DKA and the Anion Gap

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

JkGrocerz

Member
5+ Year Member
15+ Year Member
Joined
Nov 8, 2005
Messages
79
Reaction score
1
So i had a new onset DKA kid yesterday...

we started a iv bolus sent of a bmp among other labs. his sodium was 126....corrected for his hyperglycemia..his sodium is about 137. the computer calculated his gap to be 9 using 126 as the sodium.

when calculating the gap, should you use the 126 or the corrected sodium of 137? according to the computer his gap was already closed BEFORE we even started the insulin. Using the corrected Sodium, his gap is more like 21.

I'm thinking the latter. what do you guys think?

Members don't see this ad.
 
I think "DKA and the Anion Gap" sounds like a good name for a band.
 
  • Like
Reactions: 1 user
Members don't see this ad :)
you need to be very careful using the anion gap to guide your insulin therapy in DKA. The problem arises when the DKA patient receives large amounts of NS which contains Na and Cl. Because there will be a relative increase in Cl from the NS the gap Na-(HCO3+Cl) can look like it has closed just because of the extra Cl given for treatment and the patient can still be met acidosis. It is much better to follow the HCO3 as a guide, a general rule would be to use a value of 20 and stop the insulin at that value. just a thought that I just recently came across in an ICU patient.
 
We actually just discussed this in our joint Critical Care/EM conference (I presented a HHNC case). The consensus is that you do NOT use the corrected sodium (high or low) for patients when calculating the Anion Gap.
 
migraine12 said:
you need to be very careful using the anion gap to guide your insulin therapy in DKA. The problem arises when the DKA patient receives large amounts of NS which contains Na and Cl. Because there will be a relative increase in Cl from the NS the gap Na-(HCO3+Cl) can look like it has closed just because of the extra Cl given for treatment and the patient can still be met acidosis. It is much better to follow the HCO3 as a guide, a general rule would be to use a value of 20 and stop the insulin at that value. just a thought that I just recently came across in an ICU patient.

I was always told that to follow the HCO3 is folly because it is not mechanistic (i.e not cause/effect) in acid base balance theory...
 
migraine12 said:
It is much better to follow the HCO3 as a guide, a general rule would be to use a value of 20 and stop the insulin at that value.


DON'T STOP THE INSULIN!!!!!!!!!!!

That is how the patient got into trouble in the first place! Because the cells can't get glucose into them (no insulin) they start using free fatty acids- and putting out ketones. The only way to keep the person out of DKA is to keep the insulin going. If you need to, give glucose. DO NOT STOP THE INSULIN!!!!!!!!!!!!

While bicarb level may not be "mechanistic" it will ensure that the patient is out of acidosis as hyperchloremic acidosis can make it appear the gap is closed.

DO NOT STOP THE INSULIN!!!!!!!!!!!!!
 
roja said:
We actually just discussed this in our joint Critical Care/EM conference (I presented a HHNC case). The consensus is that you do NOT use the corrected sodium (high or low) for patients when calculating the Anion Gap.

so if you're going to use the uncorrected sodium, should you still expect to have a wide anion gap metabolic acidosis? The gap was calculated to be 9 Using the uncorrected sodium in this patient...which is normal :confused:
 
I happened to ask this question to a chief IM resident and Pulm/Critical care attending today. They both said to use the corrected sodium to calculate the AG.

Anyone have any literature on this? I'll try to look.
 
NKMU said:
I happened to ask this question to a chief IM resident and Pulm/Critical care attending today. They both said to use the corrected sodium to calculate the AG.

Anyone have any literature on this? I'll try to look.

I got my answer from Duke Internal Medicine, saying use the uncorrected.

We need an authoritative answer - not consensus.
 
Not sure about the literature, but I will ask around. However, the head of our ICU, two other ICU attendings and 2 fellows (plus several ED attendings) all concurred that you don't use the corrected sodium.

The explaination for this was that the 'corrected sodium' is used to gauge water def. However, in calculating the anion gap, you want the actual sodium, which is the uncorrected value. (representative of the true serum sodium)

Instead of the calculated which is to help gauge the water def.
 
