Why do we have to provide an admitting diagnosis?


Category : Medical Rants

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.


Comments (5)

The Desk Chair Mafia frown on R/O’s/

Taking a page from Larry Weed’s approach, if I am not certain of the patient’s underlying condition or process after the initial admission evaluation, then I will enter a symptom (e.g., “fever”) or syndrome (e.g., “SIRS”) as the admitting diagnosis. This can be refined later as the patient’s clinical course evolves. I then delineate my differential diagnosis in the History and Physical.

I will admit that I rely on my colleagues in Care Coordination to provide me guidance regarding “medical necessity” and “inpatient status” (a separate but related issue).

Wait – The Gods of Big Data and ICD-9/10 Demand a Diagnosis to link to a Procedure/Therapy.

One of the problems is that, at least on MediCare patients, the DRG and thus the LOS is determined by the admitting diagnosis. Discharge planning relies on knowing the LOS, and coordinating post-hospitalization care. Thus the diagnosis is key to the whole admission, and if it isn’t determined until day 2 or 3 of the hospitalization, and the LOS is 3 days, you are now behind the proverbial 8 ball in terms of discharge planning. Concurrent coding can help. but only if you are making appropriate diagnoses to count as CCs or MCCs.

Of course none of this has anything to do with patient care, but it’s a reality of hospital care today.

Spot on!

Post a comment