Capabilities / D · Medical practice
D.2 Prescription and insurance review Active
Given prescriptions with the relevant labels, interaction rules and insurance payment restrictions, the agent produces a review list of issues and logs any missing information. Graded by pairing each listed issue and information-gap entry 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 detectionA.6 Research evidence semanticsB.3 Tool invocation
- Required by
- E.1
- Internal task families
- rx_review
- External anchors
- PhysicianBench, HealthAgentBench
- Version
- 0.1.0 · 0.1.0
Definition in detail
Pre-dispensing prescription review and insurance review catch dosing, interaction and payment-restriction problems before they reach the patient or the claim. Both depend on checking each line against the label and rules actually supplied.
Inputs a task gives
- Prescriptions or medication orders with relevant patient attributes and laboratory values
- Drug label excerpts
- Interaction rule tables
- Insurance payment restriction rules
Artifact required
A review list (issue_id, prescription_line, issue_type such as indication, dose, contraindication, interaction, duplication or payment restriction, rule_source, evidence) and an information_gaps list (missing field, the check it blocks).
How it is graded
A program pairs each issue and each information-gap entry with the reference. Missed issues and entries with no reference counterpart are both counted. 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
- Checking each line against supplied label, interaction and payment rules
- Logging information whose absence blocks a check
- Citing the rule and evidence for each issue
Out of scope
- Appraising candidate treatments for a clinical decision - graded under D.19
- Consistency of medications with phenotype and history over time - graded under D.12
- Writing a corrected prescription - prohibited under X.3
- Diagnosis and procedure coding - graded under D.9
Typical failures
Expected failure patterns; not measured rates.
- Applying general knowledge of a drug instead of the supplied label excerpt
- Assuming normal renal function when the value is missing instead of logging the gap
- Flagging a payment restriction whose conditions the record shows are met
Standards: std-rxnorm · std-nhsa-codes
Benchmarks
Grading references
No grading reference is mapped to this capability.