Think That You Might Be Wrong

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - Dana Smetherman, MD, MPH, FACR
Dana Smetherman, MD, MPH, FACR

A breast imager offers the subspecialty as a model for radiology’s transition from volume to value.

A few years ago, handmade, anonymously produced guerilla art proclaiming, “Think That You Might Be Wrong,” sprouted on telephone poles and street signs around the city where I live.  As this message seeped into our public consciousness, the placards reminded me of times when I might have misjudged another person, neglected to consider a different perspective or allowed my own biases to color my actions or opinions.  This notion also could apply to the current practice of radiology. 

For years, many radiologists have focused strongly on productivity, striving to interpret studies as efficiently as possible, our success measured largely by the number of exams we read.  As healthcare in the U.S. shifts from measuring volume to value, radiologists may need to reassess how to demonstrate our worth to patients, referring physicians and the healthcare system.

For years, breast imagers have marched to the beat of a somewhat different drummer among the radiology community.  Adherence to the structured reporting of BI-RADS delivers consistent, actionable reports to referring clinicians and allows us to benchmark our performance against national norms. 

In many practices like mine, the breast imaging section takes ownership of navigating the patient through the process of screening, diagnosis, extent of disease evaluation, surgery, response to treatment and, finally, back to surveillance.  We communicate the results of our imaging studies not only to referring clinicians, but also to patients themselves as written summaries in lay language. We recommend needle biopsies that we will perform and discuss the pathologic results of those biopsies directly with patients. 

What would it look like if all of radiology followed this breast-imaging model?  Standardized reporting might be possible but the outcomes data available in breast imaging could be more difficult to produce for other types of exams.  Breast imaging studies evaluate one organ for the presence of one disease—carcinoma.  Breast imaging quality metrics, such as sensitivity, specificity, recall rate, cancer detection rate and positive predictive value of biopsies, gauge our success in one outcome—accurately diagnosing breast cancer without generating either too many false positive or false negative interpretations.

Number of challenges

There are some imaging studies for which the anatomy and organ systems covered and the potential diagnoses are so broad that it would be difficult to generate the outcomes data available in breast imaging.  Think about an abdominal CT performed for a patient with abdominal pain.  The patient’s symptoms could arise from any organ, from the skin to the spine, and the potential etiologies are almost endless: infection, trauma, vascular lesions, congenital anomalies, neoplasm, autoimmune and inflammatory diseases and more. Other studies also may not have a readily available gold standard (such as a pathologic result from a breast biopsy) against which to benchmark one’s accuracy. 

Nonetheless, there may be pockets of radiologic practice where the clinical question, anatomic focus and potential disease entities are narrow enough to consider adopting a reporting and quality assessment model similar to that utilized in breast imaging.  For example, my colleagues and I have discussed applying a similar strategy to lung cancer screening with CT, thyroid or pelvic ultrasound, lumbar spine imaging for back pain and perhaps vascular imaging studies.  

Another challenge, of course, is the significant financial investment necessary for the infrastructure to support breast imaging database and reporting requirements.  Would our departments be able to bear the cost of implementing similar systems throughout radiology? 

Some other aspects of breast imaging practice—such as more direct interaction with patients—may be more easily translated to all of radiology.  Patient portals in electronic health records, for instance, facilitate and even necessitate communication (at least digitally) between the radiologist and patient. 

Health and Human Services Secretary Sylvia Burwell announced this year that the timeline for implementation of value and quality based initiatives of the Affordable Care Act has been accelerated.  As these changes become reality, all of radiology may need to demonstrate our quality and the value we bring to patient care by rethinking the