Bedside Rounds Adam Rodman, MD, MPH, FACP

What does it mean when a computer can make better medical decisions than a human? The progress in large language models, and in particular the popularity of ChatGPT, has brought these questions to the forefront in 2023, but we’ve been discussing this for over 50 years. In this episode, Dr. Shani Herzig and I are going all the way back to the early 1970s with the invention of AAPHELP, the first real clinical decision support system, and the first time doctors had to contemplate working with – or competing against – computer systems.

What happens when a patient far from surgical care – say, at the bottom of the Pacific ocean on a submarine, or at a research base in Antarctica in the middle of the winter – develops a surgical abdomen? This dilemma was the impetus to build the first truly effective clinical decision support system – and to grapple with what it means when a computer can make better medical decisions than a doctor. In this episode, part one of three, we discuss the dramatic stories of appendectomies at Novolazarevskaya and aboard the submarine Seadragon. Also, there’s a brand new #AdamAnswers about the origin of the word “scut.”

American doctors spend the majority of their time during the day on the computer, either writing or reading notes about their patients; only a small fraction is spent with the human beings in their care. Technology itself – especially the electronic medical record – has often been blamed for this. But in this episode – a recorded grand rounds that I gave at the San Francisco VA in 2022 – I argue that this alienation has its roots in the way we’ve decided to organize clinical data, and the assumptions that we’ve made about the nature of medical care. In particular, I’m going to discuss one of the most influential medical thinkers of the second half of the 20th century, Larry Weed, his invention of the problem-oriented medical record and the SOAP note, and how his insight – that medical documentation fundamentally influences how we think about our patients – changed the way we think about our patients.

In the past episode, cultural and medical historians Lakshmi Krishnan and Mike Neuss discussed the history of the actual work of the doctor – Holmesian detective, data entry clerk, or something else altogether. In this episode, we conclude our discussion by talking about what type of metaphors are best suited for clinical work. Plus a brand new #AdamAnswers about the reason that American doctors are so obsessed with using, well, the # symbol in our notes.

What do doctors actually do? Are they Sherlockian detectives, hunting down obscure clues to solve intractable cases? Are they virtuosic experts, training for half a lifetime to bring the latest science to bear to cure disease? Or are they clerks, whose main job is to collect and enter data into the electronic health record? In this episode, Adam is joined by medical and cultural historians Lakshmi Krishnan and Mike Neuss to discuss the stories we tell about our own work – and how this often conflicts with the realities of clinical practice.

How do doctors actually think? And if we can answer that, can we train a computer to do a better job? In the post-WW2 period, a group of iconoclastic physicians set about to redefine the nature and structure of clinical reasoning and tried to build a diagnostic machine. Though they would ultimately fail, their failure set the stage for the birth of the electronic health records, formalized the review of systems, and set up a metacognitive conflict that remains unresolved to this day. This episode, entitled “The Database,” is the second part of this on the history of diagnosis with Gurpreet Dhaliwal.