This study aimed to establish a predictive nomogram model for recollapse of fractured vertebra after posterior pedicle screw fixation in thoracolumbar fractures (TLFs). Patients undergoing posterior pedicle screw fixation for TLFs at our hospital between January 2016 and December 2021 were retrospectively reviewed. Patients were divided into 2 groups according to the presence or absence of recollapse of the fractured vertebra at the final follow-up. The predictors for fractured vertebra recollapse were identified by univariate and multivariable logistic regression analysis, and a nomogram model was developed. The prediction performance and internal validation were established. A total of 224 patients were included in this study. Of these, 46 (20.5%) patients developed recollapse of fractured vertebra. Age, thoracic and lumbar injury severity score, screw distribution in the fractured vertebra, and anterior vertebral height compression ratio were associated with vertebral recollapse. These predictors were used to construct a predictive nomogram. The area under the receiver operating characteristic curve of the nomogram model was 0.891. The concordance index was 0.891, and it was 0.877 with bootstrapping validation. The calibration curves and decision curve analysis also suggested that the nomogram model had excellent predictive performances for fractured vertebra recollapse. A clinical nomogram incorporating 4 variables was constructed to predict fractured vertebra recollapse after posterior pedicle screw fixation for TLFs. The nomogram demonstrated good calibration and discriminative abilities, which may help clinicians to make better treatment decisions.
Read full abstract