Medical Device Daily

Physicians could improve their ability to accurately diagnose pulmonary embolism, which kills about 600,000 Americans a year, if they would extend the commonly used chest computed tomography (CT) to the legs and combine it with the physician's own clinical assessment of the patient, according to a new study funded by the National Heart, Lung, and Blood Institute (NHLBI) of the National Institutes of Health (NIH; both Bethesda, Maryland).

The Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) II study of 824 patients is the largest study ever conducted to assess the role of contrast-enhanced multidetector CT angiography for diagnosing pulmonary embolism, the institute said. The study was published in the June 1, 2006, issue of the New England Journal of Medicine.

“What brought this particular investigation to mind is the fact that people are using CT angiography to make the diagnosis, but literature had been a little bit unclear – had been a lot unclear – as to how accurate is the diagnosis,” Paul Stein, MD, lead author of the paper, director of research education at St. Joseph Mercy Oakland Hospital (Pontiac, Michigan), and professor of medicine at Wayne State University (Detroit), told Medical Device Daily.

Although small studies had been done at individual hospitals or even groups of hospitals, Stein said “there had never been any tests with a sufficient number of patients to be certain.” Also, he said that most of the studies that were published were conducted using “first-generation CT scanners, and we needed to know the accuracy with more up-to-date CT scanners.”

The PIOPED II study researchers reported in the NEJM that chest CT angiography alone detects suspected pulmonary embolism (PE) in only 83% of patients. In contrast, combined results of the chest CT angiogram and the leg CT scan detect clots in 90% of patients.

The researchers recommend that physicians consider additional test results before ruling out PE in patients whose scan does not detect clots but whose clinical assessment suggests a high likelihood of PE, i.e., the extension of CT scans to the legs.

According to Stein, PE in 90% of patients originates from blood clots in the leg veins, or what is commonly known as deep vein thrombosis (DVT). When these clots in the legs travel to the lungs through the right chambers of the heart, into the artery and then into the lungs, the “clot gets caught” there because the “veins taper off” in the lungs, he said.

The extension of the CT scan to the legs would only take about an additional 2-3 minutes of both the patient's and the technicians' time, and it would require only one injection of a contrast agent that ultimately will travel to the legs after leaving the chest area, Stein said. It would necessitate only the moving down of the scanner over the legs, and the patient is already lying on the X-ray table.

“The patient does get more radiation, and we have ways of reducing that, but there is a radiation burden to it,” Stein said.

There would also likely be an additional “reading fee” by the radiologist by adding the legs, and the hospitals that own the scanners might charge an extra fee, he said.

In PIOPED II, researchers compared the accuracy of three ways to diagnose blood clots in 824 patients suspected of having pulmonary embolism; chest CT angiogram alone; chest CT angiogram with venous-phase imaging, or leg CT; and chest CT angiogram with an objective clinical assessment known as the Wells Score.

The NHLBI said that the Wells Score is a “validated tool to determine the likelihood that a patient has PE based on characteristics such as signs and symptoms, heart rate and risk factors. A high score indicates that a patient has a high probability of having pulmonary embolism.”

The NHLBI said that to determine their accuracy, the tests were compared with the participants' composite results from other validated diagnostic tests of PE.

Overall, the sensitivity, or the ability to detect clots, of the combined chest CT and leg CT was 90%, compared to 83% sensitivity of the chest CT angiogram alone. The specificity, or the ability to rule out the presence of clots, of the chest CT alone compared to the chest CT and leg CT combined was similar, or about 95%.

A high clinical probability combined with positive chest CT correctly indicated PE in 96% of study participants. However, in patients with a high clinical probability, a negative chest CT result did not “confidently” rule out a diagnosis of PE, the researchers said.

“I hope that [the study] calls to physicians' attention that contrast CT is, in fact, a good study, and generally accurate,” Stein told MDD. “It's more accurate if done in combination with the leg studies. My [sense of it] is more people are identified [with leg studies] because one might see the clot in the legs before they get to the lung or see something in the leg that would be causing problems . . .”

Still, Stein said that as with any test, “it's not perfect, and there are some patients that will be missed and some patients that will be diagnosed incorrectly.”

“The addition of a clinical assessment in combination with the CT helps reduce those errors,” Stein said. “And the combination strengthens the diagnosis as when the physician believes the disease is present, and the X-ray shows it, then we can be more confident of the diagnosis. The reverse is also true.”

Pulmonary embolism results in death for nearly one-third of untreated cases, but if it is detected early, that death rate can be lowered to between 3% and 8% with such therapies as anti-clotting medications and injections of clot-busting agents.

Stein said that the study used multidetector CT scans including those by Siemens Medical Solutions (Malvern, Pennsylvania) and GE Medical (Waukesha, Wisconsin).