Analysis of predictive factors in children with high suspicion of pulmonary embolism

Abstract

AbstractBackground and AimSince there is a limited number of predictive models designed for pediatric patients with pulmonary embolism (PE), healthcare providers have resorted to utilizing the Wells criteria and PE Rule‐out Criteria (PERC) that have been validated for use in adults. The objective of the study is to identify the predictive factors for PE and assess the effectiveness of the PERC, Wells, and pediatric‐specific PE (PPE) criteria.MethodsThe study included individuals between the ages of 0–18 years who underwent pulmonary CT angiography and/or ventilation–perfusion scintigraphy with suspicion of PE. Demographic characteristics, symptoms, physical examination findings, risk factors, and laboratory and imaging results were retrospectively analyzed, and PERC, Wells, and PPE criteria were applied to the study population.ResultsOf the 110 patients included in the study, 27 (24.5%) had PE. Saturation, albumin, cough, calf swelling, central catheter, and malignancy were found to significantly contribute to the model. The total weighted risk score, which represents the sum of all predictive scores, ranged from 0 to 16 with a mean of 5.41 ± 4.02. When the cut‐off >6.5, the model had good discrimination power for positive PE (AUC 0.79, 77% sensitivity, and 70% specificity). In our study, the Wells criteria showed a sensitivity of 96% and a specificity of 24%. The PERC exhibited a sensitivity of 96% and a specificity of 21%, while the PPE demonstrated a sensitivity of 74% and a specificity of 30%.ConclusionWhile the sensitivity of Wells, PERC, and PPE models was higher than our developed model, their specificities were considerably lower than our model.