Search

Introduction: Precise prediction of hospital stay duration is essential for maximizing resource utilization during surgery. Existing lumbar spinal stenosis (LSS) surgery prediction models lack accuracy and generalizability. Machine learning can improve accuracy by considering preoperative factors. This study aimed to develop and validate a machine learning-based model for estimating hospital stay duration following decompression surgery for LSS. Methods: Data from 848 patients who underwent decompression surgery for LSS at three hospitals were examined. Twelve prediction models, using 79 preoperative variables, were developed for postoperative hospital stay estimation. The top five models were chosen. Fourteen models predicted prolonged hospital stay (14 days), and the most accurate model was chosen. Models were validated using a randomly divided training sample (70%) and testing cohort (30%). Results: The top five models showed moderate linear correlations (0.576-0.624) between predicted and measured values in the testing sample. The ensemble of these models had moderate prediction accuracy for final length of stay (linear correlation 0.626, absolute mean error 2.26 days, standard deviation 3.45 days). The c5.0 decision tree model was the top predictor for prolonged hospital stay, with accuracies of 89.63% (training) and 87.2% (testing). Key predictors for longer stay included JOABPEQ social life domain, facility, history of vertebral fracture, diagnosis, and Visual Analogue Scale (VAS) of low back pain. Conclusions: A machine learning-based model was developed to predict postoperative hospital stay after LSS decompression surgery, using data from multiple hospital settings. Numerical prediction of length of stay was not very accurate, although favorable prediction of prolonged stay was accomplished using preoperative factors. The JOABPEQ social life domain score was the most important predictor.

Introduction: The efficacy of minimally invasive surgeries for thoracolumbar flexion-distraction injuries (FDIs) has been reported, but those surgeries were monosegmental fusion surgeries of two adjacent vertebrae with bone grafts or temporary fixations using percutaneous pedicle screws (PPSs) that were at least bisegmental. Our idea was to fuse the fracture itself, not to fuse the fractured vertebra with an adjacent vertebra or to stabilize the fractured vertebra by bridging rostrally/caudally adjacent intact vertebrae, specifically when the displacement is minimal. This study aimed to present the surgical techniques of reduction and temporary monosegmental fixation of neurologically intact thoracolumbar bony FDIs using multiaxial PPSs, which can minimize the surgical invasiveness and preserve all motion segments, as well as report three cases treated with this procedure. Technical Note: When the fracture extended from the vertebral body to the spinous process at the same level, screws were placed into the fractured vertebra rostrally to the fracture along the rostral endplate, and the caudally adjacent vertebra was instrumented beyond the fracture line. When the fracture extended from the vertebral body to the spinous process of the rostrally adjacent vertebra, screws were placed into the fractured vertebra caudally to the fracture line, and the rostrally adjacent vertebra was instrumented. The kyphotic deformity was reduced through ligamentotaxis by using MPPSs in the rostral vertebra as rigid joysticks to apply direct buttress leverage to the rostral endplate. Intraoperative blood loss was minimal. The correction of kyphotic deformity and its durability were acceptable, and the segmental range of motion of the two affected vertebrae from flexion to extension was maintained after implant removal. Conclusions: This surgery can act as the least-invasive option for the management of thoracolumbar bony FDIs to allow early ambulation without external bracing and to preserve all the motion segments.