2501. Optimizing Post-Discharge Follow-Up Using Operational Research and AI-Driven Risk Stratification
Invited abstract in session TD-13: Machine learning in healthcare, stream OR in Healthcare (ORAHS).
Tuesday, 14:30-16:00Room: Clarendon SR 1.01
Authors (first author is the speaker)
| 1. | Matilde Santos
|
| UNIDEMI, NOVA FCT | |
| 2. | Mariana Peyroteo
|
| UNIDEMI, Department of Mechanical and Industrial Engineering, NOVA School of Science and Technology, Universidade NOVA de Lisboa | |
| 3. | Luís Lapão
|
| Intelligent Decision Support Systems Laboratory, Universidade Nova de Lisboa |
Abstract
Hospital readmissions pose significant operational challenge, demanding efficient resource allocation and smarter decision-making in healthcare organization. This study applies Operational Research (OR) methodologies to optimize post-discharge follow-up through iFollow@Care digital platform - a decision-support system integrating machine learning-based risk stratification with workflow optimization. The digital platform was implemented with streamlit and python libraries. Using three years of patients’ follow-up data, we developed a classification model (Logistic Regression, AUROC: 0.9531) to dynamically prioritize patient follow-up needs, reducing unnecessary interventions while ensuring timely care for high-risk cases. The application of predictive analytics, and resource optimization enabled the reduction of nurse workload by at least 79%, reallocating time toward more value-added patient care. Additionally, evidence-based validation demonstrated its potential to enhance patient outcomes and hospital efficiency. This work highlights the role of AI and OR in data-driven healthcare decision-making, offering a scalable framework for real-time risk management and proper resource allocation in hospital settings.
Keywords
- Machine Learning
- Medical Applications
- Risk Analysis and Management
Status: accepted
Back to the list of papers