Related Experiment Video
Updated: Jan 17, 2026
11:30
Multilevel Oblique Lumbar Interbody Fusion in Degenerative Lumbar Disc Disease with Instability
Published on: July 25, 2025
1.1K
Predicting Surgical Site Infection after Lumbar Laminectomy and Discectomy: A Cutting-edge Algorithmic Approach by
Ali Haider Bangash1, Kyle Mani2, Samuel N Goldman2
1Department of Neurosurgery, Montefiore Medical Center, Albert Einstein College of Medicine, 3316 Rochambeau Ave, Bronx, NY, 10467, USA.
Neurosurgical Review
|September 18, 2025
Summary
This study developed an advanced machine learning model to predict surgical site infections (SSIs) after spine surgery. The novel approach achieved high accuracy, aiding in better patient outcomes for degenerative spinal disease.
Area of Science:
- Spine Surgery
- Machine Learning
- Infectious Disease Prevention
Background:
- Surgical site infections (SSIs) are a significant complication following lumbar laminectomy and discectomy for adult degenerative spinal disease (DSD).
- Accurate prediction of SSIs is crucial for effective patient management and resource allocation.
Purpose of the Study:
- To develop and validate an algorithmic approach for predicting SSIs in patients undergoing lumbar laminectomy and discectomy.
- To incorporate ensembled stacking into state-of-the-art automated machine learning (aML) for enhanced predictive accuracy.
Main Methods:
- Utilized a prospective multicenter surveillance dataset of SSIs after lumbar laminectomy and discectomy.
- Employed nine algorithms (XGBoost, LGBM, NN, CatBoost, RF) with hyperparameter tuning within an aML framework.
- Implemented ensembled stacking, combining stacked and ensemble models, evaluated using five-fold cross-validation and macro-weighted average Area Under the Receiver Operating Curve (mWA-AUROC).
Main Results:
- A stacked ensemble model achieved an mWA-AUROC of 0.994, accuracy of 98.7%, sensitivity of 90%, and specificity of 98.81% in predicting SSIs.
- Key predictors identified by the top model (XGBoost-20) included operative time, smoking status, and patient age.
- The algorithmic model architecture is available on GitHub for external validation.
Conclusions:
- The novel algorithmic approach integrating ensembled stacking into aML effectively predicts SSIs after lumbar laminectomy and discectomy for DSD.
- The high performance of the stacked ensemble model suggests its potential as a valuable clinical tool for improving decision-making and patient outcomes in spine surgery.
- Further validation in diverse settings and integration into clinical practice are recommended.
