Classification of Repeat Imaging: Implications for the ED
In February 2013, the ACR® Harvey L. Neiman Health Policy Institute (HPI) released a brief¹ proposing a classification system for repeat medical-imaging studies. Richard Duszak, MD, CEO of HPI, says, “We wanted to propose a way to study the incidence, utilization, and appropriateness of repeat imaging for research and policy purposes. While that was our primary audience, and they would be researching this retrospectively, there is a lot of opportunity to do this in a prospective, real-time fashion for quality-improvement purposes.”
Emergency-department imaging is, Duszak says, “definitely an area that merits some further exploration.” The explosion of imaging utilization in the emergency-department setting over the past few years has been well documented; less studied is the degree to which this imaging is appropriate. “There is a lot of evidence to support the value of imaging in the acute setting. Imaging is very helpful in reducing triage time and cutting down on admissions,” Duszak says. “Therein lies the opportunity—but also some of the difficulty.”
Setting-specific Challenges
The emergency department, as a setting, is unique in health care, Duszak explains. Physicians are having one-time interactions with patients, often without the context of their clinical histories. “The emergency-department physician is busy, and even if he or she has access to an electronic health record (EHR), there may not be time to mine it manually for prior studies,” he notes. “On top of that, often, the patient isn’t in a position to provide his or her clinical history. My hunch is that if emergency-department physicians had real-time, easy access to prior imaging studies, their requests for imaging for each encounter would decrease.”
Emergency-department physicians are also particularly vulnerable to medical-liability (malpractice) suits, which might result in their adoption of a defensive approach to care. “These physicians have to make decisions based on limited information, unlike (for instance) a family practitioner who might have a longstanding relationship with the patient,” Duszak says. “As a rule, physicians in one-time, transactional relationships with patients are more vulnerable to lawsuits because there isn’t that trust and understanding there.”
Duszak says that many emergency departments are facing a high volume of patients who need expedient triage. The situation is complicated further by the fact that these patients arrive with a mixture of true emergency conditions and disorders that could be handled elsewhere in the care-delivery system.
“Unfortunately, there is a lot of volume in the emergency department that isn’t best dealt with in that setting,” he notes. “It has to do with limited private-physician office hours, access to care, and insurance: Those factors tend to siphon patients into the emergency department, and therein lies one of the reasons imaging is used so widely. The system is mixing patients with acute disorders with patients who could be managed in an office setting much more cost effectively, and imaging has been extremely good at helping to differentiate between those—to make sure everyone gets the best care.”
Data-driven Dialogue
More data are needed to support the dialogue surrounding emergency-department imaging, Duszak says. “We are moving into an environment in health care where information is king, and that requires us to be very thoughtful in our analyses,” he says. “Big data can be good or bad, depending on whether they are used intelligently. If you’re simply mining the data to say that this patient had two tests within two days, so the second test must have been wasteful, that’s overly simplistic.”
Computerized decision support, better health-record integration, and informatics support for emergency-department physicians that will allow them to mine EHR data efficiently would all be likely to have an impact on imaging utilization in the emergency-department setting, Duszak says. The issue, however, needs to be examined “with a microscope, not from the 30,000-foot perspective,” he adds.
He points to a 2012 study2, which he coauthored, on the utilization of CT exams of the abdomen/pelvis in the emergency department for patients with complex conditions. “We wanted to use real data to show that in some situations, more imaging is appropriate,” he explains. “We have to be careful and thoughtful in our approach to this. If emergency-department imaging is up, is that good or bad? The answer, of course, is that it depends.”
Going forward, Duszak and his colleagues at the HPI hope that their proposed classification system will make a more nuanced discussion of imaging utilization possible across multiple settings. “You don’t want to bring a chainsaw to a problem that should be addressed with a scalpel,” he says. “The need for advanced medical imaging is higher in acutely ill patients. Screening mammography is another example: There are areas where people of lower socioeconomic status are actually not getting enough repeat testing. There are people who do need more imaging, and there are also opportunities to reduce low-utility imaging in the emergency department. The overall message is not to throw the baby out with the bathwater.”Cat Vasko is editor of Medical Imaging Review and associate editor of Radiology Business Journal.