An Epidemic of CT Pulmonary Angiography: Physicians and Patients Beware!

Although CT pulmonary angiography has been a valuable addition to our diagnostic armamentarium for pulmonary embolism, permit me to say, “Enough is enough.”

As a primary care internal medicine attending, I have been experiencing a rash of negative CT angiograms in patients who even hint minimally at either chest pain or shortness of breath without substantive evidence for risk factors. I am gratified to present a study1 and related editorial2 squarely focusing on the overuse of this diagnostic modality.

NOT A HARMLESS TEST

What is known? The application of CT pulmonary angiography has “increased dramatically.”1 The test itself is not harmless. It places persons at risk for ionizing radiation exposure, contrast nephropathy, soft tissue injury, and anaphylaxis.1 More so, a false-positive diagnosis of pulmonary emboli exposes patients to risks inherent in anticoagulation.

Adams and colleagues1 reviewed 3500 consecutive CT pulmonary angiograms to classify their cohort as either “pulmonary embolism unlikely” or “pulmonary embolism likely” based on a revised Geneva score (a validated tool to assess pretest probability for pulmonary emboli and included with the paper). The angiograms that were performed were either concordant or discordant with Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) II guidelines on the basis of the following two criteria:

•Pulmonary embolism likely or unlikely based on pretest probability.
•A highly sensitive D-dimer test that was positive.

A total of 1592 out of 3500 angiograms were consistent with the recommendations of PIOPED II. However, 1908 of the studies were performed on patients with low Geneva scores (militating against pulmonary embolism) without a D-dimer test or even after a negative D-dimer test. The potential false-positive rate was also high for patients whose revised Geneva score was less than or equal to 10 with a negative D-dimer value.

THE DANGERS OF RELYING ON GESTALT

The results are disturbing. More than half of 3500 studies were discordant with PIOPED II recommendations. Although physicians scoff at the volume of guidelines, for pulmonary emboli, they are evidence-based, efficient to apply, and invaluable. The authors presumed most emergency physicians bypassed the Geneva score (or other choices such as the Wells score) and replaced them with gestalt. In this instance, gestalt seems terribly inadequate. The editorialist made insightful observations as well.2 I begin with his italicized quote, “If a patient has no clinical risk factors for pulmonary embolism and has a negative D-dimer result, the probability of pulmonary embolism is less than 1%. You simply should never order a computed tomography angiogram (CTA) for such a patient.” He also proposed a caveat based on a young patient he evaluated recently. The patient had so many visits to the emergency department over 2 years with shortness of breath followed by CT angiography to rule in or out pulmonary emboli that standard dose estimates determined she received more radiation than a 1945 Hiroshima resident 5 miles from ground zero.

The guidelines from PIOPED II take advantage of validated pretest probability. Ignoring them and relying on gestalt leads to unnecessary studies. The downside of these studies is real. It is time for a change in practice to protect patients from radiation, contrast complications, and inappropriate
anticoagulation.