Accuracy: A Moving Target in Radiology

I wrote last week about traveling to Iceland for a semi-successful viewing of the solar eclipse. While it probably won’t be my last visit, I will wait until the place gets a little less popular/pricey. One can easily drop $40 on a sandwich. A friend who went just a couple of years ago insists that this is a drastic change.

No matter where my voyages take me, sooner or later I find myself thinking about what it would be like to live there. I am not looking to flee my home or even crafting doomsday scenarios in which I would be forced to do so. It is just a flight of fancy.

Sitting on the tour bus, I had more than my usual amount of time to dwell on the matter and imagined what it might be like to spend the latter portion of my radiology career just shy of the Arctic Circle. The whole point of teleradiology is being able to read cases remotely. I would be far from the first to switch countries while doing it.

Of course, laws and regulations have a few things to say about that. Certain telerad setups are restricted or outright disallowed, sometimes for good reason. For the purpose of my relocation daydreams, I don’t get bogged down with such details.

Since one of our guides talked about the local health-care system (a public/private hybrid), and more docs are needed pretty much everywhere, I imagined some sort of deal wherein an American expat would be allowed to do his telerad thing if, for example, he was willing to spend a certain percentage of his time working for the Icelandic system.

This flowed into a path of thought I have trodden in various ways over the years: Suppose I am reading a 50-50 split of telerad cases. Half is for USA clients I have been serving for the long haul and half for some other country. I will clearly be able to tell which studies come from where. How likely is it that I will treat them all the same way?

Outsiders or newbies to the field might think the question is preposterous, even disturbing. Ivory-tower academicians might condemn me for even asking. Of course, they should all be handled the same way. Radiology, like any other scientific field, has right and wrong answers. Gray/unknown areas have established protocols (ACR white-paper guidelines for pancreatic cysts, BI-RADS category 3).

Ask anybody dwelling in health care’s “real world,” though, and the waters muddy quickly. A radiologist’s reading habits tend to change with various peers, supervisory types, and, yes, lawyers looking over his or her shoulder.

A Lebanese med school mate once explained to me that the attitude toward bad medical outcomes was much different where he came from. In the United States, where docs and their hefty malpractice coverage are very juicy targets, if things go wrong, the knee-jerk response is often “This shouldn’t have happened. Who can I sue?” Back home, he recalled, the same situations would often be apologetically dismissed with an “InshaAllah.” The bad outcome was beyond our mortal control.

Suppose you start in an environment where non-egregious medical errors are blameless misfortunes. You move to one where everything that goes wrong spawns an inquisition. You might adopt behaviors that you think, rightly or wrongly, will protect you. In other words, defensive medicine.

Now, suppose you make the opposite move. You go from a litigious Manhattan-area career to some rural or third-world environment where they are just grateful to have you and probably paying you less, so you are less of a juicy medmal target. It might take you longer to let down your defenses, but eventually your reports will probably have less fluff and defensive hedging.

This doesn’t have to take months or years. I saw it happening as early as my residency training, and I imagine most rads reading this did too. Attending radiologists overseeing us had widely differing styles. They ranged from nitpicking perfectionists to real-world pragmatists. If you knew who would be reading out your cases at the end of your shift, you might very well tailor your preliminary reads to suit them.

The perfectionist, for instance, might hassle you over every tiny ditzel you didn’t specifically address. Meanwhile, if you did precisely that, and presented the pragmatist with paragraphs of incidental findings for an otherwise negative trauma CT, he or she might be irritated with you for wasting his or her time.

They wouldn’t tell you they’re being subjective and quirky. They both think they are reading cases accurately, the way one should. One might conclude that an informal part of your training is to learn that everybody has got his or her own metrics of accuracy. After their tutelage, your sense of accuracy is likely to be an average of theirs, perhaps more heavily weighted to the ones you respected and identified with the most.

Now multiply that handful of attendings into dozens of peers, the medicolegal system at large, and however your local government oversees health care. If you don’t adapt to that, you will face routine hassles (or legal, even disciplinary action) every time your concept of accuracy clashes with theirs.

Imagine, for instance, you start working someplace where the powers that be have decided that BI-RADS 3 isn’t a thing. However, they have decided to justify themselves. Their rule is that tissue isn’t “probably benign.” It’s either benign or malignant, and your job as a radiologist is to make your best prediction.

Everyone who has been working in their system is accustomed to it, especially rads who were trained there. It doesn’t matter how well you arm yourself with scholarly write-ups or how energetically you fight City Hall. You are either going to do things their way or you are going to stop doing breast imaging (voluntarily or when they punish you by revoking those credentials).

At least under such circumstances, accuracy is in the eye of the beholder, and subject to a less often referenced variant of the Golden Rule. He or she who has the gold makes the rules.