Growing concern over the radiation exposure and costs associated with climbing rates of computed tomography scans in the emergency department create a dilemma for physicians and payers alike, but no consistent method can determine who should have a head scan.
Jeremiah Schuur, MD, an assistant professor of emergency medicine at Harvard Medical School and an attending at Brigham and Women's Hospital in Boston, said he is not opposed to measuring quality and efficiency in hospitals and emergency departments, but said he disagreed with a recent rule promulgated by the U.S. Centers for Medicare and Medicaid Services that sought to reduce the numbers of CT scans for atraumatic headache patients. He and emergency physicians at 21 hospitals sought to determine the validity of the rule and recently published the negative results in the Annals of Emergency Medicine. (2012 Feb 23. [Epub]; see FastLinks.)
"We did not undertake this project because we are against Medicare's desire to measure," he said. "I, and most of the authors of this report, do not think that is wrong. I think there's a real role for Medicare to engage in quality and efficiency measurement, but we think this is an odd place to start. If you look at the literature and guidelines, there are not a lot of good studies of this issue."
The authors pointed out that the National Quality Foundation did not endorse the measure because no scientific evidence supported it. But CMS said headache continues to be a frequent reason for ED use. (See FastLinks.) The agency noted:
- The National Hospital Ambulatory Medical Care Survey found a CT scan was done in 13.9 percent of the estimated 117 million ED visits in the United States in 2007. Forty-eight percent of those were head CTs (approximately 7.8 million).
- Physicians generally have a lower threshold for ordering neuroimaging for headache in the ED because of physician time constraints and the need to operate efficiently.
- Physicians in the ED may be inclined to use CT scans unnecessarily for the sake of time and caution.
The measure, called OP-15, is meant to be an efficiency measure with a numerator of ED visits with a primary diagnosis of headache with a coincident brain CT study (i.e., brain CT studies on the same day for the same patient) and a denominator of ED patient visits with a primary diagnosis code of headache. Excluded from the calculation are visits with a lumbar puncture, dizziness, paresthesia, subarachnoid hemorrhage, complicated or thunderclap headache, focal neurologic deficit, trauma, or CT performed with intravenous contrast, among others.
Dr. Schuur and his colleagues reviewed 767 visits that CMS identified as having inappropriate CTs. They excluded those with unavailable records or incorrect information, and found that a number of the visits had process and clinical exclusions not identified in administrative data used by CMS. "The CMS measure was 34.6 percent reliable when calculated from administrative data compared [with] record review," the authors wrote. They found that 479 of 748 patient visits had one or more clinical exclusion, and that the measure's validity was 47.5 percent in 259 patient visits with no recorded exclusions. Overall, 65 percent of the patient visits they evaluated had a documented exclusion not identified in the data used in the CMS administrative information.
"Overall measure accuracy from administrative data was approximately 17 percent, and hospital-level performance as reported by CMS did not correlate with performance as determined by record review. According to these findings, we are concerned that OP-15 is not a reliable, valid, or accurate measure of quality or efficiency," they wrote.
Dr. Schuur noted that older adults are excluded from most studies, yet are at the highest risk of bleeding and tumors. "I develop quality measures, and I'm in charge of reporting on them for my emergency department," he said. "I believe in this, but there are some basic tenets with which you must start. You have to have good evidence, and this process did not do that."
The rate of imaging has gone up dramatically in emergency departments, he said, adding that the rate of increase is above the rate of critical findings. He said one approach would be to create a measure for topics for which there is good evidence, such as pulmonary embolus and mild traumatic head injury, both of which have tools that clinicians can use to decide whether to get a head CT. Another approach would be to measure the total number of CTs and not go into the detail of whether they are appropriate or not. "I would have fewer issues with that," Dr. Schuur said. "You could measure the hospital against itself over time."
The current CMS measure also fails to consider anticoagulated patients, a problem that will only grow as the population ages, said Dennis Hanlon, MD, an associate professor of emergency medicine at Drexel University College of Medicine and Temple University School of Medicine. Often, he said, the medical staff never records that the patient is taking blood-thinning medication.
"What about the patient who falls and bumps her head? If the CT is negative, what do you with that patient?" he asked. "Neurosurgeons say you should admit them all, and rescan them automatically. Others will say that you can watch the patient, and then rescan before sending him home with good instructions."
It can depend on how anticoagulated they are, Dr. Hanlon said. "It's like a sliding scale."
He agreed that the CMS efficiency measure was ill considered, and advised looking at it more closely with scientific accuracy. "Putting out recommendations that are not evidence-based or lack appropriate cautionary factors is difficult and dangerous," Dr. Hanlon said.
Experts point to a recent Annals of Emergency Medicine study as a good way to approach the issue. Researchers led by Dr. Vicenzo G. Menditto of the Ospedali Riuniti di Ancona in Ancona, Italy, found that 24 hours of observation and a repeat CT scan of 97 patients on anticoagulants with minor head injury detected all but two cases of delayed bleeding. (Ann Emerg Med 2012 Jan 13. [Epub].)
"It's a tough area," said Dr. Hanlon. "They might come up with better guidelines by segmenting the population and having separate guidelines for the elderly, the anticoagulated, or a combination of the two."
- Comments about this article? Write to EMN at email@example.com.
- Read the Schuur et al findings for CT scans for atraumatic headache at http://bit.ly/CTheadache.
- Review the CMS PowerPoint on CT scans for headache at http://bit.ly/CMSct.
EMN Quick Poll
What do you think of the CMS rule to reduce the number of CT scans for atraumatic headache patients? Vote in our Quick Poll on www.EM-News.com.