In 1949, J. Deryl Hart, MD, then Chair of Surgery and future President of Duke University, opened Duke’s Surgical Instrument Shop, which was charged with building “anything to aid [healthcare practitioners] in their work and studies.” Alongside myriad surgical instruments, its products included a well-known “Duke University Inhaler” for administering trichloroethylene anesthesia and a folding operating table. As a medical student, I walked by the original site of the Instrument Shop every day, and as a trained surgeon I’ve appreciated how surgical instruments extend my ability to reach into difficult places, dissect tissues at odd angles, and rejoin tissues after injury.
But while the Instrument Shop shaped (and still shapes) metals that extend our capabilities in the operating theater, we now live in a world where data—all those bits of information about our physiologic state, our demographics, our social environments, our genetics and epigenetics—increasingly comprise the raw materials for making new tools for helping our patients. In fact, one could argue that data scientists are the metallurgists and instrument makers of our time.
A little more than two years ago, when we founded Duke’s center for actionable health data science, we named it “Duke Forge” with this metaphor in mind. We are, in fact, instrument makers—not so very different than the craftsmen of the Surgical Instrument Shop. This may sound quaintly artisanal and not remotely “tech” or “data science-y”, but it is very deliberate. “Artificial intelligence,” “machine learning,” “statistics,” “data science”—whatever you want to call it—is an instrument. It’s an extension to a fundamentally moral activity: healing our fellow humans.
As such, I see our job at Forge as building and assessing the instruments of data science in human terms. When I pick up a surgical “right angle clamp,” it’s curved to fit in my hand; it has rings for my fingers; it’s rounded and smoothed to be held comfortably. What it is not is merely an abstract capability designed to dissect or clamp in a plane orthogonal to one’s arm. It absolutely performs this task, but only because it intuitively conforms to the surgeon’s hand.
Many of the algorithms we see discussed in the news or at machine learning conferences are still abstract capabilities that may theoretically outperform humans but are, for all intents and purposes, currently impossible to use in daily healthcare delivery. There is no intuitive congruity to the clinician’s hand—or to our daily workflow.
This is where our hard work lies: to take the tremendous theoretical capabilities available to us and make them useful and intuitive to already overburdened and often burned-out clinicians. As I’ve noted in previous blog posts, there are huge cultural and technical shifts that must happen. For instance, that right angle clamp has relevance and utility in the operating theater, but is sorely misplaced at a CVS Minute Clinic. We need to figure out the right contexts for the algorithms we use, and we’re still at a very early stage in this journey. There’s no doubt that data science will fundamentally change how we practice medicine, and this obligates us to figure out the intuitive connections that link the professional, the instrument, and the patient.
As the new Director of Duke Forge, I am profoundly indebted to the mentorship and example of Rob Califf. Rob already has made monumental contributions to the goal of making clinical research useful to the world by founding and directing the Duke Clinical Research Institute and Duke Translational Medicine Institute, as well as serving as Commissioner of the FDA, among many other roles. Rob’s vision for Duke Forge encompassed data science as an enterprise done in the service of our obligations as healthcare professionals. Early on, when we were explaining the underlying concepts that drive Duke Forge, Rob used an analogy that continues to resonate for me: “we take rigid structures, melt them down, and make them into new tools.” Medicine is changing. If data science is going to help change it for the better, we will need hot fires, and much vigorous pounding on anvils.
As we consider the next chapters in the application of data to healthcare—really the reformation or the recasting of healthcare—we must anticipate the challenges we will face. These will include:
- Identifying and managing algorithmic bias;
- Addressing the possibility of “de-skilling” in the health professions as algorithms assume roles in clinical decision and operations support; and
- Squarely addressing the deficiencies of electronic health records in exacerbating clinician burnout, their failures in providing useable data, and their impeding innovation in patient and clinician-facing applications.
The novelty presented by the incredible capabilities of data science—as with any new technology—does not mean it circumvents our responsibilities as clinical and scientific professionals. I look forward to renewing and extending the Forge’s mission of providing a multidisciplinary home for the “conscience of data science” and fashioning a new health data science culture for Duke and beyond.