Centro Médico
A connected clinical experience from first appointment to follow-up.
Executive summary
Reception needs actual availability for every clinician and location. Clinical staff need relevant records, tests and consent. Patients need a clear route before and after a visit. The proposed platform joins those journeys in a modular product with visible ownership, status and traceable records.
The design begins with scheduling and records, then extends the same identities to reminders, payments and reports. Scheduling models duration, preparation and shared resources. Clinical records retain authorship and versions. Confirmation messages follow explicit events and patient channel consent.
Rollout is phased: map workflows and data, pilot one location and specialty, train staff, verify integration, then expand. Clinical and administrative leads define every integration. No-show, handling-time and data-quality baselines are measured before improvements are reported.
The challenge
Disconnected systems create duplicate patients, booking collisions and difficult reconciliation. An apparent free slot may still require an occupied room or shared equipment. Baselines must record cancellations, rescheduling, identity errors and administration time before targets are set.
Architecture
Patient, reception and clinical interfaces share a persistent identity and role-aware API. PostgreSQL handles transactions, PostGIS supports site search and medical history is versioned. Concurrent reservations need transactional checks. Laboratory, messaging and billing connectors use idempotent event identifiers to prevent duplicate bookings or charges.
- Headless WordPress as the content layer, with a Next.js 14 and TypeScript frontend for patients and clinical staff.
- PostgreSQL 16 with PostGIS to geolocate sites and route patients to the nearest one.
- OAuth2 and JWT with role-based access control for clinicians, front desk, administration and management.
- Multi-site scheduling with optimistic locking that prevents double booking, even with several front desks working in parallel.
- Electronic clinical records with digital signature, versioned history and a full audit trail for every change.
- Reminders through the WhatsApp Business API, email and SMS as a fallback.
- Billing integrated with Chile's SII and Argentina's AFIP.
- Infrastructure on Google Cloud Run, Cloud SQL, Cloud Storage and Cloud CDN, with AES-256 encryption at rest, TLS 1.3 in transit and backups every six hours.
- Observability through Sentry, Cloud Logging and Prometheus to catch issues before a patient ever notices them.
Six modules
Six modules cover multi-site booking; admission and consent; versioned medical records; reminders and confirmations; service billing and SII connection; and an operations dashboard for capacity, waiting lists and referrals. Each module can be piloted independently with explicit read and write rights.
Engineering decisions
Booking consistency takes precedence over an optimistic display. Reservations must commit before notification; messages can be retried later. Records consist of attributable entries. Legacy systems use adapters rather than dictating patient-facing design. Failure of an external provider should leave a visible reconciliation queue.
Security
Clinical data requires separate clinician, reception, billing and management permissions. Sensitive roles use multifactor authentication. Every clinical change and download is audited. Encryption, managed secrets, tested restores, retention periods and local privacy review form part of deployment.
Performance
Performance checks distinguish initial load, patient search and reservation confirmation. Long histories are paginated, common searches indexed and large images loaded when requested. Booking queries must be load tested against expected location and staff volumes; thresholds are agreed during the pilot.
Observability
Sentry tracks application errors without capturing sensitive clinical notes. Grafana monitors latency, error rate and connector queues. Correlation identifiers trace a booking to confirmation. Alerts distinguish delayed invoicing from a critical double booking; backup restores and reminder delivery are monitored.
Results and measurement
The before-and-after table describes workflows, not verified statistical outcomes. A pilot should measure no-shows by specialty, administrative minutes, reconciliation errors, duplicate records and patient feedback against a comparable baseline. Every reported figure needs dates, sample and limitations.
| Before | After | |
|---|---|---|
| Scheduling and records | Separate systems, no traceability | Unified, with versioned history |
| Reminders | Manual, over WhatsApp | Automatic via WhatsApp, email and SMS |
| Billing | In a separate system | Integrated with SII / AFIP |
Use cases
Dental clinics coordinate chair, clinician and imaging. Specialty centres support referrals with context. Multi-site networks allow rescheduling with instructions intact. Physiotherapy manages session cycles. Labs deliver results to the record. Telemedicine links consultation and note; billing reconciles services and management sees aggregate demand.
Roadmap Q1 / Q2 / Q3 2026
Q1 2026: workflow discovery, data inventory and concurrent booking prototype. Q2: focused pilot for records, consent, reminders and priority connectors with training and security review. Q3: compare indicators, correct friction and expand by site. Dates are a reference plan subject to approvals.
Frequently asked questions
Does the record replace clinical judgement?
No. It organises access and records. Licensed clinicians retain responsibility for diagnoses and treatment.
How are old records migrated?
Inventory, field mapping and sample validation precede migration. Unstructured files may remain identified attachments.
What if a location loses connectivity?
A continuity policy defines limited access, temporary safeguards and subsequent reconciliation.
Can current systems coexist?
Yes. Adapters support migration by specialty or location and log integration errors.
How is double booking prevented?
Transactional confirmation checks both clinician and room before showing a successful reservation.
Who may see a record?
Rights follow clinical need and job function. Sensitive access is auditable.
How is impact measured?
Define indicators and baselines before the pilot and compare like periods and populations.
What starts the project?
Map locations, current scheduling, connected systems, clinical owners and privacy requirements.
Let’s talk
Let us review sites, specialties and integrations, then define a first pilot, owners and measurable criteria for the next stage.