Validation
Measured, not magical.
OpAgent publishes measured results with the methodology attached. RadiologiQ production volume is live. CodeIQ OR figures are scoped engine validation — not proof that CodeIQ or OpAgent Wrap is deployed through Essentials.
MEASURED PERFORMANCE
Weighted combined agreement across 359 held-out surgical cases spanning 5 procedure families, scored against the operating surgeon's own billing-record coding — a historical agreement benchmark, not clinical ground truth.
n=359 locked test set · April 2026 · scoped to CodeIQ OR, not implied across agents
PRODUCTION
Live production count since April 2026. Every report reviewed and signed by a physician.
How this was measured
The validation dataset is 3,951 de-identified operative cases (2018–2026) from a single high-volume arthroplasty and trauma practice, held as a Limited Data Set. Cases were split with a fixed seed into few-shot, evaluation, and locked-test partitions; the locked test set was physically quarantined and never viewed during prompt iteration. The comparison target is the operating surgeon's billing-record CPT assignments. Prompt changes were accepted only when the evaluation set improved. 4.2% of cases self-flagged as ambiguous and routed to human review. Historical surgical coding agreement, not clinical ground truth, not adjudication, and not implied across other OpAgent agents.
| Procedure family | Test n | Combined agreement |
|---|---|---|
| Hip arthroplasty | 196 | 98.5% |
| Knee arthroplasty | 140 | 96.4% |
| Hip trauma | 11 | 90.9% |
| Knee non-arthroplasty | 10 | 100.0% |
| Hip infection | 2 | 100.0% |
| Weighted | 359 | 97.5% |
Scope: this is scoped engine validation of CodeIQ OR surgical coding only. It is not a claim that CodeIQ, documentation closeout, or OpAgent Wrap is currently deployed through Essentials. The comparison target is historical billing-record CPT assignments, not clinical ground truth. It is not a claim about other OpAgent capabilities, other specialties, or national representativeness. Sample size, methodology, and date update together when the validation is re-run.
In production
RadiologiQ has drafted 6,026 imaging reports in production since April 2026 — every one reviewed and signed by a physician. The count above is live from the production system, not a marketing snapshot. It is not a paid-rate lift.