2052. Towards personalised pathway management for acquired brain injury paediatric rehabilitation via Markov models
Invited abstract in session MB-13: Medical services and applications, stream OR in Healthcare (ORAHS).
Monday, 10:30-12:00Room: Clarendon SR 1.01
Authors (first author is the speaker)
| 1. | Bruno Salezze Vieira
|
| Decision Analytics and Risk, University of Southampton | |
| 2. | Edilson Arruda
|
| Department of Decision Analytics and Risk, University of Southampton |
Abstract
Acquired Brain Injury (ABI) is a leading cause of disability in pediatric patients, requiring personalized rehabilitation approaches. This study introduces a Markov model-based framework for managing ABI rehabilitation pathways, enabling dynamic, patient-specific predictions. Unlike traditional models, which assume static recovery trajectories, our approach adapts to new patient data over time, capturing inter- and intra-patient variability.
Using anonymized data from the Bristol Royal Hospital for Children (2016–2018), we analyzed 575 weekly rehabilitation records from 53 patients. Multinomial logistic regression was employed to estimate transition probabilities between Kings Outcome Scale for Child Head Injury (KOSCHI) states. The model achieved 93% accuracy and a log-likelihood of 0.84, demonstrating strong predictive performance.
By offering a probabilistic view of recovery, our approach supports clinicians in making informed decisions, adjusting treatment plans dynamically, and improving patient outcomes. This study advances personalized pathway management through data-driven, adaptive modeling.
Keywords
- Health Care
- OR in Sustainability
- Analytic Hierarchy Process
Status: accepted
Back to the list of papers