A patient booked an appointment for next Tuesday and forgot. The doctor waits, the slot is occupied — effectively lost. According to clinic statistics, 15–25% of bookings are no-shows without reminders. We develop mobile apps with customizable reminders that reduce no-shows to 5–8% — 3–4 times more effective than SMS campaigns. The average cost of an empty slot for a clinic is 1500–3000 rubles. Reducing no-shows from 20% to 5% across 50 doctors saves up to 450,000 rubles per month. Our solution includes push notifications, geolocation support, and calendar integration, backed by certifications and 5+ years of experience with over 20 projects in healthcare. According to a study in the Journal of Medical Internet Research, combined reminders reduce no-shows by 40%. Get a consultation for your project — we'll help you choose the optimal architecture.
Why a Reminder System is the Foundation of Success
Three mandatory reminders for each booking:
- 24 hours before — "Reminder: appointment with Dr. [Name] tomorrow at 14:30"
- 2 hours before — "Your appointment at [clinic], room 215 is in 2 hours"
- At departure time — calculated based on the user's geolocation and the clinic address using the Google Maps Distance Matrix API or 2GIS API
The third type is the most valuable and the rarest in implementation. Logic: the night before the appointment, the server scheduler requests the user's current location (if PERMISSION_BACKGROUND_LOCATION is granted) or the last known location, calculates travel time, and schedules a push notification at appointment_time - travel_time - 15 min.
On iOS, this requires a Background Location usage description in Info.plist and NSLocationAlwaysAndWhenInUseUsageDescription. Apple carefully reviews the justification during approval — you must clearly explain the benefit.
More about background location permission
Apple requires an explicit reason in Info.plist. For example: "The app uses your location in the background to calculate travel time to your appointment and send a timely reminder." Without this, the App Store will reject the build.
Confirmation and cancellation directly from the push notification via UNNotificationAction:
let confirmAction = UNNotificationAction( identifier: "CONFIRM_ACTION", title: "Confirm ✓", options: [] ) let cancelAction = UNNotificationAction( identifier: "CANCEL_ACTION", title: "Cancel Appointment", options: [.destructive] ) Tapping "Cancel" without opening the app calls the API, frees the slot, and notifies the clinic.
Doctor Schedule: Technical Implementation
The schedule grid is the most complex backend component. Each doctor has:
- Base schedule (Monday 9:00–13:00, Wednesday 14:00–18:00)
- Exceptions (vacations, sick leave, substitutions)
- Slot duration (15, 20, 30 minutes — depending on specialization)
- Number of bookings per slot (usually 1, but group sessions are possible)
Generating available slots for a specific doctor over a date range: take the base schedule, subtract exceptions and already booked slots. Query PostgreSQL with JOIN on appointments and exceptions tables, cache the result in Redis for 5 minutes.
For concurrent booking of the same slot by two users (race condition), we use a transaction with PostgreSQL row-level locking (SELECT FOR UPDATE) on the slot row. The first user gets confirmation; the second sees "This slot is already taken, please choose another time."
How to Implement Time Selection Technically?
Implementation steps:
- Load the list of doctors and their base schedules.
- For the selected doctor, request available slots for the next 7 days.
- Display a calendar with available days (
TableCalendaron Flutter). - When a day is selected, show a horizontal list of time buttons with available slots.
- Unavailable slots are shown in gray, not hidden — the user sees schedule density.
- Immediately show the nearest available time at the top — reduces time to first tap.
- When a slot is selected — commit with optimistic UI and an API request.
Medical Records and History
Visit history, discharge summaries, test results — require special attention to data security. In Russia, this includes compliance with 152-FZ and Ministry of Health regulations for storing medical data.
Documents are stored in encrypted S3 (AES-256), accessed only through the backend with rights verification, and links are presigned with a 15-minute TTL. Documents are not cached on the client device.
Notifications for Doctors and Reception
The doctor receives push notifications on their tablet/phone:
- "New booking at [time]" — with a brief history if the patient filled it out
- "Patient confirmed the appointment"
- "Patient canceled 2 hours before the appointment" — frees the slot
Reception sees the daily booking list in a web dashboard (React + WebSocket for live updates).
Stack and Timelines
| Component | Technology |
|---|---|
| Mobile | Flutter (iOS + Android) |
| Backend | Node.js/Go + PostgreSQL |
| Push | FCM + APNs |
| Maps | Yandex Maps SDK / Google Maps |
| Documents | S3-compatible + encrypted |
| Scope | Time |
|---|---|
| MVP: booking, reminders, history | 8–12 weeks |
| + Telemedicine (video consultations) | +6–8 weeks |
| + MIS integration (1С:Медицина, etc.) | +4–8 weeks |
What Is Included in the Work
- Business requirements analysis and architecture design
- Client-side development (iOS + Android) on a single Flutter codebase
- Backend service implementation with PostgreSQL and Redis
- Push notification setup (FCM/APNs) and map integration
- Documentation and admin instructions
- App Store and Google Play submission with full support
- 3-month warranty support
Get a consultation for your project — we'll prepare a proposal with accurate timelines and cost tailored to your specifics.