Applied Radiology

Dr. Mirvis is the Editor-in-Chief of this journal and a Professor of Radiology, Diagnostic Imaging Department, University of Maryland School of Medicine, Baltimore, MD.

Why isn't it mandatory within the certification requirements of The Joint Commission that requests for all diagnostic studies be accompanied by the current clinical signs and symptoms that prompted the study to be obtained? There are a zillion (or some other really big number) rules about how far boxes need to be from automatic sprinklers, how bottled gases need to be arranged, and so on. What is it about requiring meaningful clinical information on imaging tests that makes it forbidden territory for mandatory oversight?

Picture yourself as an internist (nothing personal against internists; some of my best friends are internists) who is seeing a patient with abdominal pain. You have several potential approaches to deal with the situation.

• Interview and examine the patient. Assess vital signs while learning the quality, location, frequency and duration of the pain. You may ask about potentially pertinent medical history, perhaps obtain some basic blood work, auscultate for bowel sound and test for guarding, rebound, etc. (all that medical school stuff). Next, you assemble all that data and formulate a differential diagnosis. Then you decide if further testing is needed to establish or eliminate the most likely diagnosis. Next, request the imaging study, providing keypositive and negative information or consult with the radiologist (you always wish you had gone into radiology) to discuss the patient's findings and arrive at the most efficient imaging work-up, if needed.
• After performing a history and physical you request the study that you believe is most appropriate, providing minimal clinical information on the request, i.e. "abdominal pain."
• After performing a history and physical, you request the study you believe is needed and provide "rule out" appendicitis as the indication. This diagnosis may be your leading diagnostic consideration or just something to exclude because it's serious. The clinical assessment may not point that way at all, however, the diagnostic imager will focus on assessing the appendix and perhaps spend less time and attention on other areas. That's a human response.

By the way, why is it that someone would want to "rule out" their primary diagnosis? I would think you would want to verify it. Having it ruled out all the time would force you to question your diagnostic skills. But I digress.

• You take the symptoms off the triage nurse's notes, glance at the patient from the foot of the bed to verify life and request multiple studies to cover any potential abdominal pathology. The order in which these studies are performed is by luck of the draw. Have the nurse write "pain" in the clinical information box on request.
• Never see the patient, have the nurse order tests for whatever body part hurts and one joint above and below, and offer "other" by way of clinical history. These are the patients whose primary physicians can never be found and everyone else you reach denies knowing or wanting to know of the patient's existence. Sometimes this circumstance requires you to write a note about a critical finding and tape it to the patient's chest. You can also mention to the patient that they have an acute aortic dissection on the off chance they will remember to mention this to any medical personnel who happen by.

Of course the entire scenario described above is hyperbole. I hope.

So why is it that expensive imaging studies, which require patients to be injected with contrast materials-or swallow obnoxious liquids or perhaps even get these liquids instilled from the opposite direction-in preparation to receive a walloping dose of ionizing radiation onthe order of a smallish solar-mass ejection get requested with missing or misleading information? Performing these studies is serious business and should be requested with some consideration for cost, appropriateness and the risk-benefit ratio. Of course, pertinent and factual clinical information should be provided to the interpreting physician. Some nonradiologists interpret studies quite well, but I would never bet on one over a board-certified radiologist, not for any patient I cared about.

When the topic of useful clinical information came up recently with some visiting radiology residents to my program (of course no resident from my institution was present) I heard some interesting examples of the less than in-depth clinical information provided on requests. Here are just a few:

• Feeling poorly.
• R/O fever.
• GI attending wants study.
• Pain (total-body CT requested).
• Over-heated.
• Patient has lipoma on back. Patient requests metastatic work-up.
• Patient stubbed toe (total-body radiographic survey requested).
• Bumped knee on fire hydrant, total-body radiographs requested. (I am glad they included the fire hydrant.)
• History of epilepsy, R/O pneumonia.

I'm certain the reader can supply many more such examples. Please send them my way as I am trying to build a world-class collection. Information technologists tell me that newer order-entry systems require clinical information before an imaging study is "ordered." Fine, except someone who actually knows the patient should enter that information. It must be accurate and, even better, should ask a specific question for the study to resolve. One begins to doubt concurrent requests on 30 intensive care unit patients who all have "atelectasis" for their clinical information. All such foolproof systems have workarounds anyway. It's sad that such measures are needed.

OK, now all you hospital regulators and risk managers out there get cracking on this problem! We have all seen more than a few imaging study interpretations messed up as a result of missing or misleading indications. If the study is important enough to perform, it's important enough to be requested correctly. After all, we're here to help the patient, not challenge the radiologist to read shadows in a clinical vacuum.

By the way, it might not be such a bad idea to leave this editorial lying around where your referring docs just might happen to run acrossit.