Validation of a novel diagnostic nomogram for pulmonary thromboembolism in the emergency department

ABSTRACT

Objective@#Diagnosis of pulmonary thromboembolism (PTE) is essential for preventing serious complications in the emergency department (ED) or intensive care unit. Contrast computed tomography (CT) of the chest is used for confirming pulmonary embolism, but there is a low specificity and radiation- or contrast-related side effects. We developed a novel nomogram to facilitate decision-making for performing contrast CT of the chest in the ED. @*Methods@#A retrospective observational study was conducted to develop a prediction model of PTE. The prediction model was derived from demographic characteristics, clinical history data and results of laboratory tests, ultrasonography and echocardiography. A nomogram was constructed from the variables of the prediction model and validated. @*Results@#A total of 326 patients were analyzed (a training cohort, 260; a validation cohort, 66). Wells’ score, D-dimer level>1,100 ng/dL, positive McConnell’s sign and D-shaped left ventricle were associated with the occurrence of PTE. The overall predictive accuracy of the prediction model was 0.802 (0.748-0.849) (area under the curve with 95% confidence interval). The calibration plots for the probability of PTE showed good agreement between the nomogram prediction and actual probability among cohorts. @*Conclusion@#A novel nomogram using risk stratification, laboratory test and sonographic examination findings is a good screening tool for predicting PTE, and it can be helpful to decide whether an ED physician should perform a contrastenhanced chest CT in the ED.