Rescheduling an appointment: availability, confirmation and limits

When a patient changes an appointment through their link, reception needs to distinguish the available time, the recorded change and the new confirmation. We tested these Clínicas Pro rules in an isolated environment and present the results and the test's limits.

A technical test of Lyntev's own product, using fictional data and SQL rules executed in a temporary in-memory database. The interface journey, real authentication and message delivery require their own checks.

What happened in the valid scenario.

In the valid scenario, the appointment moved from 10:00 to 11:00 the following day. The new date and time were saved; the status was 'scheduled' and patient confirmation was still pending. The history retained the previous date and time, and a notification for the clinic was prepared in the sending queue.

A representation of the states confirmed by the test. “Day +5” and “day +6” are days relative to the test date.
  1. Original appointment

    Day +5, at 10:00

    Status: confirmed

    Fictional clinic, professional, service and patient; appointment calendar set up from 08:00 to 18:00.

  2. Choosing a slot

    Day +6, at 11:00

    Time returned by the appointment calendar

    The availability query included 11:00 among the slots for the following day.

  3. Rescheduled appointment

    Day +6, at 11:00

    Scheduled; confirmation pending

    After the change, the record was scheduled and had no patient confirmation.

  4. Continuity for the clinic

    History and notification

    Change recorded; notification queued

    The event saved the previous date and time. The queue contained a notification with the new time; this test did not send any email.

The notification remained in the sending queue. This result does not prove that the clinic’s email was sent or received.

What was accepted, refused or remained pending.

The results below were confirmed by the test assertions for that edition. The rules and limits described correspond to the scenario tested.

Test checks using fictitious data
ConditionObserved result
Appointment confirmed for 10:00, five days after the testThe rules returned an indication that the appointment could be rescheduled.
Slots for six days after the testThe availability list included 11:00.
Request for forty days after the testThe list was empty; this search offered no slots.
Change to 03:00 or to an invalid time, 25:00Both operations were refused.
Valid rescheduling for the following day, at 11:00New date/time saved; status 'scheduled'; patient confirmation set to false.
Rescheduling historyThe change event retained the previous date and time.
Notification queue for the clinicNotification prepared with the new time, without the patient's or service's name in the tested content. Delivery was not verified.
Second change and an attempted third changeAfter the second change, rescheduling was no longer available; the slot query and the third change were refused.
Clinic with rescheduling disabled, in a separate scenarioThe link no longer offered rescheduling.
Slot query without the expected server roleAccess using the simulated authenticated role was refused; the tested function was reserved for the server.

How we performed the verification.

Performed on 20 November 2026, between 21:51:16 and 21:51:18 UTC.

We prepared a temporary in-memory PGlite database and applied the 76 schema and rule migrations from the version identified below. The clinic, professional, service and patient were data created for the test.

We ran the existing test without changing it. Function calls and the authentication identity were represented directly in that database. The assertions checked the saved states, events and queue, including situations that should be refused.

The rules in this version define a 30-day search window. The test checked a request for six days later and another for forty days later; it did not test every boundary of that window.

The run finished with exit code zero and two blocks of checks passed. These numbers describe this technical run; they do not count customers, real appointments or journeys passed in a browser.

Download the protocol and test results

The document includes the scope, results and version references. It does not provide private code or customer data.

Technical identification of this execution
Code edition tested
80e011318a511ef85b1191f8d09fb10c05558cf8
Test SHA-256
179bd5080e06e0bb9bd7e5d07b0da7d8032f1ebf22cf08103f04370c08203820
Environment preparation SHA-256
54c384a79bcb54b561b5b09970b363e36d1542e56ee7e08c583220f5b4cbc18d
Migration manifest SHA-256
50d5eb735dfae101e46d36b69b2695bda9e36e95edd04dfbef270d3e060babca
Downloadable protocol SHA-256
7ab4e2809c59e493fa4e59ae8600e88cd4855d5a96f06cab693819efb449458b

These fingerprints identify the files associated with the execution record; they are not external certification.

The scope of this evidence.

  • This is a software test with fictional data, not a clinical trial or a customer testimonial.
  • Real login, Supabase authentication, HTTP requests and a user interface were not exercised. The diagram on this page represents the verified states; it is not a screenshot of the product.
  • The notification was checked in the queue. No email was sent or received, and there is no evidence that a notification was delivered.
  • Some scenarios reset test data to assess the clinic's setting separately. That preparation does not mean that a user can delete the history or bypass limits.
  • Concurrent appointment booking, performance under load, remote consultations and availability of the hosted application were not assessed in this run. No gains in revenue, reduction in missed appointments or health outcomes are inferred.

What you can check in your evaluation.

For reception, changing the time and confirming the new appointment are separate steps. When assessing the system with fictional information, check the new time in the appointment calendar and see whether the team can distinguish an appointment that is still awaiting confirmation.

A notification prepared for sending is also different from a received message. Before relying on that channel, test sending, receiving and failure handling separately, using a contact controlled by the team.

The cancellations and rescheduling guide helps design the clinic's process. This case shows a specific check of the rules implemented in one version of our product; the terms of use and plans remain on the commercial page.

The demonstration lets you explore the interface with fictitious data. It is a separate journey from this technical execution.

Wellinton Oliveira

Written by

CEO of Lyntev

Technology with context, original analysis and sources you can consult.