Say What You Mean or Else..
Data mining radiology reports may result in coffee-break interventions.
Today is the last day that I'm using wordsby Madonna in "Bedtime Story"
They've gone out, lost their meaning
Don't function anymore
Feeling intrepid, I set out to see if there was truth to the claim that "clinical correlation" is a commonly used phrase, at least within my own department.
Donning a jaunty deerstalker hat, I performed a search for the phrase "clinical correlation" as far back as possible on our data-mining software. As far back as possible turned out to be through the end of 2009. I left out "required," "suggested," "recommended," "needed," and "advised" (i.e., words that are often combined with "clinical correlation"). Out of 88,042 studies read through early September 2015, 32 of my reports contained some variant of "clinical correlation." That works out to 0.036 percent, or 1 in every 2,751 reports.
I don't know what you consider the threshold value to be for "commonly used," but I think we can agree that less than 0.05 percent qualifies as rare. It shouldn't be too hard for me to produce zero reports that include the phrase-that-shall-not-be-mentioned from now on. There's a new performance target I can set for myself.
How about the rest of my department? Almost every staff member has fewer than 100 reports containing "clinical correlation" over the same time period. A few people have more than 100, less than 1 percent of their report volumes. Only five are bona fide outliers, with the number of guilty-as-charged reports ranging from 624 to 1,492. The individual with the highest rate uses the verboten verbiage 5.7 percent of the time or, looking at the flip side, does not use the forbidden phrase in 94.3 percent of reports — infrequently, but not commonly.
How often do you use or see "clinical correlation" in reports in your practice? Maybe someone in your group is a rogue radiologist recklessly repeating wrong remarks. Kidding and alliteration aside, we can all look at our own reports critically with respect to choices in phrasing and commit to changing them. But how do we get our radiology colleagues to replace phrasing that, admit it, none of us really likes, but every now and then, it just kind of comes out of our mouths, especially when we say it only rarely? Go ahead, one last time, a little guilty pleasure. Say it! You know you want to. I just did.
Now, having data in hand, I will have a chat over a cup of coffee with an incredibly talented colleague who says this one itty-bitty phrase more than the rest of us. And I will definitely choose a word other than "meaningless" as I try to gently persuade this colleague to use "clinical correlation" less, if at all. Because words do matter. I especially like the phrase "with respect and kindness," which is how, dear fellow radiologists, we should try to treat each other during these times of great change, including revisiting what we say (and what we should avoid) in our reports.
What do you think? Are you more likely to change your reporting behavior based on journal articles advising you to do so or do you find other means more compelling (such as a one-on-one conversation with a trusted colleague, a consensus statement from a professional society, or feedback from referring providers)? Have you ever changed your reporting behavior, such as eliminating "clinical correlation"? How can JACR engage with you, its valued readers, to help facilitate positive change in radiology reporting language?
Very inventive. I love the whimsical style. Many hearty congratulations on starting something new for JACR