November 2013

On Common Sense in Radiology
By Stephen J. Gordin, MD
Radiology Today
Vol. 14 No. 11 P. 12

Like many of you, I receive copies of a few printed journals each month, and I scroll through a couple online as well. Despite this plethora of knowledge, there may be three or four actually useful articles that help me in my practice. The rest are useless, scary, or dangerous.

The useless have titles such as “The Use of 5-Hydroxy-Propyl-Octo-Pepto-Bismol Imaging in Albino Rats With Radon Exposure at 7T: A Meta-Analysis of Multicenter Trials.” Trust me, I get it. We have to do basic research if for nothing else than to provide jobs for people. But in an era of cost cutting, I have my doubts that 7T imaging of anyone will ever develop into a pragmatic approach for evaluating anything. However, no one suffers either—except for the rats.

The scary articles are those that create impossible or impractical standards. For example, when Dr Big Name at Big Name University points out a 3-mm lung lesion or low-density renal lesion—things we see 10 times a day—that he or she biopsied and discovered cancer, I pull the covers over my head. Sometimes, the lesion looks round, smooth, nonenhancing, and fluid dense, but then the researcher found cancer anyway.

I think I can save a lot of time and money by reminding my academic colleagues that any finding could be something bad. Yet how is this research helpful to patients and our referring physicians if we recommend performing a biopsy on every finding? Besides, if no one knows that a lesion is malignant, what is its biologic danger? We could follow up every abnormality forever, but then we expose patients to radiation levels that rival Hiroshima survivors and to considerable expense. We have authors who pay lip service to this concern, but evidently that’s where it stops.

The dangerous articles arise when some claim to have developed the new standard in evaluating certain pathology. CT angiography (CTA) of the chest, for example, was touted as extremely accurate with high sensitivity and high specificity. Emergency physicians took to ordering them like chest X-rays, not realizing or caring about the massive amount of patient radiation exposure and the costs. The lay press took to promoting it as well, frightening people with the dangers of pulmonary emboli and extolling this great new test to detect them. Yet the emergency physicians couldn’t understand why we were often more wishy-washy than we were on ventilation/perfusion (V/Q) scans. We explained that what many researchers evidently didn’t take into account was the fact that real people would have to be scanned by real techs, and we didn’t have the latest winning number in the multidetector scanner lottery.

In my practice, 300-plus-lb women who were 5-foot-2 and had DD-size breasts were scanned by techs who couldn’t always figure out the timing of the bolus and how to adjust for scanning factors. However, our clinical colleagues had the faulty notion that they would be getting the “real” answer for the first time ever. And truthfully, when articles claim that we can see emboli in fifth-order vessels, I understand why they felt that way.

CTA is a good test but not as good as some claim. And the cheaper cost and lower radiation dose from V/Q scans make them a good choice in many situations. In other words, our focus should be on the best manner to work up the patient, not the best theoretical test to evaluate a pathology.

In my opinion, we should change the entire direction of our academia. The Affordable Care Act will demand more from radiologists as far as being consultants for our referring colleagues and giving them guidance on the most efficient way of working up their patients. I already get such calls, as I am sure most of you do. It doesn’t help referrers if I quote an arcane article about rat intestines as if it’s somehow relevant to their patient with rectal bleeding. As a profession, we have to concentrate on making imaging efficient and practical. Those of us who deal in busy private practices know we can’t be afraid of making a call on a case.

How should we handle a 3-mm lesion in the kidney of a 65-year-old with multiple renal cysts? Forget it. Will we miss a cancer? Perhaps, but how many thousands of people are we willing to put through the system to find it? And that’s just in our practice. When we think about the millions of folks already in the system and the many more coming into it, we could bankrupt the country just by looking for a handful of cancers.

Imaging has gotten so good that we’re seeing pathology so subtle that we have no idea whether it will ever hurt us. Yet articles give fodder to plaintiffs’ attorneys and make us scared of every shadow, including our own. In response, we feel forced to make these calls in our reports. We see standards being developed under controlled, sterile conditions when, in fact, we should be testing them out in the field with real radiologists and real techs performing exams on real people. Only then will we really know how good—and how practical—the new imaging technology is. Most importantly, we need to counsel and promote the use of common sense in our every day practice.

I understand that vested interest likely will continue to look the other way for a while, but I think that those who figure it out sooner will be ahead of the game. And the sooner we front-line private practice radiologists get that cover from our academic brethren, the better off and more efficient our profession can become.

— Stephen J. Gordin, MD, is a radiologist in Campobello, South Carolina, with 19 years’ experience in private practice. Outside of practice, he is a writer and musician.