2025 - 2027 PARSLAI: AI Analysis of Care Pathways in ALS Patients in the Centre-Val de Loire and in Pays de la Loire Region: Towards Continuity of Care optimization
ARSLA - Partners: LIFAT and CHRU Tours.
ARSLA
Amyotrophic lateral sclerosis (ALS) is a devastating neurodegenerative disease that causes gradual muscle weakness, leading to severe disability and ultimately death. Managing ALS is complex because patients require care from multiple healthcare providers, including neurologists, respiratory therapists, dietitians, and physical therapists. Coordinated multidisciplinary care has been shown to improve outcomes by slowing functional decline, reducing hospitalizations, and enhancing patients’ quality of life. However, access to coordinated care can differ widely depending on geographical location and availability of specialized resources.
This project aims to evaluate and compare the care pathways for patients with ALS in two French regions characterized by medical desert:
- Centre-Val de Loire (CVL), where the ALS reference center is located in Tours. The region benefits from an established neurocentre network, ensuring seamless coordination between hospital and home care.
- The eastern half of Pays-de-la-Loire (PDL), with its ALS reference center in Angers (where the PRIOR platform is located and where there is a mobile team dedicated to rare neurological diseases). Unlike CVL, this area lacks a structured network but relies on a mobile team specialized in rare neurological diseases.
We hypothesize that patients living in the CVL and PDL regions experience variations in the type and quality of care they receive, particularly regarding regularity, continuity, and care coordination, including potential interruptions and delays along their care pathways. Understanding these differences is essential to identify gaps and implement improvements, while also considering the specific challenges posed by medically underserved areas. By leveraging comprehensive administrative and clinical data from the clinical data warehouses of both participating centers—including hospital records and detailed patient measurements—we will analyze care trajectories over time. Our analysis will track key events such as clinic visits, hospitalizations, procedures (e.g., gastrostomy or mechanical ventilation), and instances of loss to follow-up. Using advanced statistical models and machine learning, we will identify subtle patterns, detect delays in care, and assess how disruptions relate to clinical deterioration, including rapid functional decline or respiratory failure.
Through process mining techniques, we will visualize patient pathways to identify bottlenecks and care ruptures that may increase risks for poor outcomes. Predictive algorithms will be developed to forecast which patients are most at risk of care disruption or emergency hospitalization, enabling early intervention. By comparing the two regions, the project aims to assess the impact of coordinated neurocenter networks and mobile team on the continuity and quality of care for ALS patients.
Ultimately, this research will generate valuable insights for healthcare providers and policymakers. It will support the design of targeted strategies to reinforce care coordination, reduce preventable hospital admissions, and enhance patient quality of life. Furthermore, the methods and tools developed may be extended to other chronic diseases requiring complex, multidisciplinary care. This project represents a major step towards personalized, patient-centered care pathways for ALS in medically underserved regions and contributes to the broader field of digital health innovation.