Development and External Validation of a Nomogram to Predict Prognosis of Patients With Urothelial Carcinoma of Bladder

Background This research aimed to create and validate nomogram predicting overall survival (OS) for urothelial carcinoma of the bladder (UCB) patients. Methods We sourced 15,606 UCB patients diagnosed between 2004 and 2015 from the Surveillance, Epidemiology, and End Results database. The patients were randomized into training (70%) and internal validation (30%) cohorts. In addition, 122 patients from Minzu Hospital of Guangxi Zhuang Autonomous Region between 2012 and 2022 were selected as the external validation cohort. Utilizing univariate and multivariate Cox regression analyses, we devised nomograms forecasting 1-, 3-, and 5-year OS. Several metrics, including the consistency index (C-index), calibration plots, area under the receiver operator characteristics (ROC) curve, and decision curve analysis (DCA) were used to validate the validity and clinical utility of the model. Patients were categorized into high- and low-risk profiles, and their survival outcomes were contrasted using the Kaplan-Meier method and the log-rank test. Results Age, marriage, AJCC stage, tumor size, surgery, and chemotherapy were identified as predictors of OS. In the training cohort, internal validation cohort and external validation cohort, the nomogram for predicting OS achieved C-index values of 0.718 (95% CI: 0.712–0.724), 0.714 (95% CI: 0.704–0.724), and 0.725 (95% CI: 0.641–0.809), respectively. In all cohorts, the calibration plots revealed high consistency between actual and predicted values. The nomogram depicted by ROC and DCA showcased superior stability, predictive value, and clinical applicability for 1, 3-, and 5-year OS. The risk stratification delineated patients into low- and high-risk brackets, revealing significant prognostic distinctions ( P  < 0.05). Conclusions Based on the SEER database and Chinese data, we developed a reliable nomogram forecasting 1-, 3-, and 5-year OS for UCB patients. The model can identifie high-risk patients, aiding clinicians in personalised treatment and prognostic evaluations.