Having cold water dashed in one’s face can result in spluttering, but it can be welcome when an effective wake-up call is needed. Both reactions to a 2010 article¹ abstract on patient-centered radiology might be expected; the authors deliver a sometimes-harsh assessment of radiology’s shortcomings, but they also offer concrete suggestions that could help protect the specialty from commoditization, reduced reimbursement, and an uncertain future.
Stephen J. Swensen, MD, MMM, and C. Daniel Johnson, MD, MMM, are Mayo Clinic radiologists and professors of radiology based in Rochester, Minnesota, and Scottsdale, Arizona, respectively. Their premise is that radiology is a target that cannot survive the forces converging on it unless it can become patient centered. Radiology must respond to patients’ wishes in everything that it delivers, and it must amass the data needed to prove that it does exactly that.
Radiology must grant the five primary wishes of patients: for enough information to make decisions, for the correct exams to be ordered, for those exams to be performed safely, for sound interpretations to be communicated effectively, and for the prices charged to be fair, according to Swensen and Johnson.
All of these wishes rely on a shift in values in radiology in which the needs of the patient supersede those of the radiologist (as well as those of the facility and the payor). Other kinds of change, no matter how effective they are in their spheres of influence, will not have the power to improve results. “All of our equipment and processes of care could be stellar, and we could still produce terrible results . . . values drive behaviors, and behaviors drive the results,” the authors write.
Information and Appropriateness
Swensen and Johnson cite mammography as an area where the patient’s need for information is most obvious. Despite its lifesaving importance overall, the sensitivity of mammography in breast-cancer detection varies hugely, so women considering screening need to know whether their radiologists interpret mammograms well or poorly.
Describing radiology as far from transparent, Swensen and Johnson call on radiologists to share their performance data with the public. Even if this step doesn’t affect patients’ choices strongly, it will certainly change radiologists’ performance, they add.
Sometimes, they continue, the best choice for the patient seeking not only the right radiologist but the right exam will be no exam at all. They write, “Our profession typically views overutilization as someone else’s problem. Sure, there are issues with cardiologists self-referring to their own imaging equipment, some cases of orthopedic surgeons with their own MRI units, and family physicians with chest x-ray machines in their back offices.” Nonetheless, the authors report, roughly 90% of commercial imaging spending goes to radiologists², and this makes appropriate utilization their responsibility.
Undeniably, overuse of imaging can often be traced to patients’ demands and to referring physicians’ fear of malpractice suits (and lack of knowledge about the best imaging choices). Radiologists, however, must help physicians and patients make the correct choices, and should be able and willing to stop them when they don’t.
If radiologists can’t take that step, the authors add, utilization managers will, particularly since they are already seen as effective in reducing inappropriate imaging. “Self-referral and overutilization are our professional space. Utilization management arose because we had not done our job,” Swensen and Johnson write. “Most estimates show that approximately 30% of examinations do not meet standard appropriateness criteria.”
For patients, overutilization creates more than additional expense and anxiety. Often, it leads to medical and surgical interventions that are of little to no benefit (or are harmful). For example, the authors note, back-pain patients in one study³ were 2.5 times more likely to have back surgery if they underwent MRI exams instead of radiography, but their outcomes (measured as general health, pain, and disability) were no better.
Swensen and Johnson write, “As a profession, we may cause more harm from overutilization than from any other action. We need to step up to it. A rational approach involves IT.” Computerized order entry using appropriateness standards is one example, but clinical prediction rules, developed by multispecialty teams, can also be applied.