The problem of admission diagnoses – a guest post


Category : Medical Rants

I received this response to a recent post. It is so good that I wanted to share it – so with Dr. Thomas Nielson’s permission I have. He makes the important point that the rush to LABEL the patient with the diagnosis has major unintended negative consequences. He says it so well that I encourage your reading and comments.

Thank you for this post. This is a problem that occurs from time to time, and I believe that the current system in place for admissions is a large part of the problem.

We are asked to diagnose people in the emergency department because we need an “admission diagnosis” so that we can make sure that we meet “admission criteria”. The people in hospital administration who require this have never taken care of patients themselves, and they have no idea what they are talking about.

There is a disconnect between administrative types and doctors, and I do not know how this can be solved in our current system.

Example: Patient presents with AMS and is found to have a massive acid base disorder in the ED. What is the diagnosis? I have no idea! I need to admit the patient to the hospital and run a bunch of tests before I can tell you what the underlying problem is. Is it ethylene glycol poisoning? Is it their renal failure? I don’t know.

The current system puts the cart before the horse in requiring an admission diagnosis. What happens is some random diagnosis is given to the patient so that we can get them into the hospital. Unfortunately, because the system demands that we treat the patient with “quality” care, this diagnosis puts a process in place. Now the patient is run through a bunch of tests relating to the admission diagnosis which may or may not be the actual problem going on with the patient. And at the end the patient is sent home, dazed and confused, without a clear understanding of WTF just happened to them.

Think I am kidding? How about this example: Patient present with shortness of breath and the radiologist in the ER says pulmonary vascular congestion. The BNP is 300. Now the patient is admitted with “congestive heart failure”. Do we know it is CHF? NO! But now, because CHF is a diagnosis which beancounters believe they can treat by protocol alone, the patient is set upon a course in which all of the “quality” measures must be met to make sure that we get paid. The patient does not know what is going on, but all of a sudden they are being put on a low salt diet, being given an ACE inhibitor, and so on. They will (hopefully) get an echocardiogram and then….what happens if the echo is normal? SYSTEM FAILURE

To get back to your post, it is often the system that is running away with the patient when we have a sense that the diagnosis is wrong. And studies have repeatedly shown that pinning a diagnosis to a patient early in the process, which is required by the system, leads to significant bias in the doctor’s treatment process and judgment.

What we need in this country is for the doctors to reclaim their rightful place in the system. All of the quality metrics BS needs to go right out the door.

How do we do this? I am afraid that we are going to have to revert to doctor-run hospitals, with all cash. Let patients deal with their own insurance companies.

Comments (7)

there are a couple of solutions, some easier than others. Having courses to familiarize admitting MD’s with how to handle the dilemmas financially would be one but you would have to do that for hundreds of providers so not’ likely to be effective. I used to get calls from the ER on diabetes and thyroid all the time and guided the docs over the phone. We could do the same for these admission problems by developing an administrative on call process. Plenty of Suits have MD’s which would give them a better grasp of this interface between payment and care. Each MD/Suit could be assigned a week’s on call time per stretch and questions on this opposed to that doctor. Presumably that would make the admitting doctor less culpable and if there are enough annoying calls of this type, the management can review the problem for a more durable solution.

The best option for patients and payment but the most cumbersome would be to change payment and quality not on an erroneous initial impression but on a confirmed final outcome. This would need some cooperation from the people we vote for and from the Third State. Hospitals are now large enough to do effective lobbying as are some of the professional societies. It’s the hardest solution but changing a plan from what you think needs to be done to what actually needs to be done serves patient and payer the best.

“All of the quality metrics BS needs to go right out the door.”

If just one medical society were honest enough and brave enough to say that, this entire house of cards might start to crumble.

I can only comment based on my own personal experience. One day I wake up with crushing substernal pain. The pain gets worse as I try to get up and walk. I get ready for work, but realize the pain hasn’t resolved. I head to ER instead. I get to the ER, ECG normal. The pain still persists. The staff once the ECG is normal, ignores me. But I speak to to the ER physician, who also doesn’t like the story. It’s the only reason I went to the ER. Still nothing has been done to resolve the pain. I’m admitted (not sure why), and they insist on a cardiac stress test. While waiting for cardio or my IM physician to see me, they finally give me sublingual nitro, which immediately resolves my pain. My wife, also a physician, and her cousin who has priveledges at the hospital I am in, pulls everything up. We realize its an esophageal spasm. Essentially self diagnosed. Once I pass the stress test I’m able to go home. Never once was I diagnosed by any of the health care team in the hospital.

Great post! Another thing to consider is that these incorrect diagnoses can follow a patient forever and cause further problems with medical care and insurance further down the road.

One possible answer: stop trying to get to the most specific possible diagnosis in the ED. ICD-10 includes diagnosis codes for most non-specific symptomatology, mostly in the R chapter. Start there; let the hospitalist drill down to the specific codes as more data becomes available. Also, make liberal use of the words “possible,” “probable,” and “suspected.” However, for our hospitalist colleagues, that means we have to be on the ball and make appropriate notations of “ruled out” to prevent inappropriate coding and inappropriate application of useless metrics (see: sepsis).

Absolutely important comment

I wish it were that easy. Too often we get pressure from the suits to provide a diagnosis.

Post a comment