Improving relationships between radiologists and referring physicians

Specialists are using clear communication, improved radiology reports, CDS and more to satisfy the physicians they depend on.

The past few years have brought advancements and enhancements in both imaging technologies and patient care, but something else is also changing in radiology: the relationship between radiologists and referring physicians. 

It’s a dynamic that has had its ups and downs, but healthcare’s recent shift toward value-based care has made leaders pay closer attention to the satisfaction of referring physicians they do business with than ever before. Does physician satisfaction translate into patient satisfaction? It very well could. When physicians believe their patients are receiving high-quality care from a radiologist, they’ll go to him or her again and again. If they don’t think their patients are being taken care of? They’ll simply go elsewhere. There’s always another radiologist or group ready to read that urgent examination. 

So what still needs to be done to make the radiologist-referring physician relationship stronger? And what role does clinical decision support (CDS) technology play in all of this?   

Knowing What Referring Physicians Want

One of the primary things referring physicians crave is more open lines of communication with radiologists. The rationale is twofold: it ensures a more positive imaging experience for patients in light of the ever-growing importance of patient satisfaction, and it helps them obtain from these specialists more detailed information and insight that will help them determine how to move ahead with individual patients’ course of treatment. 

“Physicians are looking for discrete sets of data as to what findings they should act on immediately as well as a sense of severity,” says Tessa Cook, MD, PhD, assistant professor of radiology in the Perelman School of Medicine at the University of Pennsylvania in Philadelphia. “Can this wait one month? Six months? A year? Or is it cancer and a treatment needs to be plugged in sooner than later? They want concrete recommendations.”

William B. Morrison, MD, professor of radiology and director of the division of musculoskeletal radiology at Thomas Jefferson University Hospital in Philadelphia, agrees with Cook, noting that radiologists aren’t always as quick with information as they need to be. “There’s a wall between referring physicians and radiologists, and it has to come down,” he says. “If we are to all work for the benefit of our patients, especially at a time when value and patient satisfaction are very much at the forefront, there cannot be communications barriers, and we need to give referring physicians more than just reports.”

Referring physicians don’t always want the same information from radiologists, Morrison adds. It often depends on the physician’s own area of specialization, so radiologists must consider who, exactly, ordered the examination. In the case of a foot infection, for instance, a primary care physician would typically ask a straightforward “yes” or “no” question about if the patient’s bone is infected. By contrast, an orthopedic surgeon or podiatrist would inquire about the extent of both bone and soft-tissue infection.

In addition, Morrison says there is a strong desire among referring physicians for assistance in  determining the best and most appropriate imaging modality or technique for answering clinical questions. Returning to the example of an infected foot, he notes the mere presence or absence of infection could be confirmed using a bone scan or non-contrast MRI. The question pertaining to the extent of the infection, however, would only be answerable using MRI (optimally with contrast depending on the patient’s creatinine clearance).

Referring physicians also want radiologists to improve when it comes to helping answer patient questions, especially now that patients are beginning to access their own reports. The University of Pennsylvania developed a software prototype that allows knee MRI reports shared with patients to be annotated with definitions, illustrations, and hyperlinks to online resources. Cook says patient satisfaction and their overall understanding of reports improved once the annotations became available. 

At Stanford University Medical Center in Stanford, Calif., radiologists are charged with working together as a unified group and engaging in a consistently high level of information-sharing with referring physicians. This ensures that referring physicians receive the same high caliber of service and detailed answers to their questions, no matter which member of the radiology team they encounter. Coaching and teaching are offered to radiologists whose abilities in this area are not quite up to par, notes David B. Larson, MD, MBA, associate professor of pediatric radiology and associate chair of the performance department at Stanford University School of Medicine.

“Referring physicians—especially primary care providers and emergency  medicine specialists—really want to rest assured that they’re going to have the same experience with one radiologist in a radiology department or private practice, as with that radiologist’s colleagues,” Larson says. “It’s part of their perception of value.”

Improved Reports Gain Ground

While referring physicians appreciate better communication with radiologists and more comprehensive answers to their questions, they also seek a higher caliber of radiology reports. “Referring physicians have traditionally pointed to a lot of problems with reports, ranging from the use of overly technical and/or confusing language and excessive recommendations for follow-up imaging to typographical errors and failing to answer clinical questions,” says Andrew J. Gunn, MD, assistant program director of the diagnostic radiology residence program and assistant professor of the interventional radiology section at the University of Alabama School of Medicine in Birmingham, Ala. “The extent of common problems also extends to notifications of report content, with complaints from referring physicians that radiologists don’t contact them about important findings cited” in the text.

Gunn notes that eliciting structured feedback from referring physicians, using a formal peer-reviewed process, can play a pivotal role in improving radiology reporting practices as well as pave the way for better communication and a stronger sense of camaraderie among parties from both camps. A 2013 American Journal of Roentgenology study co-authored by Gunn bears out the validity of these assertions and illustrates that referring physicians may not be entirely satisfied with the reports being provided by the radiology community (AJR Am J Roentgenol. 2013 Oct;201(4):853-7). 

