[Construction and validation of acute pulmonary embolism diagnostic model based on clinical and laboratory...

Objective: To establish and validate the diagnostic model of acute pulmonary embolism (APE) based on clinical and laboratory variables. Methods: This retrospective analysis was conducted on patients with suspected APE who underwent CT pulmonary angiography at the First Affiliated Hospital of Henan University of Chinese Medicine between February 2015 and December 2023. The patients were randomly divided into a training set and a validation set at a ratio of 7∶3. Clinical and laboratory data of the enrolled patients were collected, and patients were divided into an APE group and a non-APE group according to CT pulmonary angiography results. In the training set, univariate and multivariate logistic regression as well as Lasso regression were used to identify risk factors for APE, and a diagnostic model was developed and validated. Receiver operating characteristic curves were plotted, and calibration and decision curves were used to assess the performance of the diagnostic model. The diagnostic efficacy of the model was compared with that of the Wells score and the revised Geneva score using the DeLong test. Results: A total of 752 patients were enrolled, aged (64±15) years, including 417 (55.5%) males. The training set included 526 patients and the validation set included 226 patients. The incidence of APE in this cohort was 48.7% (366/752), with 366 cases in the APE group and 386 in the non-APE group. Multivariable logistic regression analysis showed that cyanosis (OR=8.88, 95%CI 2.04-49.11), elevated neutrophil count (OR=1.82, 95%CI 1.06-3.15), elevated creatine kinase isoenzyme (OR=3.45, 95%CI 1.76-6.91), decreased partial pressure of carbon dioxide (OR=12.88, 95%CI 7.64-22.34), elevated age-adjusted D-dimer (OR=2.53, 95%CI 1.10-6.20), prolonged thrombin time (OR=4.08, 95%CI 2.06-8.33), and positive lower limb venous ultrasound for thrombus (OR=4.39, 95%CI 2.59-7.58) were risk factors associated with APE. The area under the curve (AUC) of the diagnostic model was 0.92 (95%CI 0.90-0.94) in the training set and 0.92 (95%CI 0.89-0.95) in the validation set. The diagnostic efficacy of this model was superior to that of the Wells score (AUC: 0.92 vs. 0.63, P<0.01) and the revised Geneva score (AUC: 0.92 vs. 0.59, P<0.01). Conclusion: The diagnostic model for acute pulmonary embolism constructed based on clinical and laboratory parameters demonstrates excellent diagnostic performance and may facilitate rapid and accurate screening in clinical practice.