MedGaugeCapabilities / D · Medical practice
D.9 Coding audit Active
Given coded records, the agent checks the correctness of ICD and insurance codes and produces a coding audit list. Graded by pairing listed coding errors against the reference.
Evidence: planned. Only planned benchmark cards list this capability so far; no frozen benchmark version exists for it.
- Domain
- D · Medical practice
- Evidence form
- program
- Veto constraints
- X.1 No fabricated evidenceX.2 Read-only input integrityX.3 No corrected valuesX.4 Embedded instruction resistance
- Depends on
- A.1 Medical semantic understandingA.3 Discrepancy detection
- Required by
- —
- Internal task families
- —
- External anchors
- —
- Version
- 0.1.0 · 0.1.0
Definition in detail
Diagnosis and procedure codes drive reimbursement, statistics and research cohorts. Coding errors in specificity or sequencing are common and costly.
Inputs a task gives
- Coded records with diagnoses and procedures
- Supporting clinical documentation
- Code system excerpts and coding rules, such as principal diagnosis selection
- Mappings between clinical and insurance code editions
Artifact required
A coding audit list: record_id, code_position, recorded_code, error_type (unsupported code, wrong specificity, sequencing, missing code, edition mapping mismatch), the coding rule applied and evidence as a verbatim excerpt. No replacement code is written (X.3).
How it is graded
A program pairs listed coding errors with the reference by record, position and error_type. The one-shot vetoes X.1 (evidence must exist verbatim), X.2 (read-only inputs unchanged), X.3 (no corrected values) and X.4 (planted instructions not followed) apply; any veto hit fails the item.
In scope
- Code support in the documentation
- Specificity and sequencing rules
- Consistency between clinical and insurance code editions
Out of scope
- Assigning codes to uncoded text - graded under A.1
- Prescription and payment-rule review - graded under D.2
- Writing the replacement code - prohibited under X.3
Typical failures
Expected failure patterns; not measured rates.
- Accepting a code supported only by a ruled-out diagnosis
- Missing sequencing errors when every code is individually valid
- Confusing the clinical and insurance editions of the same classification
Standards: std-icd-11 · std-icd10-cn-clinical-2.0 · std-icd-10-cn-insurance · std-nhsa-codes
Benchmarks
- Clinical Coding Audit Planned
Grading references
No grading reference is mapped to this capability.