Intake, Pharmacy Queue & pharmacist decisions
Collect consented information, hand off cases and keep every decision with a pharmacist rather than AI.
On this page · Jump to a section
Before you begin
Approved environment gates, a reviewed protocol and a pharmacist responsible for the queue.
Where to go
Pharmacy Queue → Case detail
Visible menus may vary with language, permissions, app version and shop configuration.Follow these steps
- Customers see the AI-not-pharmacist notice and consent before intake. Questions come one at a time from an approved protocol, not a model-invented set.
- Server rules check red flags, missing and conflicting fields. The customer confirms the latest summary before queueing; summary confirmation is not medicine approval.
- A pharmacist claims the case and reviews source messages, summary, gaps, conflicts and evidence, then may request more information, edit, reject, refer or escalate through the UI.
- Approve/reject/refer follow guarded states and actors. Only a licensed pharmacist may approve, and approved customer wording is the pharmacist's own typed text.
- AI medication suggestions are pharmacist-only drafts: explicitly generate, use/edit and approve them. They never reach customers or create orders automatically.
How to verify success
The test case follows valid states, audit metadata omits raw health data and unlicensed users cannot approve.
Limits & things to check
Intake/AI/protocol gates default off and require deployment-specific DB/provider sandbox evidence. Reading this guide alone does not authorize production rollout.
Ready to try it in your shop
Lessons are free to read. Actual operations require sign-in and the appropriate permission. Use test data for training.