Putting the horse behind the cart   (September 30, 2007)

A while back, I received a question about someone with diabetes who had blood glucose levels that were bouncing all over, from the 60s to the 500s. It seems that he/she was not receiving any basal insulin, and is getting insulin based only on blood glucose levels:

* 70-150 -no insulin
* 160-250- 4 units
* 280-350 - 8 units, etc
* above 450, call the doctor.

That ensures two things: if the blood sugar is normal, and no insulin given, the next reading will be high. And that the doctor will be called (and probably be unavailable to help!) exactly when the doctor's orders have caused the greatest possible screw-up and put the person at unnecessary risk.

Why do physicians use sliding scales? Because that's what we were taught in medical schools and residency training. When I was first in practice, the sliding scales were based, believe it or not, on urine sugar output (if the urine has trace to 1+ sugar, give...). But the concept was never studied in a systematic way as far as I could find. It was merely tradition, doing something because that's the way it was done before.

But it's wrong. You don't stop giving an antibiotic simply because the patient's fever has returned to normal. Why should you stop giving insulin because the sugar has returned to normal?