Credit Tom Fitzsimmons
My favorite moment in residency was hearing about my friend’s favorite moment in residency.
Excited to have finished our first year as doctors — the most unsettling and demanding in medical training — we swapped stories of our most memorable experiences. He told me about his final evening in the intensive care unit, the end of a grueling month. He had made plans to unwind with a friend from out of town whom he hadn’t seen in years. Before leaving, he checked in on a patient he’d admitted the previous week, an elderly woman with dementia and a bad pneumonia that was getting worse. Her oxygen levels had now dipped so low he feared her heart would stop beating.
He’d grown close to the patient’s daughter, who had been at her mother’s bedside each day. The daughter had been struggling with whether her mother should be intubated if her condition deteriorated, which seemed likely. My friend told her he was transitioning off service and would be leaving for the night, but assured her that the next team would take good care of her mother.
She hesitated. Her brothers and sisters were flying in from around the country that night, she told him. They planned to discuss their mother’s life, her condition, her wishes moving forward. And she wanted him — the junior member of the team — to lead the family discussion.
He called his friend. He wouldn’t be making dinner. Then, he had his most meaningful conversation as a doctor.
As he gathered with the family, they told stories of who their mother was, what had been most important to her, and how she would want to die. Ultimately, they decided against intubation and focused on keeping her comfortable in her final days.
What struck me about my friend’s story is not only that he acted as an exemplary physician and helped his patient die a dignified death. It’s that it was important that he was the one having that conversation.
Too often in medicine, you feel like part of a machine, a cog in a massive bureaucracy. We cover each other’s shifts, we maintain a hospital’s patient flow — and at the end of many days, you feel nothing would have been different if another doctor subbed in.
This isn’t necessarily a bad thing. I don’t want a patient to fare differently simply because I’m on call rather than my friend. Much of medical training is an exercise in reducing this kind of variability from one doctor to the next. We start medical school with creative and distinct ways of thinking, but we soon learn to recognize patterns and approach problems in a standardized way: when you see X, think Y, and do Z.
Risk calculators, diagnostic algorithms and treatment guidelines support us in this role. Surgical checklists can prevent infections; timely stents can save lives; computers can reduce medication errors. But not always. Sometimes checklists don’t help, stents hurt, and computers lead to overdoses.
But standardized care, by definition, is not personalized care: it fails to acknowledge patients’ individuality. A calculator can predict your risk for disease, and a clinical trial can reveal the possible side effects of treatment. Neither, though, can tell individual doctors or patients what to do — what tradeoffs to make, what quality of life to accept.
Standardization can also strip doctors of a sense of ownership and autonomy. In a health environment bustling with protocols and metrics, we sometimes feel less like doctors caring for people than technicians generating outcomes. With a growing reverence for algorithms comes a perception that physicians are somehow replaceable, or at least interchangeable. But individual doctor’s judgments, patients’ preferences, and connections between clinicians and patients are what make health care meaningful.
New technologies will likely further complicate the issue. Standardized care may soon give way to computerized care. Already, hospitals are teaming up with IBM’s Watson — the computer that won “Jeopardy!” — to digest new medical knowledge, collect data, diagnose disease, adjust medications and check for errors. A recent report by McKinsey & Company found that almost half of all activities American workers perform could be automated by currently available technologies. Contrary to prevailing thought, it’s not just low-skilled occupations at risk: physicians, financial managers, senior executives and the like will have significant amounts of their work automated.
Will doctors, then, soon be replaceable?
A better question is how best to incorporate new technologies into the day-to-day work that doctors do. The best medicine is an essential art, and algorithms, if thoughtfully deployed, can free us to make more of it. The great contribution of technology, then, may not be improved efficiency and safety. It may be crystallizing what only doctors — as humans — can offer: critical thinking, clinical intuition, empathic care, exploring what’s important to patients so they can make the decisions that are right for them.
We haven’t yet found the right balance between standardized care and personalized care — between automation and autonomy, algorithms and art. We know that protocols can improve care, but also that they can diminish individuality. We shouldn’t think of them as replacing what we do, but rather, as making room for what only we can do.
Dhruv Khullar, M.D., M.P.P. is a resident physician at Massachusetts General Hospital and Harvard Medical School. Follow him on Twitter: @DhruvKhullar.