The Quality of Diagnosis

Posted on Feb 16th, 2008 at 9:50am.

Has it ever struck you that most of what we work on in quality assumes that the diagnosis is correct--diabetes, colon cancer, mycoplasma pneumonia, rheumatoid arthritis....etc.--and that our principal quality challenge is simply to deliver the right evidence-based care, safely, to the person who has that condition? Have you ever wondered about the quality of the process that resulted in reaching the diagnosis itself?

Here's an illustration of what I mean. If two doctors started with identical patients, at the very same stage in the evolution of the very same disease, and one doctor came to the correct diagnosis over 1 month, in two visits, after $150 of lab tests, and the other doctor took 6 months to reach the same diagnosis, despite 4 referrals to various other doctors, $2500 in lab and imaging studies, and two costly, invasive and dangerous procedures, wouldn't you say that the quality of the diagnostic process was better in the first instance?

In a recent article (JAMA 2008;299:338-340) Eric Holmboe, Rebecca Lipner and Ann Greiner of the American Board of Internal Medicine surface this question of "diagnostic quality" as they consider whether physician knowledge and clinical judgment have an impact on quality.  I have argued for years that being Board certified in whatever specialty has relatively little impact on the likelihood that a doctor will reliably execute an evidence-based treatment plan for a common condition. That aspect of quality (delivering the treatment plan reliably for a given diagnosis) is far more dependent on the systems and teams with which that doctor works.

But I have also felt strongly that being Board certified (in the case of the ABIM, passing a fairly stringent test of clinical knowledge and judgment) would make a difference in how quickly, efficiently, and safely a doctor arrived at the right diagnosis, especially for less common conditions, or for unusual presentations of common diagnoses. It is surprising to me that so little work has been done to understand variation in the quality of the diagnostic process, and to assess what factors predict efficient, accurate diagnosis. Here are some of the questions I would like to ask, just to understand the variation, for starters: For patients with the same eventual diagnosis...

  1. How many visits did it take to reach the correct diagnosis?
  2. How many referrals to different specialists?
  3. How many laboratory tests?
  4. How many imaging studies?
  5. How many procedures? (biopsies, endoscopic examinations...)
  6. How long did the whole process take? (this is the "sleepless nights" question)

Once we had some sort of picture of variation, we could begin to study factors that predict the quality of diagnosis, and perhaps even begin to test changes that would improve the quality of the diagnostic process. This arena (the quality of diagnosis) might well be something of a "next frontier" for quality work, and I'm very interested in hearing from anyone who is working on this problem.


We had the beginning of a discussion on quality of diagnosis today at our Board meeting. BTW, our CEOs reflect the value of the work you've done with them over the last 2 years. Thank you!

Posted by Sherry Bright (Jan 16th, 2009 6:30am)

Please enter the six characters in the image above.