PECSS: Pulmonary Embolism Comprehensive Screening Score to safely rule out pulmonary embolism among suspected patients presenting to emergency department

Abstract

Background: Pulmonary embolism is a severe cardiovascular disease and can be life-threatening if left untreated. However, the detection rate of pulmonary embolism using existing pretest probability scores remained relatively low and clinical rule out often relied on excessive use of computed tomographic pulmonary angiography.

Methods: We retrospectively collected data from pulmonary embolism suspected patients in Zhongshan Hospital from July 2018 to October 2022. Pulmonary embolism diagnosis and severity grades were confirmed by computed tomographic pulmonary angiography. Patients were randomly divided into derivation and validation set. To construct the Pulmonary Embolism Comprehensive Screening Score (PECSS), we first screened for candidate clinical predictors using univariate logistic regression models. These predictors were then included in a searching algorithm with indicators of Wells score, where a series of points were assigned to each predictor. Optimal D-Dimer cutoff values were investigated and incorporated with PECSS to rule out pulmonary embolism.

Results: In addition to Wells score, PECSS identified seven clinical predictors (anhelation, abnormal blood pressure, in critical condition when admitted, age > 65 years and high levels of pro-BNP, CRP and UA,) strongly associated with pulmonary embolism. Patients can be safely ruled out of pulmonary embolism if PECSS ≤ 4, or if 4 < PECSS ≤ 6 and D-Dimer ≤ 2.5 mg/L. Comparing with Wells approach, PECSS achieved lower failure rates across all pulmonary embolism severity grades. These findings were validated in the held-out validation set.

Conclusions: Compared to Wells score, PECSS approaches achieved lower failure rates and better compromise between sensitivity and specificity. Calculation of PECSS is easy and all predictors are readily available upon emergency department admission, making it widely applicable in clinical settings.

Trail registration: The study was retrospectively registered (No. CJ0647) and approved by Human Genetic Resources in China in April 2022. Ethical approval was received from the Medical Ethics Committee of Zhongshan Hospital (NO.B2021-839R).

Keywords: D-Dimer; Emergency Department; Pretest probability; Pulmonary Embolism; Screening Score.

Fig. 1

Study workflow. The workflow of excluding patients passing the exclusion criteria

Fig. 2

Univariate association of each clinical predictor with PE status. Odds ratio (OR) and 95% confidence interval (CI) were obtained from univariate logistic regression model. Hgb, hemoglobin; WBC, white blood cell count; NEUT, neutrophil count; MONO, monocyte count; PLT, platelet; CRP, c-reaction protein; abnormal BP, abnormal blood pressure; pro-BNP, N-terminal (NT)-pro hormone BNP; ALB, blood albumin; ALT, Alanine transaminase; GLO, globulin; Cr, creatinine; BUN, Blood Urea Nitrogen; UA, blood uric acid; cTNT, cardiac troponin test

Fig. 3

Flowchart summary of the PECSS + D-Dimer screening approach

Fig. 4

Distribution of PECSS among patients with different PE severity grades in the derivation and validation set. Shaded areas indicate patients proportions where PE can be safely ruled out and no CTPA is needed: PECSS no higher than 4, or 4 < PECSS ≤ 6 and D-Dimer ≤ 2.5 mg/L