Dka

[jok200]

I understand that DKA patient are hypokalemic even though the bmp will show hyperkalemia because of the shift for the acidic environment. How do I know what the actual potassium actually is though? I initially give iv fluids but wait on the insulin because if I give them the insulin and they are hypokalemic I could further exacerbate that state and cause an arrhythmia. So I wait for the bmp which will show a false elevation in potassium, so their was no point in waiting right? Normally i would give the fluids, then oral potassium and insulin, but i am confused on how do I know the actual potassium in the DKA patient and do I have to wait for the potassium level before giving insulin?


thanks, sorry I know a simple question, just keep getting confused about it.

---

Members don't see this ad.

 

[Blackcomb]

You have to estimate the actual potassium based on the level drawn on the initial BMP and the degree of acidemia. Most algorithms state potassium repletion for levels below 4.5 (there are varying cutoffs between 4-5; of course you would never replete K in normal patients at these levels but you assume a correction factor in DKA). You make an estimated guess, replete K empirically, and repeat a BMP as soon as you think is clinically indicated (could be 2 hours, could be 4 hours based on how urgent/dangerous the degree of hypo/hyperkalemia).

Remember that the primary issue in DKA is acidemia, and the only way to reverse that acidemia is to provide insulin to the body (as opposed to HHS). You should almost never delay in starting insulin.

 

[jdh71]

I understand that DKA patient are hypokalemic even though the bmp will show hyperkalemia because of the shift for the acidic environment. How do I know what the actual potassium actually is though? I initially give iv fluids but wait on the insulin because if I give them the insulin and they are hypokalemic I could further exacerbate that state and cause an arrhythmia. So I wait for the bmp which will show a false elevation in potassium, so their was no point in waiting right? Normally i would give the fluids, then oral potassium and insulin, but i am confused on how do I know the actual potassium in the DKA patient and do I have to wait for the potassium level before giving insulin?


thanks, sorry I know a simple question, just keep getting confused about it.

Treatment recommendations came out in 2009 look them up

Basically get to resuscitation first, you need to start dumping fluids to get the UOP to >50cc/hr, because if the kidneys have gone all crumpy-pants on you, dealing with K could get complicated, but in most DKA pateints too much UOP has been the problem, right? So start the fluids. Get your labs. If K is < 3.3 give 20-30mEq/hr until the K is above 3.3, then you can start your insulin. If K is > 5.3, hold any K replacement and start your insulin treatment, check K every two hours. If your K is >3.3 and < 5.3 put 20-30mEqs per hour into the patient, probably with each liter of fluid you give along with your insulin treatment. Try to keep K between 4 and 5. Done.

 

[jok200]

jdh71... i looked up the guidelines and I practice exactly what you have stated. So in reality their is no way to exactly approximate the actual total body potassium, it is simply assumed that the patient is total body potassium depleted because of a number of issues:
1-episodes of vomiting
2-osmotic diuresis
3-skeletal muscle utilization of potassium decreases significantly because the Na/K pump is ineffective with decreased levels of ATP because glucose is required intracellularly for ATP generation which is the active catalyst for the pump to work at all.

= this is where the theory of total potassium depletion originates. The algorithm you stated I have seen many times before but, I didn't know where the "why" so to speak originated.


-->oh .. their are two issues with DKA: 1) acidemia 2) hyperglycemia
.. The insulin corrects the hyperglycemia and the fluids remove the ketoacids which are acidifying the blood, people think the insulin corrects the acid, it corrects hyperglycemia only the fluids eliminate the ketoacids that acidify the blood.

thanks again guys-

 

[VentdependenT]

When the K hits 5 i start replacing.

insulin pumps K into cells


for every drop in K by 1, you need to replace 100-300meq K

Ive never thought about it anymore than that

 

[VentdependenT]

-->oh .. their are two issues with DKA: 1) acidemia 2) hyperglycemia
.. The insulin corrects the hyperglycemia and the fluids remove the ketoacids which are acidifying the blood, people think the insulin corrects the acid, it corrects hyperglycemia only the fluids eliminate the ketoacids that acidify the blood.

thanks again guys-

Wait..what?

Water follows glucose. Put glucose in cells water follows. 2/3rd fluid in body is intracellular (40% body weight). Thats why these folks need so much damn fluid. YOU metabolize ketones. Insulin restores cellular metabolism and inhibits further ketosis, meanwhile the body metabolizes the ones floating around. You dont just piss out all the ketones.

