Saturday, December 6, 2008

Case based reasoning through care pathways: a patient of toxic epidermal necrolysis

Have been busy with some interesting patients these few days.

Although I am a unit incharge I prefer having complete charge of a few patients and tend to sometimes also work like a resident for them. Right now have this interesting patient with Toxic epidermal necrolysis who developed severe sepsis with acute tubular necrosis and non oliguric renal failure soon after admission to our ICU. Other than a posible cutaneous source of infection I found that a few attempts had been made to catheterize him with a foley's as per our ICU protocol (which I discovered to my horror advocates foley's in all patients which the residents tried to follow blindly).

His urine cultures also showed muti drug resistant E coli as a result and I had to start him on imipenem for the UTI along with linezolid for his presumed cutaneous infection.

This is one danger I believe of going by average information in protocols that is necessary for learning in the initial phases but again more subtle layers of information are needed to guide the execution of protocol based information. At the moment we need to develop this layer. One way would be: if health professionals share their experiences a form of experiential learning could be promoted. It is taking place but needs to be promoted on a wider basis. Most of the times our case based reasoning and subsequent sharing is very limited to our local friends and colleagues (but again can be remarkably augmented with the power of the net).

Again coming back to the same patient yesterday I noticed that his potassium had suddenly gone up to 8.9 mmol/L. I got an ECG done and found it normal and relaxed thinking it (his high potassium) couldn't be true especially since one report the day before was 3.4 mmol/L. However on closer look at the report of 3.4 mmol/L I realized it was someone else's report wrongly stuck on to this patient's file (doesn't speak well of our ICU staff including myself I know).

The patient's relative had obtained this wrong report from the lab without noticing the name, handed it over to the ICU sister who stuck it to the file without noticing and I had glanced through the day before without noticing the wrong name either. When we collected the correct report that was also raised (which proved that it was steadily rising and we rushed him for a hemodialysis for severe hyperkalemia).

Yesterday night I reflected on the appearance of severe hyperkalemia in this patient and that too without ECG changes and sure enough found some reliable literature in the net describing severe hyperkalemia without ECG changes and also realized that this person had been started on an enteral protein supplementation and it was possible that he was gaining a lot of potassium through that. I have changed his protein supplement powder to one that contains low potassium along with continued other drug management for his hyperkalemia.

The point of all this case discussion is to a) illustrate the simple, complex to chaotic evolution of care pathways (this was off course acute care) and b) reiterate the inadequacy of average text book information which may not cater to real time point of care requirements and c) emphasize the urgency of creating online experiential patient centered learning networks to facilitate better health care outcomes (and improve medical education in general which still hangs on to gleaning average information in intermittent boluses of lectures/discussions in weekly/monthly conferences etc). d) Obviously all of us need not go through all the experiences of others but only that which may be relevant to our point of care case based needs. There is a very interesting article on case based medical informatics that is being developed and needs our support in its further development at this link:


DJ Elpern said...

T.E.N. is one of the most difficult skin disorders to treat. We usually use prednisolone or solumedrol, although there is no evidence these are effective. IvIg may be of some value. But what is more important is good nursing care - similar to what one would give to a burn patient. Most cases are drug-induced. I can remember one that I caused many years ago with Dapsone. Fortunately, the patient survived. Photos of this patient would have been helpful. Good luck and keep us posted.

Avinash Ghimire said...

I recently took care of someone with chemo induced TEN.We treated him with steroids.We had dreaded sepsis but he did well with supportive care.

Anne Marie said...

Great to see your blog. It is fascinating to read and a great way to learn for even a humble GP like me.
Thank you,
Anne Marie