Reception is a relay, not a form. A patient moves from arrival to identity check, visit creation, a queue, a service point, and sometimes payment or another desk. Every handoff can lose context or expose too much of it.
Judge the software by that relay.
Takeaways
- Test patient matching before celebrating fast registration.
- Give every queue handoff an owner and visible state.
- Keep front-desk displays and spoken details to the minimum necessary.
- Reconcile downtime records without creating duplicate patients or visits.
This is an evaluation worksheet, not clinical, identity, legal, billing, ABDM, or procurement advice. The hospital’s authorised owners set the workflow and escalation rules.
Name the arrival states
A single “checked in” flag hides too much. Ask the vendor to show a patient moving through explicit states.
ARRIVAL RELAY
expected appointment / walk-in / referred arrival
-> patient search and match review
-> demographic verification or controlled registration
-> visit and service selection
-> payment, eligibility, or exemption handoff if applicable
-> queue placement
-> service-point acceptance
-> completion, transfer, cancellation, or exception
Each transition needs a timestamp, actor or system source, reason, and next owner. A queue token should not become the patient identity. A payment state should not become clinical clearance.
NIC’s e-Hospital module list describes registration work such as OPD and casualty registration, appointment booking, confirmation, and cancellation. The same page describes a Queue Management mobile application that can add people to queues with or without ABHA. Those are facts about that government platform, not mandatory design rules for every hospital or proof of Patient Square integration.
Search before creating
Use synthetic records to test four arrivals:
| Arrival | Expected behaviour |
|---|---|
| exact known patient | show controlled match evidence and current details |
| possible duplicate | pause creation and route an authorised review |
| no reliable match | create under policy without inventing identifiers |
| wrong selection found later | correct through a retained, authorised history |
Change the order of a name, omit a mobile number, and use a shared family phone. Ask which fields drive a match and how staff avoid learning more than they need. A product should show uncertainty instead of silently selecting the closest record.
Break the happy path
Run an arrival drill with invented data:
- a booked patient arrives at the wrong department;
- a walk-in needs an authorised priority escalation;
- a payment service is unavailable;
- a patient does not want their name announced;
- the queue display fails after token creation;
- the main service returns after paper downtime registration.
For every case, record who owns the patient now, what appears to the next team, what remains private, and how staff close the exception. The clinic queue guide examines token flow in more depth. The visitor-management test covers non-patient entry and should remain a separate lane.
Treat the waiting area as a privacy surface
Reception exposes data through monitors, labels, printed slips, spoken questions, overheard calls, and unattended screens. During a demo, stand where a visitor would stand. Check what can be read or heard. Test screen timeout, print pickup, queue-display configuration, role access, and correction history.
India’s Digital Personal Data Protection Act, 2023 and final Digital Personal Data Protection Rules, 2025 call for date-specific implementation review because commencement is phased. Qualified counsel should map the active provisions and other applicable duties to the hospital’s actual purposes, roles, notices, safeguards, rights process, incidents, retention, and vendors.
Reconcile a downtime morning
Create two synthetic paper arrivals while the system is unavailable. When service returns, search again before entry, attach each record to the right patient or controlled new registration, preserve the original time and source, and close the paper copy under policy. Try entering the same patient twice. The software should expose the collision without deleting history.
Score the drill on correct identity, preserved arrival order, accountable overrides, privacy, and full reconciliation. Speed comes later.
Patient Square does not publish a standalone hospital-reception module. Do not infer registration, kiosk, payment, eligibility, token-display, or queue capability from Hospital Copilot. Use the hospital management system explainer to separate modules and inspect current security information on the India security page. The strongest reception workflow is not the one with the fewest clicks. It is the one that keeps the right person, visit, queue, owner, and exception connected.