The accurate prediction of neurological outcomes in patients with cervical spinal cord injury (SCI) is difficult because of heterogeneity in patient characteristics, treatment strategies, and radiographic findings. Although machine learning algorithms may increase the accuracy of outcome predictions in various fields, limited information is available on their efficacy in the management of SCI. We analyzed data from 165 patients with cervical SCI, and extracted important factors for predicting prognoses. Extreme gradient boosting (XGBoost) as a machine learning model was applied to assess the reliability of a machine learning algorithm to predict neurological outcomes compared with that of conventional methodology, such as a logistic regression or decision tree. We used regularly obtainable data as predictors, such as demographics, magnetic resonance variables, and treatment strategies. Predictive tools, including XGBoost, a logistic regression, and a decision tree, were applied to predict neurological improvements in the functional motor status (ASIA [American Spinal Injury Association] Impairment Scale [AIS] D and E) 6 months after injury. We evaluated predictive performance, including accuracy and the area under the receiver operating characteristic curve (AUC). Regarding predictions of neurological improvements in patients with cervical SCI, XGBoost had the highest accuracy (81.1%), followed by the logistic regression (80.6%) and the decision tree (78.8%). Regarding AUC, the logistic regression showed 0.877, followed by XGBoost (0.867) and the decision tree (0.753). XGBoost reliably predicted neurological alterations in patients with cervical SCI. The utilization of predictive machine learning algorithms may enhance personalized management choices through pre-treatment categorization of patients.

BACKGROUND: The risk of recurrence is overestimated by the Kaplan-Meier method when competing events, such as death without recurrence, are present. Such overestimation can be avoided by using the Aalen-Johansen method, which is a direct extension of Kaplan-Meier that accounts for competing events. Meningiomas commonly occur in older individuals and have slow-growing properties, thereby warranting competing risk analysis. The extent to which competing events are considered in meningioma literature is unknown, and the consequences of using incorrect methodologies in meningioma recurrence risk analysis have not been investigated. METHODS: We surveyed articles indexed on PubMed since 2020 to assess the usage of competing risk analysis in recent meningioma literature. To compare recurrence risk estimates obtained through Kaplan-Meier and Aalen-Johansen methods, we applied our international database comprising ~ 8,000 patients with a primary meningioma collected from 42 institutions. RESULTS: Of 513 articles, 169 were eligible for full-text screening. There were 6,537 eligible cases from our PERNS database. The discrepancy between the results obtained by Kaplan-Meier and Aalen-Johansen was negligible among low-grade lesions and younger individuals. The discrepancy increased substantially in the patient groups associated with higher rates of competing events (older patients with high-grade lesions). CONCLUSION: The importance of considering competing events in recurrence risk analysis is poorly recognized as only 6% of the studies we surveyed employed Aalen-Johansen analyses. Consequently, most of the previous literature has overestimated the risk of recurrence. The overestimation was negligible for studies involving low-grade lesions in younger individuals; however, overestimation might have been substantial for studies on high-grade lesions.