A new study suggests that CT angiography (CTA) can be better targeted to patients with risk factors for pulmonary embolism (PE). The study also indicates CTA may be overused in many patients with suspected PE. Thromboembolic risk factor assessment could suffice for many patients, eliminating CTA and helping reduce radiation exposure and CTA costs. Results of the study appear in the online edition and August print issue of the journal Radiology.
Because of its high sensitivity and specificity CTA has become the preferred modality to diagnose PE. However, the increasing use of the examination has fueled concerns over procedure costs and radiation exposure to patients, along with risks associated with the use of contrast agents. Overuse of CT also can back up patient flow in the ER and radiology department, said Eric vanSonnenberg, MD, chief medical officer and chief of interventional radiology at Kern Medical Center/University of California Los Angeles (UCLA) and a lead researcher for the study.
In 2006, the Prospective Investigation of Pulmonary Embolism Diagnosis II study ( PIOPED) underscored the need to accurately determine pretest clinical probability for PE to target CTA. The recent study sought to determine if a brief thromboembolic risk factor assessment could accurately gauge pretest clinical probability for PE and, thus, more appropriately target CTA.
“This study suggests that physicians could better utilize a very valuable, but expensive, test that is all too commonly negative for PE,” said vanSonnenberg. The data also indicate that CTA is overused at times. Patients can be triaged, and CTA can be more appropriately requested, continued vanSonnenberg. According to the study, thromboembolic risk factor assessment could cut the frequency of CTA for PE diagnosis, resulting in reduced radiation exposure and lower costs.
In the retrospective study, researchers evaluated the possibility of using thromboembolic risk factor assessment to reduce the number of CTAs positive for PE and sought to identify risk factors that make a positive CTA more likely.
Researchers reviewed the electronic medical records of 2,003 patients who underwent CTA for possible PE at St. Joseph’s Hospital in Phoenix, between July 2004 and February 2006. The patient population was comprised of 847 men and 1156 women, with a mean age of 51 and 49 years respectively.
The assessed risk factors included immobilization, malignancy, hypercoagulable state, excess estrogen state, a history of venous thromboembolism, age and gender. Logistic regressions were conducted to test the significance of each risk factor. The number of risk factors was quantified and recorded for each patient. Researchers also reviewed D-dimer test results for the 99 patients with available results.
Overall, CTAs were negative for PE in 1,806 of 2,003 patients, or 90.2 percent. Among the 197 patients with CTAs positive for PE, 192, or 97.5 percent, had one or more positive risk factors. All risk factors except gender were determined to be statistically significant. Age of 65 years or older and immobilization were the most common risk factors in positive PE patients. Of the 192 patients with studies positive for PE, 6 percent originated the ED, and 13 percent were inpatients.
Of the 1,806 patients with CTAs negative for PE, 520 (28.8 percent) had no risk factors. Furthermore, excluding age and gender, 1,119 (62 percent) had no risk factors. Risk factor assessment had a sensitivity of 97.5 percent and a negative predictive value of 99 percent in all patients.
Without any thromboembolic risk factors, there was only a 0.95 percent chance of a CTA positive for PE. A negative D-dimer, in addition to the absence of thromboembolic risk factors, further reduced the chance of a CTA positive for PE, added vanSonnenberg. In fact, the D-dimer test yielded 100 percent sensitivity and negative predictive value in this study.
“Thromboembolic risk factor assessment is an effective clinical method to determine when to perform CTA for PE,” Mark D. Mamlouk, MD, lead author and radiology resident at the University of California, Irvine in Orange, Calif. “Risk assessment can be performed when clinicians acquire their patients’ history. It takes only a few minutes, and there’s no cost.” This model avoids downsides of CTA, added vanSonnenberg.
This study amplified the previous research published in the American Journal of Roentgenology