I had an interesting interaction with an ER physician recently. A scan came up on my worklist with the history “R/O stone”. When I reviewed the images, it was apparent that the patient had neither a stone nor an obstructed ureter, on the other hand, there were a number of dilated bowel loops with air-fluid levels, suggesting that the patient had a small bowel obstruction. I called the referring physician. I have spoken with him many times before and he has always struck me as being a logical and practical physician with a keen analytical sense. This time, though he made my flesh creep! When I asked him which side he thought the patient’s kidney stone was on, he cheerfully replied, “Oh, I don’t think he has a stone at all, I think he has a small bowel obstruction!” I muttered back, “ Oh, that’s strange, because the history that came across said “ Rule out stone”. He laughed and responded without batting an eyelid, “ Oh that’s because that’s the only way I can order a scan without oral contrast at night and get it approved without having to move heaven and earth. Our radiologists insist on giving every patient oral contrast, and in patients who are throwing up because they’re obstructed that can be a challenge, also it prolongs the wait time by two hours, so I’ve decided that the only way I can get around this is to lie!!!”
Convenient, I thought to myself, except that it puts the radiologist in an awkward position, one where incorrect information is provided because of a protocol-driven technicality. It means that one is reading a scan not just in a vacuum, but in a Black hole where the chances of arriving at a correct diagnosis are correspondingly lowered.
Of course, I didn’t say any of this aloud. What I did say is “ Dr., I wish there was some kind of code you could use to let me know what you really think the patient has, and still keep it secret from the in-house radiologists!” He laughed again and agreed to think about, although I wasn’t really convinced.
Another time, I was reading a chest CT which came across with a history of chest pain. The scan showed a hemopneumothorax and multiple rib fractures. When I called to speak with the ER physician, the response from him was “ Ooops, sorry, I forgot to mention that the patient fell off a 2nd floor window ledge” Hmmmm.
The issue that this brings up is that of how important it is to have relevant and detailed clinical information in order to interpret scans accurately. When one looks at a hand radiograph, for example it makes a big difference to the interpreting radiologist whether the history that is provided is simply “trauma”, or whether it is “injury to the third digit with soft tissue laceration and possible foreign body”.
My colleague Dr Aschkenasi has this to say about a study he recently reported : “ When I called the surgical resident, I discovered many historical surprises which explained many unexpected findings on the exam. I think about three times in the conversation the surgeon said: “Oh yeah, and were you aware that the patient had thus-and-such two weeks ago?.” Of course, my answer was “No,” because the history just said “Abd pain and nausea.”
It seems hard to believe that a clinician would not realize that it is important for the radiologist to know in a patient who underwent a CT for right lower quadrant pain whether the appendix is in or out, and whether the patient with free air on an abdominal CT had a recent laparotomy or paracentesis.
When a large amount of money is being spent on performing an expensive investigation (not to mention radiation exposure and contrast) with the goal of getting the correct answer, it would seem logical that a tiny bit of time and effort should go into also providing relevant history to the interpreting radiologist, simply to increase the odds of the answer being correct…
Is it just that the ER physician is too busy? Perhaps so..
Every time I visit one of our client hospitals and meet with a radiologist colleague, I bemoan this issue, asking for their support in getting their clinicians to give us more history. And I typically get (a variation of) the same response, which is “ Hey, guess what – we face exactly the same problem right here in the hospital – maybe YOU can help US out!”
And so of course, this is a problem not just related to teleradiology, as the teleradiology-bashers would have us believe, but rather a systemic issue that needs to be tackled across the entire healthcare enterprise. My hope is that the IHE initiatives will help in this regard, by creating seamless interfaces between radiology, pathology and clinical information systems so that relevant information can be conveniently obtained by conscientious radiologists, improving their reporting accuracy and providing them peace of mind of a job well done.