Assessment of Right Ventricular Strain by Computed Tomography Versus Echocardiography in Acute Pulmonary Embolism. (11th March 2017)

• Record Type: Journal Article
• Authors: Dudzinski, David M.; Hariharan, Praveen; Parry, Blair A.; Chang, Yuchiao; Kabrhel, Christopher
• Editors: Mitchell, Alice M.
• Is Part Of: Academic emergency medicine. Volume 24:Number 3(2017)
• Page Start: 337
• Page End: 343
• Publication Date: 2017-03-11
• Subjects: Emergency medicine -- Periodicals; 616.02505
• DOI: 10.1111/acem.13108
• Languages: English
• ISSNs: 1069-6563

Abstract

Abstract: Background: Right ventricular strain (RVS) identifies patients at risk of hemodynamic deterioration from pulmonary embolism (PE). Our hypothesis was that chest computed tomography (CT) can provide information about RVS analogous to transthoracic echocardiography (TTE) and that RVS on CT is associated with adverse outcomes after PE. Methods: Consecutive emergency department patients with acute PE were prospectively enrolled and clinical, biomarker, and imaging data were recorded. CTs were overread by two radiologists. We compared diagnoses of RVS on CT (defined as right ventricle:left ventricle ratio ≥ 0.9 or interventricular septal bowing) to echocardiography (defined as right ventricular hypokinesis, right ventricular dilatation, or interventricular septal bowing). We calculated the test characteristics (with 95% confidence interval) of CT and TTE for a composite outcome of severe clinical deterioration, thrombolysis/thrombectomy, or death within 5 days. Results: A total of 298 patients were enrolled; 274 had CT and 118 had formal TTE. Of the 104 patients who had both CT and TTE, the mean (±SD) age was 58 (±17) years; 50 (48%) were female and 88 (85%) were Caucasian. Forty‐two (40%) had RVS by TTE and 75 (72%) had RVS by CT. CT and TTE agreed on the presence or absence of RVS in 61 (59%) cases (κ = 0.24). Using TTE as criterion standard, the test characteristics of CT for RVS were as follows: sensitivity = 88%, specificity = 39%, positive predictive value = 49%, and negative predictive value = 83%. Fourteen (13%) patients experienced severe clinical deterioration or required hospital‐based intervention within 5 days. This occurred in 30% of patients with RVS on both TTE and CT, 20% of patients with RVS on TTE alone, 3% of patients with RVS on CT alone, and 4% of patients without RVS on either modality. Conclusions: In acute PE, CT is highly sensitive but only moderately specific for RVS compared to TTE. RVS on both CT and TTE predicts more events than either modality alone. TTE confers additional positive prognostic value compared to CT in predicting post‐PE clinical deterioration.