A third of ED imaging for PE is avoidable

By Sarah Guy

Data from 12 hospitals in the USA indicate that one-third of imaging performed in the emergency department (ED) for suspected pulmonary embolism (PE) is avoidable, according to a National Quality Forum (NQF) measure.

The measure says that computed tomography (CT) is unnecessary in hemodynamically stable (systolic blood pressure at least 90 mmHg) patients with a low pretest probability (PTP; Wells score less than 2) of PE, in whom D-dimer testing is either not performed, or is negative.

However, in a study of 3710 ED patients, 65% underwent CT imaging for PE, 32% of which was avoidable under the NQF criteria.

"Rising health care costs from advanced imaging and improved understanding of the risks associated with ionizing radiation and intravenous contrast are driving efforts to improve imaging efficiency," write Christopher Kabrhel (Massachusetts General Hospital, Boston) and colleagues.

They suggest their findings indicate that there is a large opportunity to improve efficiency in this area, and that future work should focus on interventions to "close this performance gap."

Among the 1205 patients in the study who underwent avoidable CT imaging, 67% did so because no D-dimer testing was performed, and 33% did so because imaging was performed despite a negative D-dimer test result.

Kabrhel and co-authors believe that the "phenomenon" of a failure to perform D-dimer testing could be explained by physician bias toward more "definitive" testing using CT of the pulmonary arteries, overestimation of D-dimer test false-positivity, or underestimation of its sensitivity.

Just 50 patients (1.3% of those imaged) who did not undergo D-dimer testing and who received avoidable imaging were ultimately diagnosed as having PE, whereas eight (0.2%) patients with negative D-dimer results were diagnosed with PE after imaging.

This validates the NQF measure and refutes the notion that high-measure performance is associated with the unintended consequence of missed PE, remark Kabrhel et al in the Archives of Internal Medicine.

In light of few patient characteristics being predictive of avoidable imaging, the team suggests that imaging inefficiency is likely related to variation in physician-level risk tolerance, patient preference, or hospital characteristics not measured by their study.

Licensed from medwireNews with permission from Springer Healthcare Ltd. ©Springer Healthcare Ltd. All rights reserved. Neither of these parties endorse or recommend any commercial products, services, or equipment.