procedure without knowing.
How is completeness / omission measured?
Define the unit, then Sample on purpose, then Judge with clinicians, then Report the margin. Report the unit (per sentence, per note, per patient), the sampling method, the severity scale, and how often the automated grader agreed with clinicians compared with how often clinicians agreed with each other.
What should I watch out for with completeness / omission?
A single average hides rare, serious errors and weaker subgroups. A good report gives: a number with a clear unit, a stated sample, a margin and the grader's agreement with clinicians.
Can completeness / omission be automated?
Only in part. It needs clinician judgement and automatable. Software can screen, but a clinician must judge the cases that matter.