Having spent the past 2 days at the 6th annual Diagnostic Errors in Medicine Conference, many thoughts have percolated as I reflect on the meeting. The conference had over 200 attendees and at such meetings one learns as much from hallway conversations as from the actual presentations.

The Society to Improve Diagnosis in Medicine is filling an important void – the emphasis on proper diagnosis. Back in the day (the 1970s and before), diagnosis was clearly job #1. Somehow focus shifted away from correct diagnosis to management strategies, performance measures and report cards. Wonks and politicians assume that we can make correct diagnoses. Unfortunately, the diagnostic process is difficult. We have problems of missed diagnoses, over diagnosis, and lack of diagnosis. One important policy influencing physician told us a story about a patient being treated aggressively without a diagnosis, a diagnosis that he (an internist) figured out with 2 questions. Having a correct diagnosis dramatically changed his relatives quality of life.

Such stories are all too commonplace. Too many physicians now put their focus on measurables.

Diagnostic accuracy eludes easy measurement. We heard “experts” (those who have written papers on the subject) disagree on how to define diagnostic error. Heated debates occurred on such definitions. We all understand that some diagnoses elude us more than others. How can we possibly assess diagnostic accuracy in a measure?

We could do what other performance measures do – look under the lamp post. We can look for missed colon cancers, or melanomas or breast cancer. We might even be able to study initial diagnoses of community acquired pneumonia or “congestive heart failure”.

But we always will suffer with defining a gold standard.

I suggested like other forms of patient safety, we need to study diagnostic mishaps, delays and errors using root cause analyses. Wrong diagnosis should become a subset of patient safety. We need to better understand the paths to incorrect diagnostic presumptions.

Clearly I do not have the answers. William Bruce Cameron clearly said (although often attributed to Einstein), not everything that can be counted counts, and not everything that counts can be counted. Diagnosis matters, whether we can count it or not. We must continue to learn about the root causes of diagnosis errors (broadly defined). We must restore a focus on proper diagnosis in medical education.