Overdiagnosis of Pulmonary Embolism by CT

Samuel Z. Goldhaber, MD


October 01, 2015

This feature requires the newest version of Flash. You can download it here.

Overdiagnosis of PE With CT Scans

Hello. This is Dr Sam Goldhaber for the Clot Blog on at Medscape, speaking to you from the European Society of Cardiology Congress in London.

Today I'm going to speak about overdiagnosis of pulmonary embolism (PE) by chest CT scan. I'd like to start by emphasizing that, overall, PE remains underdiagnosed because the symptoms and signs are nonspecific and because patients who die suddenly from PE are often signed out on the death certificate as having acute myocardial infarction.

There is, however, a problem with overdiagnosis of PE on chest CT scans. A recently published study[1] did an expert review of about 175 patients who were classified as having PE by chest CT scanning. One out of four of these patients were subsequently deemed by a panel of expert readers to have artifacts that mimicked PE but that weren't really PE.

The biggest problems occurred when there was only a single so-called PE lesion on the chest CT scan, or in patients who had only subsegmental defects without any segmental or lobar defects on the scan. The number-one cause of these artifacts was a motion artifact from breathing, but also there was some problem with beam attenuation. When the beams hit the superior vena cava or the right atrium or the right ventricle, they could give the appearance of a cardiac pulsation artifact. There was also a problem with artifacts due to poor mixing of contrast.

Implications for Practitioners

What are the implications for us as practitioners? First of all, we should read our CT scans on our patients. If it's obvious to us, as nonexperts on CT scan reading, that the patient has PE, we can go ahead and treat that patient with confidence. If we're really not sure whether there is a PE when we look at the scan ourselves—particularly if we can only see one possible defect, or only subsegmental defects—these we should probably review, ideally in person, with an expert radiologist or specialist in cardiovascular imaging to make sure that the patient has PE before we embark on anticoagulation.

I think we have to realize that PE suffers from two curses: the curse of underdiagnosis, and also a smaller but present possibility of overdiagnosis.

This is Dr Sam Goldhaber, signing off for the Clot Blog.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.