Field Notes: Easy answers ≠ easy medicine.
How simple conversations make a difference--and why physicians really do have job security from AI.
On the medical boards, if you choose the right answer, the patient lives. In real life, knowing the right thing is not enough. You have to get stuff done.
Here’s an example. Ordering a STAT ultrasound to rule out an organ-threatening condition like a twisting of the ovary is not the last thing your emergency room clinician has to do. It’s the first thing. After the order gets entered, the clinician has to pick up a phone (or else physically hurry over to the radiology suite) to make sure they understand that in this case, the word “STAT” really means now. Next, someone has to get the patient over there. Then, the radiologist has to receive the study (i.e., someone has to know to expedite the images to the front of the line). Finally, the results have to be reported back to us verbally (not just updated into the computer where the information could languish). A lot has to happen to get it right. We do this all the time, but it takes effort at every step.
An easy answer is not the same thing as easy medicine.
Recently, I treated a patient with profound anemia—low red blood cells, the ones that carry oxygen. She really needed a blood transfusion. That was the easy part. What unfolded next was a lot harder, and a great example of how artificial intelligence doesn’t threaten the interesting and important parts of my job as an ER doctor.
Here’s the story of how I ethically worked with a patient who said no to a blood transfusion…
Keep reading with a 7-day free trial
Subscribe to Inside Medicine to keep reading this post and get 7 days of free access to the full post archives.