Revision of Clinical Pre-Test Probability Scores in Hospitalized Patients with Pulmonary Embolism and SARS-CoV-2 Infection

Abstract

Background: The need for computed tomography pulmonary angiography (CTPA) to rule out pulmonary embolism (PE) is based on clinical scores in association with D-dimer measurements. PE is a recognized complication in patients with SARS-CoV-2 infection due to a pro-thrombotic state which may reduce the usefulness of preexisting pre-test probability scores.

Aim: The purpose was to analyze new clinical and laboratory parameters while comparing existing and newly proposed scoring system for PE detection in hospitalized COVID-19 patients (HCP).

Methods: We conducted a retrospective study of 270 consecutive HCPs who underwent CTPA due to suspected PE. The Modified Wells, Revised Geneva, Simplified Geneva, YEARS, 4-Level Pulmonary Embolism Clinical Probability Score (4PEPS), and PE rule-out criteria (PERC) scores were calculated and the area under the receiver operating characteristic curve (AuROC) was measured.

Results: Overall incidence of PE among our study group of HCPs was 28.1%. The group of patients with PE had a significantly longer COVID-19 duration upon admission, at 10 vs 8 days, p = 0.006; higher D-dimer levels of 10.2 vs 5.3 $\mu$ g/L, p $<$ 0.001; and a larger proportion of underlying chronic kidney disease, at 16% vs 7%, p = 0.041. From already established scores, only 4PEPS and the modified Wells score reached statistical significance in detecting the difference between the HCP groups with or without PE. We proposed a new chronic kidney disease, D-dimers, 10 days of illness before admission (CDD-10) score consisting of the three aforementioned variables: C as chronic kidney disease (0.5 points if present), D as D-dimers (negative 1.5 points if normal, 2 points if over 10.0 $\mu$ g/L), and D-10 as day-10 of illness carrying 2 points if lasting more than 10 days before admission or 1 point if longer than 8 days. The CDD-10 score ranged from -1.5 to 4.5 and had an AuROC of 0.672, p $<$ 0.001 at cutoff value at 0.5 while 4PEPS score had an AuROC of 0.638 and Modified Wells score 0.611. The clinical probability of PE was low (0%) when the CDD-10 value was negative, moderate (24%) for CDD-10 ranging 0-2.5 and high (43%) when over 2.5.

Conclusions: Better risk stratification is needed for HCPs who require CTPA for suspected PE. Our newly proposed CDD-10 score demonstrates the best accuracy in predicting PE in patients hospitalized for SARS-CoV-2 infection.

Keywords: SARS-CoV-2 infection; pre-test probability scores; pulmonary embolism.

Fig. 1.

Flowchart of the included patients who underwent CTPA due to clinical suspicion of PE. CTPA, computed tomography pulmonary angiography; DVT, deep vein thrombosis; ICU, intensive care unit; PE, pulmonary embolism; RT-PCR, real-time reverse transcriptase-polymerase chain reaction.

Fig. 2.

Receiver operating curve analysis according to PE. (A) Original CDD-10 regression model. (B) CDD-s10 scoring system.