DKA and Potassium

I work in the dirty nasty central-city ER and ICU.

We probably wouldn't treat the K+ of 6.3 on a DKA patient in the ER but we don't start fluids w/KCl either. After rehydration with 3-4L of NSS, the patient's fluid does get switched to one w/KCl once they're in the unit.

Oh, and I have to comment on your statement that DKA is pretty serious as far as emergencies go ...

At my hospital, about half of the DKAs don't even go to the unit, just to a tele floor. And we have a small handful of frequent flyers in the ER who show up with blood sugars > 600 but simply get rehydrated, 1 or 2 doses of insulin, a lecture on compliance, and then discharged.

Our orders for DKAs in the ER are if the K is > than 6.4, we treat with k/x, start insulin drip, and send to ICU, mainly to monitor for cardiac sequelae (no particular reason for 6.4, I honestly just think someone pulled the number out of their rectal vault).

In the unit, we change the fluids to reflect changes in K and glucose. One of the worst nights I ever had was when a newbie (knew it all, just ask'm) didn't put the pt on fluids with D5 in it, and the glucose fell like a rock -- started getting cerebral edema, VS and orientation changes. Told the idiot you wanted to drop the sugar slow, but they wanted the patient to get 2 negative acetones so we could toss them upstairs. Thanks to them, we almost tossed the patient into the morgue.

I hate morons.

I work in the dirty nasty central-city ER and ICU.

We probably wouldn't treat the K+ of 6.3 on a DKA patient in the ER but we don't start fluids w/KCl either. After rehydration with 3-4L of NSS, the patient's fluid does get switched to one w/KCl once they're in the unit.

Oh, and I have to comment on your statement that DKA is pretty serious as far as emergencies go ...

At my hospital, about half of the DKAs don't even go to the unit, just to a tele floor. And we have a small handful of frequent flyers in the ER who show up with blood sugars > 600 but simply get rehydrated, 1 or 2 doses of insulin, a lecture on compliance, and then discharged.

Thanks for sharing your experience!

I have also d/c'd for several patients where noncompliance was the main issue, including an 18 year old who basically refused to take her insulin... although mortality rates have really dropped in the past 20-30 years from DKA, it's still 2-5% which is not exactly a sprained ankle.

We just learned DKA in school the way the OP said. No mention of k/x ever being used in DKA no matter what the K level.

In the unit, we change the fluids to reflect changes in K and glucose. One of the worst nights I ever had was when a newbie (knew it all, just ask'm) didn't put the pt on fluids with D5 in it, and the glucose fell like a rock -- started getting cerebral edema, VS and orientation changes. Told the idiot you wanted to drop the sugar slow, but they wanted the patient to get 2 negative acetones so we could toss them upstairs. Thanks to them, we almost tossed the patient into the morgue.

I hate morons.

Don't you have an order of some kind for what kind of fluids to put them on?

Where I work, we have flowsheets. As long as the C02 is above 8 they come to the floor, any lower and they go to ICU. They get 2-3 liters of NS pretty quickly and then once there BS below 300 we switch over to fluids with D5. In the meantime, we draw chemistries often and monitor C02 and K. If K is

Don't you have an order of some kind for what kind of fluids to put them on?

It's on our DKA protocol to change the fluids in response to the labs, but this particular person thought that if they didn't start the D5 when the fsbs was below 250 that they'd get 2 negatives faster and get them out of the unit sooner, so they just didn't change them to D5 half NS + 40 of KCL and kept them on NS + 40 KCL. Spectacular non-grasp of the basics of DKA, and also non-grasp of what the word "order" meant.

In my ED we treat our DKAs with K+ 6.5 or greater with sodium bicarbonate and calcium gluconate, in addition to 0.9% NSS fluid resuscitation and insulin gtt. Remember that with an insulin gtt, it's going to take some time to drive the potassium into the cells along with the glucose. Even a 1 point deviation in K+ from normal range is associated with significant mortality, so it's especially important to treat critically high K+. When the patient gets to the ICU, continuing monitoring of K+ levels will determine when to switch to KCl containing fluids.

The ED doc writes the orders for the initial treatment, then the hospitalist writes the orders for the ongoing treatment. The ideal situation is that it should be seamless with each physician and unit understanding their role and working as a team. The focus in the ED is preventing cardiac complications r/t hyperkalemia, and patient should be in the ICU long before we need to be worried about hypokalemia r/t K+ shifting back into the cells.

One major ED rule is that the patient is not staying! They're either going back from whence they came, transferred to an inpatient unit, or transferred to that big unit in the sky (or down below, whichever seems appropriate).

A delay in transfer would definitely change the course of treatment in the ED. We'd have to think about rechecking labs, switching fluids, etc.

Just want to add that Kayexalate takes a while to work. It's not likely to cause a precipitous drop in serum K+, especially while K+ levels are being monitored Q2-4 hours in the unit. You're going to catch a trend before it becomes a problem. Also, even though things like Na Bicarb and Ca Gluconate act rapidly, serum K+ levels can rebound back up. Hyperkalemia does need to be addressed initially in the ED setting, but the continued monitoring needs to happen in the inpatient unit (ICU or tele) with ongoing parameters and protocols in place.

This is one of many good examples of why delays in transfer from ED to ICU should be avoided.

Remember, nobody stays in the ED!

Hi,

In my ER we have a protocol for DKA that says we bolus 2L NS and start insulin drip. We check blood sugars q2 hours and titrate insulin gtt accordingly. After the 2L NS bolus, we start 1/2 NS with 20 of K no matter what the K is since they will drop as they rehydrate and normalize their insulin. When they get sent to a room (whether it's ICU or a tele unit) they automatically get placed on electrolyte replacement protocols. We would NEVER give kayexelate for a high K on a DKA patient!

Yikes, I shudder to think!

Many of our DKA's are on hourly litres of N/saline, so their K+ dilutes down, as well as the intracellular insulin-driven shift. Until the BSL is under 15mmol/L (270mg/DL) we replace K+ at a rate of 20mmol/hr if lab K+ is under 3.5, 10mmol/l if 3.5-5. and 5mmol/L if K+ is over 5; I've never had a patient in DKA with a glucose above 15 who had K+ replacement withheld, particularly if they're still acidotic.

Our DKA patients are nursed on the endocrine ward (mine!) unless their pH is under 7.0 in ED, in which case they spend some time in ICU.