When one studies diagnostic errors, one quickly discovers that premature closure occurs very often.  We all run the risk of assuming that the initial diagnosis is correct.  We have to train ourselves to avoid this problem.

My observation, over the years and in different hospitals and working with different residencies, is that the problem occurs primarily because insurance companies, and therefore hospital administrators, require a diagnosis for admission.  Often on the second day of admission we have to define diagnoses prior to working through the diagnostic process.

Our current residents know that “community acquired pneumonia” is my favorite example.  So often we have delays in working through the diagnosis process because we assume that the proffered diagnosis is correct.

So what is the solution?  We should delay labeling patients with diagnoses until we have enough information to assess a high probability to our diagnosis.  I propose that we advocate for allowing admissions without a specific diagnosis.  We should encourage the concept of undiagnosed disease manifested by certain sign, symptoms and test abnormalities.  We should encourage a clear understanding that the diagnostic process is underway and necessary.  I fear that without some changes in our processes we will continue to under emphasize correct diagnoses.  As I have written in the past, diagnosis is job #1.  We must work hard to get at correct diagnoses.  We must remove any barriers to achieving proper diagnosis.