Stories from the Front Lines: The Journey to Meaningful Use

Dr. Jeff Hummel Talks About A Heart Attack in the Palouse: Blind Spots, Then and Now

My first job after passing the boards was in a small town on the Washington-Idaho border with an old GP well past retirement age. It was a short and unstable partnership, notable for a lack of communication.  The final straw occurred after I had made a house call to an elderly farmer living up a dirt road who was having a heart attack.  There was no ambulance so I drove him 15 miles to the critical access hospital that had a cardiac unit, and admitted him.  When I returned to the hospital the next morning I discovered that my GP partner had been there before me and sent the patient home despite elevated enzymes and an abnormal EKG.  When I called the man he said, “Doc told me he had his heart attack at home and I could do the same.” That was over 30 years ago, and although the farmer survived, the partnership did not.  Still, that episode stuck with me.  What in the world was he thinking? 

I recently found new insight into the matter by reading Clayton Christiansen’s book, The Innovator’s Prescription, in which he points out that there are two completely different types of medical encounters; one in which the job is to establish a diagnosis, and the other in which the job is to manage a condition for which the diagnosis is known.  This second type of visit is a relatively recent addition.  Before 1950 making the diagnosis was what medicine was all about. People either recovered or they didn’t, but effective treatments for diseases like heart attacks, diabetes and cancer didn’t exist.  As new therapies converted previously fatal diseases into chronic ones, we have simply added managing these conditions to the work of an office practice, yet nearly everything about these two types of visits is completely different including the necessary clinical skills, the information requirements, the best people to do the work, and the business models on which they operate.

The job of making a diagnosis involves a complex interaction between information gathering and pattern recognition that is guided by evidence-based diagnostic algorithms using pre-and post-test probabilities in a highly intuitive process.  Physicians spend years learning to do this in formal training, and many more years in practice perfecting the art. Once the patient’s diagnosis is known however, the job changes completely to one that requires teaching the patient self-management skills and monitoring for complications over long time intervals with a mechanical adherence to protocols using dashboards and checklists.  No wonder my old GP colleague told my patient he could have his heart attack at home. The job of making a diagnosis was done.  From his perspective the patient would either get better or die, but the details of managing his heart attack for him were a blind spot. 

For those of us trained after 1970 there is no such blind spot, yet even today, the differentiation between population-based chronic disease management and acute episodic care doesn’t quite capture the stark contrast between how information is used to establish a diagnosis and how it is used after the diagnosis is known.  A good example of this is the review of systems.  Physicians have an ambivalent relationship with the review of systems because it is seen as time consuming, it often feels “off topic”, the answers are usually negative and a positive answer may open a “can of worms”.  As a result, failure to do a review of systems is common. 

When a practice implements an electronic health record invariably the question is raised, “can my medical assistant enter the review of systems?”  This question often sparks a lively debate between compulsive providers who don’t want to delegate information gathering to their support staff and those who don’t want to spend excess time entering data into the computer themselves.  These discussions usually can’t be resolved because the two sides are talking past each other. If however, we ask question differently, “How is information from the review of systems used, and who should gather it in each of the two types of visits?” then the answer becomes not only clear, but more importantly one on which both sides may agree.  For the visit in which the job is to make a diagnosis, the review of systems is a disciplined search by the diagnostician for symptoms that may not have seemed relevant to the patient in reporting the history.  It is the job of the diagnostic expert to elicit and evaluate potential clues in the context of the diagnostic puzzle.  Delegating this task to a medical assistant would not be in the patient’s best interest.  In contrast, for a visit in which the diagnosis is known, the review of systems serves as a checklist to monitor for signs of a complication, or to indicate that a patient may have strayed from a standardized treatment pathway for a known diagnosis.  In this case clinical support staff, or even the patient might document the review of systems as long as someone evaluates every positive answer for potential clinical significance.  This is just one example of how the same information can have very different uses in these two types of visits, that we still think of as just office visits. 

Today I better understand my aging colleague with whom I attempted to work so long ago, and I’m glad our patient didn’t have a tragic outcome as a result of his blind spot.  It is important however, to remember that medicine today is changing faster than it did even under his watch. Now it is we who must recognize that we have blind spots of our own, many of which (just like his) are the result of vestigial remnants of ways of thinking better suited to how we practiced medicine 30 years ago than they are for what successful medical practices will need to look like in the future.