In the spring of 2007, I was very interested in how physicians gather information and learn at the point of care. As I thought through what was needed, I sketched this on the back of a napkin.
The box represents a computer screen that is divided into three parts, each with a slightly different purpose. The left third of the screen represents the traditional electronic medical record interface where a physician would enter information on vital signs, details from the history or physical, and patient expectations. Because the meaningful use of EMRs is now mandated, and by 2015 EMRs will serve as the nervous system of medical practice through which all information flows, it seems like a logical starting place.
The middle of the screen is designed to respond to the details of the EMR. As a physician enters new information, a search algorithm would crawl the existing body of knowledge (everything we have learned about medicine) and provide suggestions and alerts. The search components of this prototype EMR would be a form of artificial intelligence, much like Google or Amazon provide us on a daily basis—if evidence exists, this middle search component would bring it immediately to the fingertips of the physician. (I came up with this long before IBM’s Watson debuted.)
Physicians could use the right third of the screen to ask questions at the point of care to a broad pool of physicians. This would turn healthcare professional social networks into a real-time learning community. The need for these social-learning components is increasing—while the amount of information that is available to a physician is overwhelming, the amount of viable, contextual evidence is often very underwhelming. The right third of the screen is designed to allow physicians to leverage the collective intelligence of their professional community.
Three things that concerned me back in the spring of 2007 still do to this day:
1. We don’t support the physicians who don’t realize they have questions that need to be asked or don’t know how to ask them. (They are wrong, but don’t know it.)
2. We don’t support physicians who have a question, but forget or misrepresent it by the time they have a chance to ask.
3. We don’t support physicians who have questions that don’t have a simple, evidence-based answer.
Any of these three concerns can undermine healthcare quality, but when viewed together it’s clear that we need a new learning support system.
My cocktail-napkin sketch intended to solve these three problems at the systems level by integrating the act of information-seeking and learning into the practice of medicine. The prototype EMR would ensure that information flows to the physician at the point of care. It would propose new ideas that trigger appropriate episodes of cognitive dissonance at the very point that high-stakes diagnosis and treatment decisions are being made. And it would offer two sources by which a physician could answer the right questions by providing real-time access to the existing evidence base (middle third) and to a community of healthcare professionals (right third) who may be able to draw upon unique personal experiences. In short, the prototype EMR would re-engineer information flow, learning, and behavior change in the medical professions.
In the past year I’ve thought through the development of this prototype and developed a list of the fundamental elements that would be required to optimize this model. In subsequent columns I will explore these elements in detail, but for now, as you think through whether new social technologies can help physicians learn, know that we are just scratching the surface of the types of systems that will bring about real change.
Brian S. McGowan, PhD, has dedicated the past 12 years to medical education as a @BrianSMcGowan.member, mentor, accredited provider, and commercial supporter. The opinions expressed are McGowan’s and do not represent the views of past, current, or future employers. Follow him on Twitter:
More of Brian's Columns: