High rates of over-diagnosis of pulmonary embolism by CT angiography

1. In a retrospective review of over 900 CT pulmonary angiography (CTA) studies in a single tertiary-care institution, image reinterpretation demonstrated a 25.9% discordance rate for diagnosis of pulmonary embolism (PE) from the original interpretation.

2. The most common cases of discordant diagnoses occurred in solitary segmental or subsegmental PEs due to breath motion and beam-hardening artifacts.

Evidence Rating Level: 3 (Average)

Study Rundown: Pulmonary embolism is a common, but potentially life threatening diagnosis among both hospitalized patients and those presenting to the emergency department. Computed tomographic pulmonary angiography (CTA) examinations demonstrate high sensitivity and specificity in cases of high clinical suspicion of PE. However, the accuracy of CTA significantly decreases in patients with low pretest probabilities of PE, with false positive rates reported at over 40%. This may result in significant over-diagnosis as well as increased morbidity and mortality from long-term anticoagulant use among those with an erroneous diagnosis. The purpose of this study was to determine the rate of over-diagnosis of PE by CTA in a single-institution, tertiary-care hospital. Researchers retrospectively reviewed over 900 CTA examinations in a 1-year period from 2012 to 2013. Examinations reported as positive for PE were subsequently retrospectively reinterpreted by 3 subspecialty radiologists. A discrepancy was noted if all 3 radiologists found a negative CTA result.

At the conclusion of the study, the overall discrepancy rate was 25.9% of all positive CTA investigations. In sub-group analysis, the most common cases of discordance were those in which solitary sub-segmental or basilar PEs were initially diagnosed. The most common reasons for false-positive interpretation was breathing motion artifact and artifacts from high-density structures adjacent to the pulmonary vessels. The results of this study demonstrate a high rate of false-positive diagnosis of PE by pulmonary CTA and given the risk of treatment for a false-positive PE diagnosis, these results support the use of additional measures to reduce the risk of PE misdiagnosis such as optimizing image quality and increased use of clinical PE pretest probability assessment tools. The study is limited by its retrospective, single-institution design. Future studies should aim to examine the most effective systematic techniques by which to reduce false-positive diagnoses of PE on CTA.

Click to read the study in AJR

Relevant Reading: CT Angiography of Pulmonary Embolism: Diagnostic Criteria and Causes of Misdiagnosis

In-Depth [retrospective cohort]: This study was a retrospective review of all CTA examinations in a 12-month period from August 2012 to July 2013 in a tertiary-care center in Ireland. All studies which had a positive finding of PE were included. Overall, 937 CTA studies were examined, of which 174 (18.6%) studies were initially reported as positive for PE. Each positive study was retrospectively reinterpreted by three subspecialist chest radiologists blinded to the original interpretation. A discrepancy was noted when there was consensus between 3 reviewers that no PE was observed in reinterpretation of the CTA. At the conclusion of the study, 45 of the 174 (25.9%) cases were discordant with the initial interpretation. Of the discordant reports, 46% were solitary PEs and were either located in the segmental or sub-segmental vessels within the lung. The most common causes of false-positive interpretations were breathing motion artifacts (42.2%), beam hardening artifacts (22.2%) and mixing of opacified and unopacified blood (11.1%) due to a poor contrast bolus.

Image: CC/Wiki/Aung Myat and Arif Ahsan

©2015 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.