Rapid prediction of deterioration risk among non-high-risk patients with acute pulmonary embolism at admission: An imaging tool

Abstract

BackgroundComputed tomography (CT) pulmonary angiography as the first-line diagnosis tool of acute pulmonary embolism (PE), might improve this discriminatory power. We aimed to developed a simply tool combining multi-CT parameters to complete individualized risk assessment of deterioration in non-high-risk patients with acute PE at admission. MethodConsecutive non-high-risk patients with acute PE who were treated in a Chinese center during 2010–2021, were collected.Prognosis-related CT parameters were reviewed. Deterioration was defined as any adverse event within 30 day after admission. Eligible patients were randomized into derivation and validation cohorts. In the derivation cohort, CT parameters were screened for importance using classification tree methodology and enrolled variables was partitioned via curve-fitting and dose-response analysis. A nomogram was developed and the predictive power in both cohorts was evaluated based on the area under the receiver operating characteristic curve (AUROC) and the corresponding 95% confidence interval (CI). ResultA total of 1001 patients were included. The preliminary analyses revealed that deterioration risk was related to the right-to-left ventricular diameter ratio at 4-chamber view, pulmonary vein filling abnormality. After a curve-fitting to deterioration risk, these parameters were partitioned and used to develop a nomogram, which had AUROC values of 0.91 (95% CI: 0.87–0.96) in the derivation cohort and 0.89 (95% CI: 0.81–0.97) in the validation cohort. A web-based version of the radiomics scoring tool was published online for use in clinical practice (https://acutepeprediction.shinyapps.io/Radiomics_Predictive_Tool/). ConclusionThis simply tool can complete rapid estimation of deterioration risk among non-high-risk acute PE patients at admission.