dka and all cause ketoacidosis

Whether Ketoacidosis is caused by Type 1 Diabetes, an SGLT2 Inhibitor, Alcohol Ketoacidosis, Amphetamines, Pancreatitis, Type 2 Diabetes with Pancreatic Toxiciy, the can all be managed very easily and similarly with one approach.

The most important issue is closure of the anion gap.  This is not a time to target glucose levels.

The diagram above is from Diabetes Canada and has been adapted specifically for ICU patients.  Follow this.  Some important issues to highlight:

• You really cannot give insulin if the potassium is <3.3. .  Insulin is going to cause a shift of potassium into the cells and patients truly can arrest.  This is the main reason we do not bolus insulin up front.  Get the potassium running IMMEDIATELY.  Start it peripherally until you have a central line.  If they can swallow KDUR and not vomit it, give them some.  If they will tolerate an NG (mot of these patients are extremely agitated form their metabolic derangements), give them KCL Elixer.

• You can start your insulin infusion as soon as the potassium is running, you just can't bolus is (in truth there is really no need to bolus it but in a patient with a higher potassium, you can, its not wrong)

• Once the potassium is above 3.3, it is ok to give bicarbonate if the patient is severely acidotic and panting to compensate.  Bicarb is going to very transiently worsen the intracellular acidosis by increasing the CO2 and the patient will have to pant faster (see carbonic anhydrase equation below).  For patients who are fatiguing, they may not be able to rise to the occasion and so you will have to go slower to allow them to compensate.  

• If they become hypoglycemic (although they should not with the instructions below), do not shut off the insulin.  Give them D50.

• Once the anion gap is closed, you can either decrease your insulin to 0.5U/Kg or if they are less agitated and able to eat, you can give a dose of long acting SC insulin and then  safely discontinue your infusion in 2 hours.  Don't forget to also discontinue to the glucose infusion if they are eating.  

• The SGLT2 Inhibitors have up to a 24 hour half life and so don't be too hasty in discontinuing insulin in them (see below)

There is no established protocol at the Ottawa Hospital that targets closure of the anion gap

There is no established protocol for ICU patients with DKA.  

Explicit Orders

• Use the ICU Insulin order set

• Delete the range and order a rate of 0.1 units/kg IV (example if the patient is 70kg, you will write a rate of 7 units of insulin/hr).

• Indicate a starting dose (so for the same 70kg person the starting dose will be 7 units

• Write "DO NOT TITRATE INSULIN" in the comments

• D50 infusion with a range of 0-50cc/hr.  Write an order to titrate up or down by 10cc/hr to maintain glucose in a range of 7-9

• Glucoscans much be hourly in order to maintain this

• Patient will require an arterial line

• DKA patients in the ICU often require central lines because of all the infusions of potassium, phosphate, often HCO3

• If the patient is delirious be careful with your sedation - If the are in the ICU they are usually severely acidotic and they are tachypneic to maintain their pH from the ketones and lactate.  Sedation may drop their respiratory drive

• Avoid intubating this patient for the same reason