2023 ISAKOS Biennial Congress ePoster
Predictors of Complete Peroneal Nerve Palsy Recovery Following Multi-Ligamentous Knee Injury Using Machine Learning: A Multi-Center Retrospective Cohort Study
Kinjal Vasavada, BA UNITED STATES
Dhruv S Shankar, BS, New York UNITED STATES
Andrew S Bi, MD, New York, NY UNITED STATES
Jay Moran, BS, New Haven UNITED STATES
Massimo Petrera, MD, FRCSC, New York, NY UNITED STATES
Joseph B Kahan, MD, New Haven UNITED STATES
Erin Alaia, MD, New York, New York UNITED STATES
Michael J. Medvecky, MD UNITED STATES
Michael J Alaia, MD, New York, New York UNITED STATES
NYU Langone Health, New York, NY, UNITED STATES
FDA Status Not Applicable
Summary
Among multiligament knee injury patinets, a random forest (RF) classifier algorithm was used to identify demographic, injury, treatment, and postoperative variables that were significant predictors of recovery from complete peroneal nerve palsy.
ePosters will be available shortly before Congress
Abstract
Introduction
Peroneal nerve (PN) palsy is one of the most debilitating sequelae of multi-ligamentous knee injuries (MLKIs). There is limited literature regarding recovery from complete PN palsy and reliable predictors for recovery.
Methods
We conducted a retrospective review of patients seen at two urban hospital systems for treatment of MLKI with associated complete PN palsy, defined as the presence of complete foot drop with or without sensory deficits on physical exam. Recovery was defined as the complete resolution of foot drop. A random forest (RF) classifier algorithm was used to identify demographic, injury, treatment, and postoperative variables that were significant predictors of recovery from complete PN palsy. Validity of the RF model was assessed using overall accuracy, F1 score, and area under the receiver operating characteristic curve (AUROC).
Results
16 MLKI patients with associated complete PN palsy were included in the cohort. Among them, 75% (12/16) had documented knee dislocation requiring reduction. Complete recovery occurred in 4 patients (25%). On MRI, nerve contusions were more common among patients without PN recovery (p<0.05), but there were no other significant differences between recovery and non-recovery groups. The RF model found that increasing age, BMI, and male sex were predictive of worse likelihood of PN recovery. Furthermore, the model was found to have good validity with a classification accuracy of 75%, F1 score of 0.86, and AUROC of 0.64.
Conclusion
MLKI with complete PN palsy remains a challenging injury to treat and continues to have life-altering impacts on patients. Our RF model found that increasing age, BMI, and male sex were predictive of decreased likelihood of nerve recovery. While further study of machine learning models with larger patient datasets is required to identify the most superior model, these findings present an opportunity for orthopedic surgeons to better identify, counsel, and treat patients with MLKIs and concomitant complete PN palsy.