“Humans are so prone to mistakes. Wouldn’t it be better if a patient could just type his symptoms into a computer and get a printout with the diagnosis and treatment?” I remember this argument vividly from my first year of medical school, sitting across the table from a tech-savvy classmate. I responded that patients would want something more: a human face to empathize with, a hand to shake. Obviously doctors need to be human to have humanity. “We could just program empathy into the algorithm. We could even create a human face to make the patients happy. It would be the same thing as a human doctor, only better.” I remember scoffing at the idea. Little did I know this argument would resurface a few years later, when the advent of a new technology would begin to challenge the necessity of physicians.
At the beginning of this year, IBM’s Watson Computer entered medical school. The same supercomputer that reigned as champion on Jeopardy began working with medical students and reading case histories and medical texts. This is all part of an experiment to see whether a computer can be trained into a master diagnostician, setting up a John Henry-esque showdown between human doctors and computer algorithms for primacy in healthcare. The news of this supercomputer preparing to take medicine by storm has forced me to revisit the argument from my first year of medical school: What makes a “good” physician, and why do we need anything more than a computer screen?
In searching for this answer, perhaps the first question to ask should be, “What is so appealing about a computer-doctor?” With the development of more and more complex laboratory and imaging studies, physicians are able to look into every crevice of the body, detecting the smallest abnormalities. This deluge of information can create a false sense of omniscience and unreal expectations of physicians. While it is understandable that patients would hope for robotic proficiency and perfection, especially when experiencing suffering and fear, such high hopes are sure to be disappointed. A “good” physician must be technically competent as a baseline, but he will have errors in judgment and missed diagnoses throughout his career. With technologies continuing to develop at a rapid clip, doctors will eventually lose ground to computers. If Watson can answer the most abstract questions about Shakespearian drama or world geography on Jeopardy, it is likely a matter of time before it can learn to diagnose appendicitis from clinical symptoms and lab values. A smart physician will not make diagnostics a battleground for the profession.
An equally important role of the physician, however, lies in the times after these diagnoses are made, when patients are in need of continued support. This support begins with listening. The doctor’s role is not only to decode the patient’s history and calculate a differential, but to truly listen to the story told through spoken words as well as tearful eyes and trembling fingers. Early in my medical career, I saw how easily one can regress from listening to merely calculating, especially in the midst of a hectic schedule. To be frank, medicine is simpler when the patient is merely a list of symptoms rather than a person with thoughts, motives and needs. Yet, a physician must have an understanding of the patient as a person in order to guide him towards the best medical decision for him. While Watson can provide calculations of risks and prognoses with strict confidence intervals, it cannot help a patient decide whether an extra month of life is worth the side effects of another round of chemotherapy. The “good” physician can take these sterile numbers and translate them into human choices.
Beyond words and numbers, the “good” physician has other things to offer. Oftentimes, merely being present can have a lasting effect. I still remember one early morning in the hospital while I was on a pediatric oncology rotation. As I walked past the room of a young patient who had just passed away, a physician emerged into the hallway. His eyes were red from having spent the night at the bedside with the family, waiting for the patient to reach her final hour. This tired doctor had given this patient and her family the gift of “being there” when words could not suffice. Wheeling a computer into this dying patient’s room, no matter how well programmed, could never fulfill this intimate role of humanity.
As medicine trends towards further complexity and unruly data, self-reflection will play a vital role for physicians, by helping them adapt to the rapidly changing landscape of their profession. After only a few moments inside a hospital, it becomes clear just how much modern medicine depends on technology. Watson is just the next step in this progression, a newer and smarter technology. However, something more than Watson is needed to care for the patient: empathy, compassion, the very things that make us human. In the face of these challenges, if physicians will strive to sustain and grow the humanity in medicine, then technologies such as Watson will serve as a supplement, not a replacement for future doctors. And hopefully these new tools will not discourage physicians, but rather strengthen their practice while pushing them to answer the question, “What can I offer that Watson cannot?”
Bryan Sisk, MD
Deputy Editor, The Living Hand