As a recently minted MD with a keen interest in health IT that began with working on Google Health six years ago, I’m decidedly on the forward-looking end of the spectrum in the physician community. I think decision-support algorithms have the potential to reduce errors and save our health care system a boatload of money by passing a greater share of patient care down the ranks to clinicians who are less-highly-trained and, therefore, less expensive, leaving physicians time to focus on more complex cases and interventions. I’m intrigued by the partnership between IBM and Memorial Sloan Kettering to train the supercomputer, Watson, to assess, diagnose, and make treatment recommendations for cancer cases. I’m optimistic about robots placing IVs or performing routine surgical procedures in the future — possibly better than human providers do.
That said, as a touchy-feely people person who romanticizes the days of the old-school general practitioner who knew your whole family, I’m sympathetic to the view that the last thing modern medicine needs is more machines and less humanity. But I think the conversation surrounding this topic has tended to create a false dichotomy between doctors and computers.
In what ways might technology bring more — not less — humanity to doctors and to the practice of medicine?
A fundamental change in the role of the physician is not unprecedented in human history. For hundreds of years, physicians were craftsmen. They learned much about anatomy and developed practical skills in diagnosis and procedures. They relied on reasoning and empiric observation to conclude, as far back as ancient Greece, that purified water and sanitation improved the health of their communities, even though they lacked a scientific understanding of the germ theory of disease, which wouldn’t debut until a couple of millennia later. Early doctors practiced their craft and took cues from the fifth century’s Hippocrates, who considered medicine to be an Art with a capital “A.”
With the arrival of modern medical science and the barrage of knowledge and data that came with it, physicians have turned to more cerebral tasks and have inadvertently become, in some sense, information storage and processing pseudo-machines. Particularly since the debut of evidence-based medicine in the late 1980s, doctors have been forced to memorize ever more medical facts and innumerable diagnostic and management algorithms (all of which are constantly updating) to offer their patients the standard of care. They now also spend much of their time interfacing with computers instead of patients to perform structured data entry in electronic medical records. These tasks are time consuming and resource inefficient, and they distract from direct patient care.
But I think we can turn this curse into a blessing. I don’t think technology will kill the craft of medicine. In fact, I think it may be the only way to save it.
Handing off data storage and processing tasks to machines that will almost invariably perform them more accurately and efficiently than we do could free us from the computational burden we’ve rightly created for ourselves. If phase one of modern medicine was acquiring the scientific knowledge and concepts to dramatically improve health outcomes, phase two is designing information systems to liberate physicians’ inner hard discs and working memories, so to speak, as well as their calendars, leaving more room for the human side of medicine, what I believe we will always do better than machines — caring for patients.
True care demands sophisticated, time-intensive interpersonal communication, relationships, and counseling with patients. It requires internalizing the important distinction between medicine and healing, between treating the body and treating the soul, yet understanding that these are two sides of the same coin, and knowing how to integrate them appropriately into routine care instead of abdicating existential human matters to a chaplain, social worker, or psych referral. When we over-focus on the medicine, we forget how to heal, we forget that merely talking and listening is therapeutic. We think we have nothing much to offer a patient with terminal cancer, or another who suffers from chronic pain without any known organic cause. But even without a cure, we can be healers for such patients.
This kind of healing may already be practiced sporadically or within certain specialties such as palliative care, but it bears little resemblance to what medicine on the whole looks like today. But I think it is better medicine. And letting computers do health care’s more algorithmic work might allow us the time and emotional space to re-establish the Art of healing, to be more, well … human. I feel modern medicine has lost this art, this humanity to science. Perhaps incorporating technology into health care in a very deliberate way can help us find the best of both paradigms, and make medicine and healing one again.
Doctors must undoubtedly still learn the ins and outs of diagnosis and treatment and should reserve the right to overrule computers using their clinical judgment. Besides, some situations simply defy the reductionism of algorithms, like when the machines say a patient is stable, but the seasoned physician just knows she’s about to crash. And, in fact, in our clinical training we are taught when to “trust the patient, not the numbers”, and to develop an instinctive approach to whether that patient in the emergency room bed is “sick or not sick”. As a student, I had plenty of attending physicians tell me to put away my notes on a patient’s vitals and lab results and ask what my gut was telling me.
Such a paradigm shift in how we practice medicine would necessitate big adjustments in how medical schools select and train students. Instead of using science GPA and MCAT scores as a primary weed-out mechanism, admissions committees could initially screen for emotional intelligence, and later delve further into applicants’ communication skills and abilities to synthesize information and observations, think and act holistically, and collaborate with machine-physician counterparts.
The Icahn School of Medicine at Mt. Sinai in New York has already begun to experiment with a program that allows liberal arts college students to gain early acceptance and skip the MCAT and undergraduate science course requirements altogether. It’s been so successful at creating students that can not only pick up the necessary science but also become more sensitive doctors, that last year Mt. Sinai announced an expansion of the program. By 2015, about half of their incoming students will be admitted through this “FlexMed” program. And perhaps selecting students this way would obviate the need for what I considered remedial communications skills training in medical school, or the clunky instructions of the scientist’s scientist trying to teach communications — for example, in my first year of school, when a handout prompted us to “place your hand on the patient’s shoulder in a reassuring manner” during the physical exam. Yikes.
If this is truly an opportunity to save the art of medicine and healing, then providers and technology experts everywhere should focus on building new products and services that give clinicians the bandwidth to be more human again, to salvage and rebuild what’s left of their therapeutic relationships with patients. One example that I encountered in my clinical training and in the literature is that both patients and doctors barely tolerate desktop computers in the exam room because of the feeling of too much time looking at a screen instead of the patient — but almost everyone loves tablets. This simple change in user interface has a profound impact on whether people feel like there’s a barrier to their interaction or an enhancement. A tablet computer sporting useful diagrams that doctors and patients look at together can enable shared decision-making, which could be the impetus for a stronger connection with the confused patient sitting on your exam table looking for medical answers, but also human reassurance.
Thoughtfully implemented technology and design – from information systems at large medical centers to the workflows in small community urgent care clinics — have the potential to solve problems such as cost and inefficiency by innovating and fundamentally changing our models for health care delivery.
But they should also just make us (both patients and providers) feel good. They should delight us the way that consumer technology already does every day. And they should help us feel more human, perhaps even by being less conspicuous. When technological innovations in health care become so intuitive and seamless that they allow doctors to spend more time with patients than they spend with computers, we’ll know we’ve made a good start.