It’s time for radiologists to stop hedging


There’s a simple way to define value. Ask why we exist.

Imaging exists because clinicians are uncomfortable with uncertainty. Imaging exists because emergency physicians feel that being 98 percent correct about the absence of pulmonary embolism is not good enough. Radiologists exist because imaging is not an assay on a Western blot with a 100 percent accuracy. Radiologists exist because information is imperfect, and clinicians do not like the imperfection.

This means that when reading a CT scan for acute appendicitis if, instead of decreasing the uncertainty of its absence, we increase it, we are doing referring clinicians a disservice. When we say we “cannot rule out minimal early sub-clinical tip appendicitis,” what we are really say is “we do not wish to share your risk. You’re on your own Jose.”

When we produce a litany of differentials and disclaimers in our reports, so that the clinician, who gave us a contract to reduce uncertainty, is even more uncertain, we are like Tony Montana from Scarface, who took money for a kill and then changed his mind. By the way, he dies in the movie.

The reason that the emergency physician orders a CTPA is because she cannot ruled out a pulmonary embolism without it. She does not enlist our help to hear this philosophical truism repeated. So when we tell her that “small isolated subsegmental pulmonary embolus cannot be entirely excluded with absolute certainty” we are having our cake and eating it. That is we are taking money to stick our neck out, but we are not really sticking our neck out.

If we are uncomfortable in sticking our neck out we shouldn’t have done the CT. It’s like a builder charging for time and material to build a roof, having a play around with the material and then saying it’s the roof that can’t be built. Why didn’t you say so in the first place? Maybe I wouldn’t have wasted money on the material and your time.

Of course, there will be occasions when we should say “cannot rule out.” These should be few and far between. Such occasions cannot be precisely defined. Guidelines cannot stipulate the circumstances in which “rule out” should or should not be used. We will have to defer to judgment.  Sorry, judgment can’t be specified. If it could, it wouldn’t be called judgment.

Of course, there’s risk of being sued when we use our judgment. A nurse documents “MD informed” because there’s merit in our being informed. It’s because we are physicians. It’s because we can be sued. We can be sued because people pay attention to what we say. This is the Code of Hammurabi.

Want a safe harbor? Want to be guarded by having process defined to the letter? Want our jobs to be reduced to a 10,000-page user manual? Want thresholds to define every pathology, incontrovertibly? Sure, go for it. In the short run, there will be security. Lots of security. In the long run, we will be replaced by people who work faster, for less and are more compliant. They can never be sued. Hurrah!

There will always be a hamster willing to run faster on the wheel for less money with a larger risus sardonicus on the face “what can I do to deliver outstanding service today? Have a nice day?” Don’t compete with hamsters. They will win.

Judgment can be neither commoditized nor outsourced. Judgment means risk. Risk means value. Refusal to accept risk means anti-value. Actually, it means worse. It means not doing our job.

So fellow radiologists let’s make a pledge not to hedge. Let’s have an international no hedge day on July 3rd — the day radiologists emancipated themselves from the perpetual fear of being occasionally wrong.


Saurabh Jha is a radiologist and can be reached on Twitter @RogueRad. This article originally appeared in ACR Bulletin.

Image credit:


View 10 Comments >

Most Popular

✓ Join 150,000+ subscribers
✓ Get KevinMD's most popular stories