Overdiagnosis and treatment of pulmonary embolism: the emperor still has no clothes

That fairy tail reference, of course, is to the famous Annals article way back in 1977. Some 38 years later we still may be overdiagnosing and overtreating PEs, but for different reasons. That is the subject of a great post overat EP Monthly. It focuses on 1) the knee jerk overuse of CTA in patients with chest symptoms (failure to apply clinical prediction rules such as the Wells score) and 2) the frequent finding, with the current heavy use of CTA, of questionable isolated small filling defects that may not need to be treated, or may not represent PE at all. It's a great read and well referenced. In reading the post, keep in mind the following:

It is not known whether such low clot burden PEs need to be treated at all. Investigators believe there is clinical equipoise and so a randomized clinical trial is now ongoing to answer the question.

Many such low burden filling defects picked up on CT would be “missed” with perfusion lung scanning. However, perfusion lung scanning is at least as sensitive (with normal perfusion or when very low probability criteria are met) as CT for clinically significant emboli and in terms of outcomes.

Despite the reported incidence of PE going way up over the years since the introduction of CT motality for PE has not changed, indirect evidence that the small lesions picked up on CT are clinically insignificant and should not be treated.

The use of CTA in preference to VQ scanning is popularity based and not evidence based. There is no evidence that CT is superior to nuclear scanning as the initial imaging modality.