Identification of the high-risk population facing early death in older patients with primary intracranial glioma: a retrospective cohort study.

Background: This study aimed to establish a diagnostic nomogram to predict the early death risk in older patients with primary intracranial glioma and to identify the high-risk population in those patients to provide them with specialized care to increase their benefit from survival. Methods: Patients aged 60 years and older with histologically confirmed intracranial glioma were identified in the Surveillance, Epidemiology and End Results (SEER) database. Initially, they were divided into a training set and a validation set in a 7:3 ratio. Next, univariate and multivariate logistic regression were employed to identify independent risk variables, which were used to develop a diagnostic nomogram further. Additional analyses were performed on the diagnostic nomogram's performance, including calibration curves, receiver operating characteristic (ROC) curves, and decision curve analysis (DCA). A mortality risk classification system was ultimately developed using the diagnostic nomogram. Results: This study included 8,859 individuals diagnosed with primary intracranial glioma. The participants were randomly split into two groups: a training set consisting of 6203 individuals and a validation set consisting of 2,656 individuals, with a ratio of 7 to 3. Univariate and multivariate logistic regression analyses on early death showed 7 independent risk variables (age, median household income, histological type, tumor grade, surgery, radiation therapy, and systemic therapy sequence with surgery) in the training set. A diagnostic nomogram for predicting the early death risk was created based on these variables. Calibration curves showed a high agreement between the expected and actual probabilities. The area under the curves (AUC) for the training and validation sets were 0.798 and 0.811, respectively. Meanwhile, the novel-created diagnostic nomogram had the highest AUC value compared to each independent risk variables, which showed that the nomogram had the best discriminatory ability. The DCA indicated that the nomogram has the potential to provide greater clinical advantages across a broad spectrum of threshold probabilities. Furthermore, a nomogram-based risk classification system was constructed to help us identify the high-risk population facing early death. Conclusions: This study created a novel diagnostic nomogram to predict the probability of early death in older patients with intracranial glioma. In the meantime, a nomogram-based risk classification system was also constructed to help us identify the high-risk population facing early death in older patients with intracranial glioma and provide them with specialized care to increase their benefit from survival.