Pending independent pharmacist review before submission.
The software pipeline, citations, deterministic checks, and adversarial reviewer have been technically verified. The verifier fails closed: if claim-vs-source review cannot complete, the finding is not rendered. Clinical validation is a separate gate. Until a licensed pharmacist has reviewed the synthetic case, this repository does not claim that the ranking or patient-specific framing has been clinically validated.
The review uses only the synthetic HF/AFib discharge fixture in data/fixtures/discharge-hf-afib.json. It contains no PHI.
The pharmacist will assess:
- Whether the synthetic case is clinically plausible.
- Whether the ranked findings are factually and clinically framed without overstatement.
- Whether warfarin plus newly started amiodarone is reasonably ranked first for this case.
- Whether the monitoring language works as a review prompt and does not read as an autonomous prescription.
- Whether a clinically important concern is missing or incorrectly prioritized.
- Whether the source expansion and audit trail are useful during medication reconciliation.
- No rendered statement is clinically misleading.
- Patient-specific uncertainty is preserved. For example, the report says INR may rise after the current reading, not that it certainly will.
- The ranking is accepted by the pharmacist or corrected before submission.
- Any requested correction is implemented and re-verified against its cited source.
- A testimonial is used only if it reflects the pharmacist's spontaneous assessment.
| Field | Result |
|---|---|
| Review date | Pending |
| Reviewer credential and practice context | Pending consent |
| Synthetic-case plausibility | Pending |
| Ranking assessment | Pending |
| Claims requiring correction | Pending |
| Missing concern identified | Pending |
| Workflow usefulness | Pending |
| Quote approved for submission | Pending |
| Video/name/credential consent | Pending |
One pharmacist review is formative product validation, not a clinical study, prospective safety evaluation, or evidence of improved patient outcomes. Sourced remains a review aid for licensed professionals and does not diagnose or prescribe.