A novel predictive model for differentiating cardiac and non-cardiac syncope in syncope-admitted patients

Abstract

Abstract Background Syncope due to cardiac causes presents a higher risk for mortality or other adverse events compared to those with non-cardiac causes. Thus, a critical focus in syncope management must be the early identification of these patients. Purpose This study sought to explore the predictive role of readily available clinical features for cardiac syncope and develop a diagnostic nomogram to facilitate its adoption in practice. Methods 877 patients who were hospitalized with syncope and had a confirmed cause were included in this retrospective study at a tertiary heart center. They were randomly divided into the training set and validation set and were categorized into two groups based on their underlying cause: cardiac syncope and non-cardiac syncope. A wide range of clinical features were analyzed and selected using LASSO regression. Subsequently, we utilized multivariable logistic regression analysis to identify independent predictors and construct a nomogram. The receiver operating characteristic (ROC) curves and the area under the curves (AUCs) were used to assess the predictive accuracy of the predictive model. Furthermore, calibration curves were used to measure the alignment between the predicted and actual observed probability, and decision curve analysis (DCA) was used to evaluate the clinical benefit of the nomogram. Results Within the entire cohort, 521 patients were diagnosed with cardiac syncope and 356 patients were diagnosed with non-cardiac syncope. Among 36 candidate variables, five independent predictors for cardiac syncope were selected using the LASSO logistic regression analysis, including BMI (OR 1.088; 95%CI 1.022-1.158), LVEF (OR 0.940; 95%CI 0.908-0.973), logarithmic NT-proBNP (OR 1.463; 95%CI 1.240-1.727), chest symptoms preceding syncope (OR 5.251; 95%CI 3.326-8.288), and abnormal electrocardiogram (OR 6.171; 95%CI 3.966-9.600). Subsequently, a nomogram based on the five independent predictors was developed, yielding AUC of 0.873 (95% CI 0.845–0.902) and 0.856 (95%CI 0.809-0.903) in the training set and the validation set, respectively. The calibration curves showcased the nomogram's strong calibration, and DCA curves revealed its potential as a highly effective tool in clinical practice. Conclusions This study has successfully developed and internally validated a novel predictive nomogram for distinguishing between cardiac and non-cardiac syncope, which holds a promise to serve as an effective tool to help facilitate early identification of such patients in clinical practice.ROC curves of the training and test setThe nomogram and web-based calculator