Pulmonary embolism at CT angiography: implications for appropriateness, cost, and radiation exposure in 2003 patients

To determine whether thromboembolic risk factor assessment could accurately indicate the pretest probability for pulmonary embolism (PE), and if so, computed tomographic (CT) angiography might be targeted more appropriately than in current usage, resulting in decreased costs and radiation exposure.

Institutional review board approval was obtained. Electronic medical records of 2003 patients who underwent CT angiography for possible PE during 1(1/2) years (July 2004 to February 2006) were reviewed retrospectively for thromboembolic risk factors. Risk factors that were assessed included immobilization, malignancy, hypercoagulable state, excess estrogen state, a history of venous thromboembolism, age, and sex. Logistic regressions were conducted to test the significance of each risk factor.

Overall, CT angiograms were negative for PE in 1806 (90.16%) of 2003 patients. CT angiograms were positive for PE in 197 (9.84%) of 2003 patients; 6.36% were Emergency Department patients, and 13.46% were inpatients. Of the 197 patients with CT angiograms positive for PE, 192 (97.46%) had one or more risk factors, of which age of 65 years or older (69.04%) was the most common. Of the 1806 patients with CT angiograms negative for PE, 520 (28.79%) had no risk factors. The sensitivity and negative predictive value of risk factor assessment in all patients were 97.46% and 99.05%, respectively. All risk factors, except sex, were significant in the multivariate logistic regression (P < .031).

In the setting of no risk factors, it is extraordinarily unlikely (0.95% chance) to have a CT angiogram positive for PE. This selectivity and triage step should help reduce current costs and radiation exposure to patients.