The radiology community's increasing concern about excessive radiation exposure has produced the successful Image Gently campaign for children and Image Wisely for adults. At the same time, the integration of clinical decision-support (CDS) software with computerized physician order-entry (CPOE) systems is making it easier to order the most clinically appropriate exam based on best-practice guidelines.
CDS software has been shown to do a good job of eliminating imaging orders that are medically inappropriate. But what happens when a referring physician has a choice of modalities for an exam that is medically indicated? Can learning about radiation dose and exam cost influence his or her decision?
A research team from the Uniformed Services University of the Health Sciences (USU) and the National Library of Medicine decided to find out. They conducted a study in which 112 family physicians were asked to order a diagnostic imaging procedure for a symptomatic hypothetical patient. The group consisted of physicians affiliated with family medicine at the University of Nebraska Medical Center and St. Louis University (Medical Care, April 17, 2013).
About half of those recruited to the study were U.S. Air Force physicians who treated both civilian and military-related patients. The researchers specifically tried to obtain a mix of civilian and military physicians. Interestingly, there turned out to be no difference in exam-ordering patterns between the two settings.
The participants were predominantly residents (64.3%), according to the authors. Nearly half (45.5%) were 30 years of age or younger, and 77.5% were younger than 40.
The researchers hypothesized that if physicians were presented with a number of equally medically appropriate imaging tests with corresponding radiation exposure and health risk information, they would avoid ordering tests with the highest radiation dose. Similarly, they believed the physicians would also avoid ordering the most costly procedures, wrote Ron Gimbel, PhD, interim chairman of biomedical informatics at USU, and colleagues.
Each participant was presented with a clinical vignette of a 22-year-old female patient presenting with a previously detected indeterminate renal mass. The study participants were given the choice of nine imaging options. After making a selection, the physician was shown the American College of Radiology (ACR) Appropriateness Criteria for the exam. The most appropriate exam choices were then displayed, and participants could modify their exam choice.
Next, either radiation dose exam information or cost information was displayed, followed by the other category (the participants had been randomly assigned to one of two groups: radiation then cost information, or cost then radiation information). Participants could again change their choice of exam at any of these stages.
A CT scan was the exam of choice for half (57, or 50.9%) of the 112 participants. Slightly more than one-third (35.7%) ordered an ultrasound exam, 8% ordered an MRI exam, and 5.4% ordered other tests. After being shown the ACR-recommended exams, those ordering a CT exam increased slightly to 54.5%, ultrasound shifted to 37.5%, MRI changed to 6.3%, and the "other" category moved to 1.8%.
In the group of 65 physicians presented with radiation dose information first, learning about CT exam dose caused more than half of the doctors to select another exam, and the number of CT orders dropped from 32 to 14 (21.5%). There was no additional change when the Medicare reimbursement cost of the CT exam ($916) was revealed.
In the same group, after dose information was disclosed, ultrasound orders increased from 25 to 36. And after the doctors learned that an ultrasound exam costs a fraction of the CT and MRI alternatives, an additional nine switched to this procedure, for a total of 45 (69.3%).
For MRI, orders increased from six to 15 after dose information was presented. However, when told its cost ($1,478), the number of physicians selecting MRI as a final choice dropped back to six.
In the group of 47 physicians given cost information first, more than 60% had initially favored a CT exam, increasing from 26 to 29 after reviewing ACR guidelines. However, after learning the cost and radiation of a CT exam, the number of orders dropped to 16 and then 15 (31.9%).
About the same percentage (36.2%) of the cost-first group had initially selected an ultrasound scan as the radiation-first group (38.5%). When the participants were told of ultrasound's low cost ($272), the number of orders nearly doubled, from 17 to 31. When told that the exam had no radiation exposure associated with it, the number increased to 32 (68.1%).
In the cost-first group, there was only one order for MRI after the ACR guidelines were shown. However, the physician selected another exam after learning of its cost.
Not everyone changes
The research team was intrigued that 29% of the physicians in the dose-first group did not change their initial order for CT. Neither did half of those given cost information first. A total of 22 of the 112 physicians stuck with their decision to order the exam, regardless of radiation dose or cost.
"This study supports an expanding literature which suggests that physicians have a real knowledge gap regarding both the estimated radiation exposure and cost of medical imaging," Gimbel told AuntMinnie.com. "In this study, we addressed both of these gaps after presentation of clinical evidence."
The research team is launching several follow-up randomized, controlled trials aimed at better understanding how the presentation of clinical evidence, safety, and cost information may shape physician ordering of medical image tests.
"In our view, the goal is 'optimal' ordering, where optimal medical imaging combines the best clinical evidence, the lowest patient exposure, and the lowest cost of an exam," he said. "This is especially true in clinical scenarios where the evidence supports more than one test option."
Gimbel acknowledged that the clinical setting in which a physician practices and related institutional policies could influence ordering behavior.
"However, my colleagues and I believe there is work to be done in exploring how to close the existing knowledge gap without impeding on physician autonomy," he said. "We believe that physicians want to make the best medical imaging decisions for -- and with -- their patients. They are often impeded by imperfect or missing information to guide the risk versus benefit equation in their decision-making."