SurgeonValue · Prior Auth Agent

Paste a clinical note.
Get a peer-to-peer letter.

Generate a medical-necessity letter ready to fax. Cites real published guidelines, addresses the most common payer denials for the procedure, and is grounded only in what the note actually documents. Every letter is matched to the criteria a reviewer is actually checking — the specific evidence they need, not a 200-page dump that buries it. A physician reviews and signs.

Payers automated the denial. Nobody automated the appeal — until now.

Insurers run algorithms that deny at machine speed and near-zero cost, while practices still appeal by hand — one chart, one letter, one peer-to-peer at a time. That asymmetry, not medical necessity, decides most outcomes: the appeals that would win are simply never filed, because the hours don't exist.

~75%
of appealed Medicare Advantage denials were overturned by the plans themselves — yet only a small fraction of denials are ever appealed (HHS Office of Inspector General)
~13 hrs
of practice staff time per physician, every week, consumed by roughly 40 prior-auth requests (AMA Prior Authorization Physician Survey)
1 in 4
physicians report prior authorization has led to a serious adverse event for a patient in their care (AMA Prior Authorization Physician Survey)

This agent doesn't out-argue reviewers — it removes the labor that made winnable appeals not worth filing, so persistence finally scales to match the denials. And because prior auth has quietly become a consolidation engine — large platforms absorb the administrative cost while independent practices hire another FTE for the same battle — making the appeal as cheap as the denial is part of how an independent practice stays independent.

The care journey has two phases. SurgeonValue operates in the second.

Pharma now frames the patient journey as Pre-RX (identify, diagnose, guide) and Post-RX(onboard, support, retain). Sanofi's iCare+ platform coordinates 20+ therapies across 50+ external partners for the Post-RX phase — reimbursement, specialty pharmacy, documentation. The model is proven at scale.

Independent surgical practices face the same Post-RX complexity — but without the pharma platform. The prior-auth letter is one node in that coordination layer: the point where a diagnosed, indicated patient can still be blocked from treatment by administrative friction. This agent drafts the appeal; the attestation architecture ensures someone accountable signs it.

Pre-RX
Diagnosis, patient identification, clinical indication. SurgeonValue does not operate here.
Post-RX
Prior auth, reimbursement coordination, peer-to-peer, appeal persistence. This is where the agent lives.

Source: Sanofi Digital & AI Patient Care Journey, July 2026. The 70% "never reach prescription" stat cited in the source is from pharma marketing and has not been verified against a peer-reviewed primary source — it is not deployed here.

Six states now require a human to review every AI denial. The architecture this agent uses is now the law.

In 2026, Alabama, Colorado, Georgia, Iowa, Utah, and Washington enacted statutes that prohibit insurers from basing coverage denials solely on AI — a licensed clinician must review every adverse decision. Indiana and Illinois added parallel requirements for AI-driven downcoding. The legal framework has caught up to what was already obvious: an algorithm can screen, but only a person with a license can deny.

AI may screen
Initial triage, pattern matching, documentation check — the automatable layer.
Human must deny
Licensed clinician review, individual medical history, accountable signature — the legal requirement.

Source: Sam Ashoo, MD, "A Complete Guide to AI Healthcare Laws in 2026," July 29, 2026. This agent's design — AI drafts the letter, a physician reviews and signs — matches the statutory requirement in every jurisdiction that has enacted prior-auth AI regulation.