For the study, five primary care providers assessed reports from abdominal CT, chest CT, brain MRI, and abdominal ultrasound exams. Only reports of imaging exams with indications of abdominal pain, shortness of breath, or headache were included in the assessment; reports from normal and follow-up exams were disqualified. A total of 48 randomly selected reports—12 from each category of indication—were distributed to the five physicians, along with the clinical scenario and a structured evaluation form for each exam.

Subjects rated the usefulness of the reports, using a scale of 1 (“not useful”) to 5 (“very useful”). While the reports themselves were deemed “clinically useful” (average, 3.8), allowing for “good confidence in decision-making” (average, 3.7), 15 percent were said to contain unclear language; 12.1 percent, typographical errors; and 7.9, to be lacking answers to the clinical question posed.  Of the nearly two-thirds (64.6 percent) of reports that did not contain recommendations for further diagnosis or treatment, 31 percent, study participants said, should have included such information. Subjects also deemed 31.2 percent of reports unclear as to whether results contained within them had been directly conveyed to the ordering physician, despite the fact that such sharing was warranted given the findings.

Gunn says it behooves all radiology departments and practices to seek referring physicians’ feedback in a quest to build more functional, actionable reports. For his part, he is “hard at work” attempting to collect data and create pilot initiatives to get the job done.

Meanwhile, the University of Pennsylvania Health System has implemented a report tracking engine designed to address the need for reliable monitoring systems to keep poor reporting from delaying the diagnosis of harmful cancers. Known as the Automated Radiology Recommendation-Tracking Engine, or ARRTE, the software helps ensure high-quality radiology reporting by monitoring radiology report compliance with structured reporting templates and follow-up recommendations. A compliance-tracking module searches each report generated to ensure that required organs are coded and that a modality and timing for imaging follow-up are specified for indeterminate lesions. Reports are flagged and reviewed by a board-certified radiologist if one or more elements of the structured template are found to be absent or duplicated. If a radiologist’s report is flagged, he or she then receives an email notification requesting an addendum with updated information. 

In a 2017 study published by the Journal of the American College of Radiology, Cook and her co-authors assessed the impact ARRTE had on a health system in 2013 (J Am Coll Radiol. 2017 May;14(5):629-636). The team tracked two types of report compliance: first-pass (reports that were compliant immediately upon finalization by a faculty member)and subsequent (reports that became compliant once an addendum had been issued). First-pass compliance increased from more than 48 percent in July 2013 to more than 72 percent the following month, after the practice of automated email notifications alerting radiologists and trainees about non-compliant reports was first introduced. First-pass compliance rates rose higher, to more than 90 percent, once the use of ARRTE was tied to a financial incentive plan.

Additionally, a gradual decrease in first-pass compliance occurred between December 2013 and April 2014, when email notifications about non-compliance were suspended, and rose again once the notification step was reinstated. Since that time, according to the study, first-pass compliance has averaged 93 percent per month, with a margin of 3.5 percent. Subsequent compliance consistently exceeds 99 percent. On average, the authors write, “fewer than 10 cases a day require manual review for compliance.”

Does Imaging CDS Help or Hurt These Relationships?

Another key factor in the relationship between radiologists and referring physicians is the growing use of imaging CDS systems throughout the United States. Even though the CMS mandate for CDS use when ordering advanced imaging was recently delayed one full calendar year to Jan. 1, 2019, the technology is becoming more and more common by both radiology departments and private practices. What kind of impact will such technology-aided ordering have on radiologists and referring physicians? 

 “In an ideal world, radiologists would be able to pick up the phone and discuss appropriate imaging modalities with referrers, but given the volume of imaging we have to do, that’s impossible,” says Arun Krishnaraj, MD, MPH, associate professor of radiology, vice chair of quality and safety, director of the division of body imaging, chief of the abdominal imaging section, and co-director of body procedures at the University of Virginia Health System in Charlottesville, Va. “Through [imaging] CDS, referrers can be educated on which studies to order, and why—and the more educated they are, the stronger the relationships with the radiology community will be.”

Some specialists aren’t so sure CDS will have that sort of impact. Gunn, for example, says it could end up making the tension worse instead of improving communication. “If it’s used as a substitute for radiologist-referring physician communication and recommendation-sharing … it could become another layer that separates us from those who refer cases to us,” he says. “That could lead to the opposite of the relationships we’re trying to form and keep.”

Overall, there are a number of catalysts that can, and will, influence the radiologist-referring physician relationships of tomorrow. Exactly what such relationships will resemble in another 5 or 10 years is a mystery, but radiologists only have one realistic option: keep pushing ahead and provide the best patient care they can. “The only thing we can do for our good, their good, and the good of patients is keep working,” Gunn says.