🏥 Tuscany Triage System — International Edition

Version 0.6.2 — Clinical Decision Support & GRC Compliance Framework

🌐 GRC Portfolio Case Study: This repository is part of a broader collection of Governance, Risk & Compliance (GRC) and AI Regulatory Assessment case studies.
👉 Explore the Main GRC Portfolio Hub

---

🧬 Vision & Current Focus

Bridging the Gap Between Care, Code, and Compliance
In high-pressure Emergency Departments, subtle clinical signs can be overlooked.
The Tuscany Triage System isn't meant to replace human intuition, but to empower it.
By integrating clinical tools like the Shock Index and GCS directly into the triage workflow, this system helps identify "silent" risks—like occult shock—before they become overt crises.

🛡️ Strategic Shift: GRC & Regulatory Governance

While the core Python implementation represents a stable baseline, the current focus of this repository is not active software development, but Regulatory Governance, Compliance & Auditability (GRC).

This repository serves as a real-world case study on how medical decision-support software is evaluated, audited, and protected under modern European regulatory frameworks:

• EU AI Act (Regulation UE 2024/1689): Categorized as a High-Risk System (Healthcare / Emergency Decision Support) with Human-in-the-Loop oversight.
• Medical Device Regulation (MDR UE 2017/745): Qualified as Software as a Medical Device (SaMD) Class IIa under Rule 11 (Annex VIII).
• ISO/IEC 27001 & NIS2: Auditability, non-repudiation logging, and healthcare infrastructure resilience.

---

📂 Repository Structure

medical-triage-system/
├── src/                                  # Source Code (Stable Baseline)
│   ├── main.py                           # UI Flow, Operator Login & HITL Override
│   ├── clinic.py                         # Deterministic Clinical Engine & Safety Logic
│   └── utils.py                          # Input Validation & Audit Logging
│
├── governance/                           # 🛡️ GRC & REGULATORY COMPLIANCE ARTIFACTS
│   ├── 01_EU_AI_Act_High_Risk_Assessment.md   # EU AI Act Art. 9, 10, 12, 13, 14 Mapping
│   ├── 02_MDR_Rule11_SaMD_Classification.md   # Rule 11 SaMD Qualification (Class IIa)
│   ├── 03_Clinical_Bias_and_Congruency_Spec.md# Algorithmic Safety & Congruency Engine
│   └── 04_Auditability_and_Logging_Spec.md   # ISO 27001 A.12.4 & Non-Repudiation Spec
│
├── docs/                                 # Executive Audit Deliverables (PDF)
│   ├── Medical_Triage_EU_AI_Act_Compliance_Report.pdf
│   └── SaMD_MDR_Regulatory_Readiness.pdf
│
├── logs/                                 # Sample Audit Trails
│   └── triage_audit_log.txt.example
│
└── README.md                             # Project Dashboard

---

🏗️ Modular Architecture & Clinical Engine (src/)

The software baseline relies on a three-tier modular structure:

• src/main.py (User Interface & Governance): Manages interaction flow, operator login, and mandatory human confirmation (Human-in-the-Loop override under EU AI Act Art. 14).
• src/clinic.py (Clinical Decision Engine): Implements medical logic, GCS calculation, Shock Index analysis, and Symptom-to-Pathway mapping.
• src/utils.py (System Services): Handles input validation, range checks, and audit trail generation (ISO 27001 A.12.4).

🧠 Key Clinical Features

• Clinical Congruency Check: Cross-references reported pain (NRS) with stable vital signs (HR/SBP). If severe pain is reported despite stable vitals, a "Dubious Congruency" alert is raised to prevent automation bias.
• Hemodynamic Safety (Shock Index): Detects occult shock (HR/SBP > 0.9) to prevent under-triage in compensating patients.
• Intelligent Pathway Mapping: Recommends clinical pathways (e.g., Stroke Unit, Cardiac, Trauma) based on symptom keyword analysis.

---

📊 Decision Workflow

graph TD
    A[Operator Login] --> B[Patient Assessment]
    B --> C{Vital Signs Check}
    C -- Critical --> D[Forced RED/ORANGE]
    C -- Stable --> E[Analyze Symptoms]
    E --> F{Assess Pain NRS}
    F --> G{Pain vs Vitals Congruent?}
    G -- No --> H[⚠️ Dubious Congruency Alert]
    G -- Yes --> I[Apply X+1 Upgrade]
    H --> I
    I --> L[Modular Audit Logging]
    L --> M[Admission Record]

Loading

---

🚀 Quick Start & Execution

1. Ensure you have Python 3.8+ installed.
2. Clone this repository.
3. Run the main script from the root directory:

   python src/main.py

⚠️ Medical & Legal Disclaimer

Educational & Portfolio Demonstration Purpose Only
This software is developed for research, educational, and Governance, Risk & Compliance (GRC) demonstration purposes. It is NOT a certified Medical Device and should NOT be used to make actual clinical decisions in real-world healthcare settings. The author assumes no liability for misuse or clinical outcomes resulting from this code.