 

[jdh71]

jdh71... i looked up the guidelines and I practice exactly what you have stated. So in reality their is no way to exactly approximate the actual total body potassium, it is simply assumed that the patient is total body potassium depleted because of a number of issues:
1-episodes of vomiting
2-osmotic diuresis
3-skeletal muscle utilization of potassium decreases significantly because the Na/K pump is ineffective with decreased levels of ATP because glucose is required intracellularly for ATP generation which is the active catalyst for the pump to work at all.

= this is where the theory of total potassium depletion originates. The algorithm you stated I have seen many times before but, I didn't know where the "why" so to speak originated.


-->oh .. their are two issues with DKA: 1) acidemia 2) hyperglycemia
.. The insulin corrects the hyperglycemia and the fluids remove the ketoacids which are acidifying the blood, people think the insulin corrects the acid, it corrects hyperglycemia only the fluids eliminate the ketoacids that acidify the blood.

thanks again guys-

I see. You were trying to figure out how to estimate total body potassium deficit. I bet there is a way to estimate, though . . . I'm not sure, and some of this is just my way of approaching these kinds of problems, why it would really matter. It would probably take too much time that you could use doing something else.

 

[surge55]

I see. You were trying to figure out how to estimate total body potassium deficit. I bet there is a way to estimate, though . . . I'm not sure, and some of this is just my way of approaching these kinds of problems, why it would really matter. It would probably take too much time that you could use doing something else.

http://spectrum.diabetesjournals.org/content/15/1/28.full

excellent article on the subject.

to answer the op, there is about a 3-5 meq/kg of body weight deficit. so if you really want to, you can precalculate how much potassium you want to give and mix it in the normal saline bags you are dumping into the patient.

the easiest way I've found is to follow along the alorithms listed above and get more frequent bmp's initially until you see a good response in the anion gap as well as see where the potassium levels are going. but this is another way if you like.

 

[Bostonredsox]

K > 5.5 I hold off on replacing, straight NS for volume
5.4 - 4.0 gets 20meq per L NS
3.9 - 3.0 gets 40meq per L NS
<3 gets 60meq per L NS
I do not give insulin until K is > 3

Straight out of wash manual of CC.
That DKA algorithm is fantastic. Simplistic, easy for nurses to follow, and has served me well throughout residency.

 

[Bostonredsox]

Wait..what?

Water follows glucose. Put glucose in cells water follows. 2/3rd fluid in body is intracellular (40% body weight). Thats why these folks need so much damn fluid. YOU metabolize ketones. Insulin restores cellular metabolism and inhibits further ketosis, meanwhile the body metabolizes the ones floating around. You dont just piss out all the ketones.

/nod. Glucose is an osmotic diuretic. Where it goes water follows. That's why HHNK 80 year old Gomers with BG of 1500 get liters on liters of fluid. The glucosiuria that ensues draws out massive amounts of intracellular volume. These pts can go into hypovolemic shock.

 

[jok200]

thanks to everyone... cleared it up perfectly.

 

[VentdependenT]

/nod. Glucose is an osmotic diuretic. Where it goes water follows. That's why HHNK 80 year old Gomers with BG of 1500 get liters on liters of fluid. The glucosiuria that ensues draws out massive amounts of intracellular volume. These pts can go into hypovolemic shock.

People seem to not understand this. They start the insulin drip and run NS at 200 an hour. I give 10units regular IV stat and dump in 4-5L up front as fast as possible and the glucose drops to 400...

 

[flipmd]

Don't be too aggressive on the insulin at first. Sugar of 1000 won't kill you. Hypovolemia will, as will the hypokalemia you induce with the insulin. I've stopped many ER docs from giving 10u of IV insulin when they "forgot" the K was 3.

 

[Bostonredsox]

Acidemia kills. Give fluids. Close the gap. Restore euvolemia. The sugar will follow. Insulin<<<volume in acidotic hyperglycemics. I've had ER docs here give 10 of insulin IV twice in an hour and call me for admission after only 500 of NS with 'the sugars down to 450 but pH is still 6.9" ....and they're kussmaul'n away with pressure starting to tank from the osmotic diuresis that hasn't been replaced with crystalloid....