Can a Decision Tool Reduce CT Use?

Maybe

By Ruth SoRelle, MPH

Can a decision tool based on a six-year-old study help emergency physicians reduce the use of CT scans in patients with minor head injuries?

Yes and no, said experts who are looking for ways to help emergency physicians make that call. Susan Van Pelt, MD, the director of quality improvement and risk management with the Emergency Physician Professional Association in Minneapolis,used the criteria in NEXUS II (the National Emergency X-Radiography Utilization Study: J Trauma2005;59[4]:954;Ann Emerg Med 2002;40[5]:505) to educate physicians about the need to reduce radiation exposure in patients and the viability of clinical criteria in deciding when to use computer tomography. But Jerome Hoffman, MD, a professor of emergency medicine at the University of California Los Angeles who took part in NEXUS II, said while the decision instrument in NEXUS II “might be a step up from very poor care, the average doctor does better or at least as well as the instrument.” The difference, he said, is minimal.

(Photo by James R. Roberts, MD)No one is denying the need to reduce radiation from medical imaging. On Feb. 9, the U.S. Food and Drug Administration announced an initiative to reduce unnecessary radiation exposure from computed tomography, nuclear medicine studies, and fluoroscopy. That approach wants device manufacturers to incorporate safeguards into their machines’ design, to develop safer technology, and to train practitioners in their safe use.

“The amount of radiation Americans are exposed to from medical imaging has dramatically increased over the past 20 years,” said Jeffrey Shuren, MD, JD, the director of the FDA’s Center for Devices and Radiological Health. “The goal of FDA’s initiative is to support the benefits associated with medical imaging while minimizing the risks.”

Both Drs. Van Pelt and Hoffman try to adhere to that goal, they said. Dr. Van Pelt’s medical group first reviewed more than 1,300 charts of patients seen with head injury and studied patients with diagnosed intracranial injury seen during 2007.They particularly looked for the variables cited in the NEXUS-II study as indicating the need for a CT scan: vomiting more than once, clinical evidence of skull fracture, abnormal behavior (including evidence of intoxication), focal neurologic deficit, altered level of consciousness, scalp hematoma, coagulopathy (including use of aspirin, other antiplatelet agents, or medical condition that could result in impaired clotting ability), and age 65 or older.

They found that scan rates varied with patient age. Eighty-nine percent of patients over age 65 had a head CT, but the rate for children under the age 3 was 37 percent. Using the NEXUS II criteria might have avoided many of these scans, they found. The researchers said physicians ordered scans on 53 percent of all patients who had none of the NEXUS-II criteria and 42 percent of pediatric patients with none of the criteria.

On the other side of the coin, however, they found that NEXUS II was relatively insensitive. It missed six of 77 patients with significant intracranial injuries, a sensitivity of only 92 percent. In reviewing the charts of missed patients, the investigators determined that four of the six had complained of a remarkable headache. Adding that to the NEXUSII criteria increased sensitivity to 97 percent.

The group then began an extensive education program for its partners, presenting the results of the study and conducting a seminar on the risks of diagnostic imaging. They then distributed written material, naming the criteria “the EPPA-9” after their association. While the rule could not supplant clinical judgment and was only 97 percent sensitive, it reassured the physicians that if they felt a CT scan was unnecessary, the rule would support their clinical decision.

When the group re-evaluated CT use in 200 patients seen with minor head injury in early 2009, they found marked reduction in use, from nearly 60 percent of patients aged 3 to 18 years in 2007 to just more than 30 percent in 2009. The rate of scanning without indication went from 42 percent in 2007 to less than nine percent in 2009. The project won EPPA the 2009 Partner in Excellence Award from HealthPartners, a prominent Minnesota cooperative insurer.

“I thought the change of practice was remarkable,” said Dr. Van Pelt, but noted that she and her colleagues still insist that the rule cannot substitute for clinical judgment. “We communicated to our partners that even this should not tell you not to do a scan on someone about whom you are worried.”

The problem is applying a decision tool to such a complicated problem, said Dr. Hoffman. “I applaud someone who is trying to find a way to cut down on head CT against his own self-interest but in the interest of the patients. NEXUS II is fine and might be a little better than off the top of your head. When you start to throw in Coumadin, an older patient, then the question is, ‘Do we have them back tomorrow to re-evaluate?’ There are so many subtleties one can add to the equation.”

Dr. Hoffman cautioned against relying on methods that disregard clinical acumen. “Our brains are pretty good,” he said. “We notice things you cannot put in an instrument. My experience is that when you try to make an instrument, as the clinical condition gets more complex and varied, the instrument performs less and less well.”

Emergency physicians also have to overcome financial incentives that encourage them to perform such tests, said Drs. Hoffman and Van Pelt, pointing out that patients often demand tests because they want the “best” care. Recent concern over health care costs hasprompted some patients to question whether tests are really needed, but more often they say, “I want the best. Do everything. I have insurance,” Dr. Van Pelt said.

Then the problem becomes convincing the patient with minor head injury and no criteria for CT that her recommendation is not to perform the test. She said she explains the evidence behind the decision, and adds, “This is what I would do for my own child, my sister.”

But even that is a double-edged sword, said Dr. Hoffman. He remembered a case in which he tried to spare a 17-year-old girl who had been in a minor car accident. Surgeons and emergency physicians at the hospital disagreed over the need to perform whole body scans on such patients. When surgeons wanted such a scan on the young woman, he opposed it. She was a minor, and Dr. Hoffman tried to explain the situation to her father, stressing that the radiation could be harmful and that he thought the test should not be done.At the end of their discussion, the father was silent for a moment, and then said, “But we have MediCal.”

“It was clear what he had heard,” said Dr. Hoffman. “He thought that I think he’s Hispanic and poor, and I don’t want to give his daughter good care.” Dr. Hoffman agreed to the scan. “We have taught people that all these tests and medicine are wonderful. We are paying for it now,” he said, pointing out that patients are as well.

Comments about this article? Write to EMN at .

In the April issue

Editorial: Time for Leadership

Dr. Susan Van Pelton why explaining the risks of radiation to patients provides an opportunity and a challenge.