Machine Learning to Predict Factors Affecting Rehabilitation Length of Stay and Healthcare Costs for Neurological Rehabilitation

Machine Learning Predictive Analysis of Key Factors Influencing Rehabilitation Length of Stay (RLOS) and Direct Hospitalization Costs for Neurological Inpatient Rehabilitation at Tertiary Care Hospital

The aim of this retrospective study is to ascertain total direct costs, rehabilitation length of stay (RLOS) and factors associated with RLOS for neurological inpatient rehabilitation at the tertiary care hospital.

The aim of the study is to identify factors that influence RLOS and the correlated costs for neurological rehabilitation in tertiary rehab using data extracted from EPIC. It is also aimed to identify the median direct costs to find out the main contributors to the costs in the local population. Lastly, the study aims to utilise artificial intelligence or machine learning to analyse the compiled data to develop a predictive model. The model aspires to understand factors associated with extended RLOS and to predict RLOS of patients who require neurological rehabilitation, aiding preemptive measures.

Inclusion Criteria:

• All patients who completed inpatient rehabilitation with the index conditions in their discharge summaries

Exclusion Criteria:

• Did not complete inpatient rehabilitation as they are discharged against medical advice