City-to-Hospital Patient Record: Seamless Digital Care Pathway 2026

The essentials in 30 seconds

Introduction

When a patient leaves the hospital and returns to their community physician, the continuity of care often collapses. The discharge summary—known in France as the « lettre de liaison »—remains a paper‑centric artifact, frequently delayed, incomplete, or unreadable. This break in the care pathway fuels readmissions, medication errors, and costly follow‑up calls.

Since 2016, Galeon has co‑created an AI‑enhanced electronic health record (EHR) with clinicians, now operating in 19 hospitals—including two university medical centers—covering more than three million patient records and supporting over 10,000 caregivers. "A coordinated digital care pathway begins with data that moves as fast as the patient does," notes Dr. Léa Martin, Head of Clinical Informatics at a participating CHU.

The challenge is not only technical but regulatory: French health law (Article R. 6317‑1) mandates a discharge summary, while the European AI Act imposes strict transparency for clinical AI. A truly seamless pathway must therefore satisfy legal obligations, clinical accuracy, and real‑time data exchange.

In this pillar article we examine where the pathway breaks today, how the discharge letter functions, what the patient health record (DMP) can and cannot do, the role of an intelligent EHR in automating documentation, and finally what the city‑based physician actually receives—and how they can turn it into actionable care.

Where does the care pathway break today?

It breaks at the hand‑off between hospital and community care. Manual transcription, incompatible file formats, and delayed delivery create a gap that compromises patient safety.

Hospitals still rely on PDF or paper copies that must be faxed or mailed. According to the French Ministry of Health, 27 % of discharge summaries arrive late, and 19 % contain missing critical information such as medication changes (source: e‑Santé 2025 report).

For Hospital CIOs

The lack of interoperable standards forces IT teams to maintain costly custom interfaces, while clinicians spend up to 30 % of their time on administrative tasks (CNIL, 2024).

What is the discharge summary (letter of liaison) and why is it mandatory?

It is a legally required, structured document that communicates the key events of a hospital stay. French law (Article R. 6317‑1) obliges hospitals to provide a concise summary within 48 hours of discharge.

In practice, many hospitals produce free‑text PDFs that fail to map to the national DMP schema, making downstream processing by primary‑care EHRs difficult.

For Primary‑Care Physicians

Without a machine‑readable format, family doctors must manually extract dosage changes, follow‑up appointments, and test results—often resulting in omissions.

What does the patient health record (DMP) solve and what does it not solve?

The DMP offers a patient‑centric, longitudinal view of health data collected across care settings. It standardises data structures, enabling secure sharing under French GDPR provisions.

However, the DMP does not force hospitals to push real‑time updates; it merely provides a repository where data can be deposited. The responsibility to generate and transmit structured discharge data still lies with the hospital’s EHR.

How does an intelligent EHR automate discharge documentation?

An AI‑enhanced EHR can extract, validate, and populate discharge fields automatically. Galeon’s DPI leverages natural‑language processing (NLP) trained on clinician‑validated datasets to produce a structured liaison letter in under two minutes.

The system cross‑checks medication lists against the French medication database (ANSM) and flags inconsistencies before final sign‑off, reducing human error by 42 % (internal audit 2025).

Technical note

Swarm Learning® – a decentralized AI training method – lets each hospital improve the model without moving patient data, preserving sovereignty while complying with the HDS certification (2024) and ISO 27001:2022 alignment.

What does the community physician actually receive and how can they use it?

They receive a structured, interoperable summary linked to the patient’s DMP. The document arrives via secure API in HL7 FHIR format, ready to be ingested into the primary‑care EHR.

This enables the GP to trigger automated alerts for contraindicated drugs, schedule follow‑up appointments directly from the summary, and document the encounter without re‑typing information.

Traditional vs Galeon‑enabled approach

Limits and challenges to be aware of

• Change management requires sustained training. Even the smartest EHR cannot compensate for users who bypass automation.
• Regulatory nuance. The AI Act classifies some clinical decision‑support tools as high‑risk; continuous documentation tools remain lower‑risk but must maintain transparency.
• Inter‑hospital variability. While Swarm Learning protects data sovereignty, differing IT stacks can affect model convergence speed.
• Integration effort. Legacy EHRs without FHIR endpoints need adapters, adding short‑term implementation cost.
• Data quality dependency. AI performance is only as good as the clinician‑validated data it learns from (see our Predictive medicine & data quality article).

FAQ

How quickly can a discharge summary be generated?
The AI‑enhanced DPI can produce a complete, structured summary in under two minutes, compared with 15‑30 minutes manually.

Is the structured summary legally recognised?
Yes. French law accepts electronic, standards‑based documents as long as they contain the mandatory content and are signed by the responsible physician.

Do community doctors need new software to read the summary?
Most modern primary‑care EHRs support HL7 FHIR; if not, a simple web viewer can be used without additional licensing.

How does Swarm Learning protect patient privacy?
Models are trained locally on each hospital’s server; only model parameters (not raw patient data) are shared, preserving GDPR compliance.

What is the ROI for hospitals?
Pilot programs report a 20 % reduction in administrative staff time and a 4 % drop in 30‑day readmissions, translating to cost savings of €1.2 M per 100 beds annually.

In summary

A coordinated digital care pathway hinges on transferring the patient record seamlessly from city to hospital and back. Galeon’s intelligent EHR, built together with clinicians since 2016, automates the creation of a standards‑based discharge summary, pushes it directly to the patient’s DMP, and delivers a machine‑readable document to the community physician. By eliminating manual transcription, ensuring legal compliance, and leveraging Swarm Learning for decentralized AI, hospitals can close the most critical gap in continuity of care while respecting data sovereignty and GDPR. The result is fewer readmissions, lower administrative burden, and a true, coordinated digital care pathway for every patient.

Explore how Smart EHRs outperform traditional systems

Sources

• ANSM medication database
• French Ministry of Health – Discharge Summary Statistics 2025
• EUR‑Lex – European AI Act (2024)
• CNIL – Health Data Governance 2024
• HDS certification guidelines 2024
• Galeon – Predictive medicine & data quality
• Galeon – Smart EHR vs Traditional EHR

‍