Templates Healthcare

Healthcare

Prior Auth Assembler

Drafts clinical justifications for prior authorization requests and flags missing fields before submission.

What's inside

  • Prior Auth Requests entity — patient, payer, procedure code, diagnosis codes, clinical history, drafted justification, missing-fields checklist, status
  • Prior Auth Review app — internal-only utility for clinicians to review AI-drafted justifications before submission
  • DraftJustification workflow — fires on each new request; AI drafts the justification, lists missing fields, and marks the request *ready* or *draft*
  • 3 sample requests: one ready to submit, one with missing fields, one mid-revision

Prior Auth Assembler

Cut the time clinicians spend drafting prior authorization justifications. Each new request kicks off a structured AI draft that produces both a payer-ready paragraph and a checklist of anything that’s likely to cause a rejection.

How it works

  1. A clinician (or upstream integration) creates a new request in the Prior Auth Requests entity with the basics: payer, procedure code, diagnosis codes, clinical history.
  2. The DraftJustification workflow fires:
    • Reads the request details.
    • Uses Claude to produce a structured JSON response containing the drafted justification, a missing-fields checklist, and a ready flag.
    • Writes the draft and the checklist back onto the record, and sets the status to ready or draft based on completeness.
  3. The Prior Auth Review app surfaces ready-to-submit requests at the top, drafts that need more work below.

What you can extend

  • Wire the ready requests into your payer submission API.
  • Add a workflow that re-fires the AI draft when a clinician edits the history.
  • Add an enum for payer-specific templates if your justifications need to vary by plan.
  • Track approval/denial outcomes on a follow-up entity and feed them back into the AI prompt.

Data handling

Clinical history and patient identifiers flow through HASP’s policy-aware AI gateway — your org’s data-handling configuration is enforced automatically before model inference. You can build and test the template on Free Evaluation; the underlying policy enforcement is the same posture every paid org gets.