Machine learning insights into patient satisfaction following lateral lumbar interbody fusion.

purposeLateral Lumbar Interbody Fusion (LLIF) has become a minimally invasive procedure for treating degenerative lumbar conditions. While it offers reduced blood loss and faster recovery, patient satisfaction following LLIF surgery shows significant variability. Identifying the factors influencing satisfaction is crucial for optimizing surgical outcomes and improving patient care. This study aims to determine key factors affecting patient satisfaction after LLIF surgery using machine learning (ML) models, including Random Forest, Logistic Regression, Support Vector Machine (SVM), and k-Nearest Neighbors (k-NN). Additionally, the study evaluates the predictive performance of these models to identify the most influential factors contributing to postoperative satisfaction.

methodsA retrospective analysis was conducted on 149 patients who underwent LLIF surgery. Preoperative, intraoperative, and postoperative variables were collected, including patient demographics, clinical measures, surgical details, and functional outcomes. Based on these variables, ML models were used to predict patient satisfaction vs. not satisfied) satisfaction (s. Model performance was evaluated using fivefold cross-validation, with metrics including accuracy, precision, recall, and F1 score.

resultsOf the 149 patients, 85.2% reported satisfaction with the surgical outcome. Random Forest achieved the highest predictive performance, with an average accuracy of 82.6%, precision of 83.6%, recall of 99.2%, and an F1 score of 90.7%. Key factors influencing patient satisfaction included the preoperative low back pain score, social life function, and postoperative improvements in walking ability and mental health. Surgical factors, such as the number of fused segments and endplate injury, had less influence on satisfaction.

conclusionsFunctional outcomes, particularly improvements in low back pain, walking ability, and mental health, are the primary determinants of patient satisfaction following LLIF surgery. In contrast, surgical factors play a less significant role. Mental health emerged as a critical factor, underscoring the importance of addressing psychological recovery through preoperative counseling and personalized postoperative care. The analysis demonstrated that ML models, especially Random Forest, are effective tools for identifying the factors most predictive of postoperative satisfaction. These findings highlight the potential of ML techniques to enhance personalized treatment planning and improve outcomes by focusing on both physical and mental recovery. Further research, including multi-center studies and the integration of psychological variables, is needed to provide a more comprehensive understanding of patient satisfaction after LLIF surgery.