Return to Work After Lumbar Microdiscectomy - Personalizing Approach Through Predictive Modeling
Monika Papić1, Sanja Brdar2, Vladimir Papić3
1Department for occupational health, Health center Novi Sad, Serbia.
Studies in Health Technology and Informatics
|May 27, 2016
Summary
Predicting return to work after lumbar disc herniation (LDH) surgery is crucial. Machine learning models identified psychosocial factors, spine mobility, and job demands as key predictors for prolonged work absence.
Area of Science:
- Neurosurgery
- Occupational Medicine
- Data Science
Background:
- Lumbar disc herniation (LDH) is a leading cause of surgical intervention in the working population.
- Predicting return to work (RTW) post-surgery is vital for patient outcomes and economic impact.
Purpose of the Study:
- To develop and evaluate machine learning models for predicting RTW after operative treatment for LDH.
- To identify key risk factors associated with prolonged work absence.
Main Methods:
- An observational study of 153 patients undergoing operative treatment for LDH.
- Classification algorithms including decision trees (DT), support vector machines (SVM), and multilayer perception (MLP) were employed.
- The RELIEF algorithm was used for feature selection.
Main Results:
- Multilayer perception (MLP) achieved the highest recall (0.86) for predicting patients who did not return to work.
- Key predictors for prolonged work absence included psychosocial factors, spinal mobility, facet joint changes, and occupational demands (standing, sitting, microclimate).
Conclusions:
- MLP models, combined with selected features, can facilitate early identification of patients at risk of prolonged disability.
- Personalized interventions can be developed for individuals at high risk of delayed return to work after LDH surgery.