Isn't the "actual sodium" in the serum, the corrected value with the uncorrected sodium representing a dilution of sodium secondary to osmotic shifts?
 
Members don't see this ad :)
I don't use the anion gap much, as it can be false for a number of reasons, and is unreliable when there are other coexisting electrolyte abnormalities (like from vomiting).

Generally I get a venous blood gas to check the pH. Venous blood doesn't differ that much from arterial in pH, and it's a helluva lot less painful. A venous blood gas of < 7.30 usually indicates an acidosis.
 
  • Like
Reactions: 1 user
JkGrocerz said:
So i had a new onset DKA kid yesterday...

we started a iv bolus sent of a bmp among other labs. his sodium was 126....corrected for his hyperglycemia..his sodium is about 137. the computer calculated his gap to be 9 using 126 as the sodium.

when calculating the gap, should you use the 126 or the corrected sodium of 137? according to the computer his gap was already closed BEFORE we even started the insulin. Using the corrected Sodium, his gap is more like 21.

I'm thinking the latter. what do you guys think?

First of all... I would highly discourage from EVER bolusing a child Insulin or IV Fluids. This is an easy way to iatrogenically cause cerebral edema, herniation, and death. Another caution with kids, is to NEVER change them over to .45 NS as you would in adults. The fluid deficit that occurs with hyperglycemia/DKA develops slowly over time and needs to be corrected slowly as well. The brain builds up pseudo-osms to compensate for the increased osmolarity in the brain (preventing fluid shifts w/ hyperglycemia which would otherwise cause the brain to shrivel up like a raisen). So if you correct their hyperglycemia too fast, or bolus too much fluids, the pseudo-osms will suck fluid into brain cells via osmosis and the kid will be 100x worse than when he was in DKA. The only time bolusing IVF is necessary is when you have a kid who is hypotensive and at risk for cardiovascular collapse. Use the Harriett Lange guide for calculating replacement fluids for dehydration based on meters-squared, and replace over 24 hrs. Or just use the 4:2:1 rule which is close enough (although it overshoots a little).

As far as anion gap goes... I agree.... use the uncorrected sodium. I've never heard of anyone ever using the corrected sodium, or read about that in the literature. If the gap is closed and the Bicarb is above 15, then it's not DKA.... just hyperglycemia or non-ketotic coma.
 
To summarize:

Use the uncorrected sodium to determine your AG
Start NS drip at large hourly rate, supplementing with insulin in either small doses or low drip (even in kids).
Supplement K+ from the beginning if the initial value is less than 4.5, otherwise add K+ approximately when glucose level halves.
Q1h glucose checks - goal is to reduce the glucose by around 100/hr
When glucose approaces 250 or so, add D5 to drip (in kids switch to D5 1/4 NS)
Let Simmer.
When Glucose approaces 200 OR remains stable between 200 and 250, re-check an AG (it should be closed by now).

Continue with maintenence fluid.

IF AG is not closed, continue with fluids, insulin, and D5, send an ABG...
 
I'm not sure why anyone would suggest to use a "CorrectED" value for AG, unless they may have just gotten the lingo (corrected, uncorrected, original, actual, adjusted, modified, etc.) misunderstood or represented in conversation. I'm no expert, but I believe the AG is indeed interested on actual Na present, not the value modified to account for hypo/hyperglycemia, etc. :)
 
NinerNiner999 said:
Continue with maintenence fluid.

IF AG is not closed, continue with fluids, insulin, and D5, send an ABG...


AAAARRRGH!!!!!! Continue with insulin even if the gap is closed!!!!

You can change them over to SQ when the gap is closed/NaCHO3>20.

Good point about starting to replace the K. Also, MG is often low.
 
you need to be very careful using the anion gap to guide your insulin therapy in DKA. The problem arises when the DKA patient receives large amounts of NS which contains Na and Cl. Because there will be a relative increase in Cl from the NS the gap Na-(HCO3+Cl) can look like it has closed just because of the extra Cl given for treatment and the patient can still be met acidosis. It is much better to follow the HCO3 as a guide, a general rule would be to use a value of 20 and stop the insulin at that value. just a thought that I just recently came across in an ICU patient.

Forgive my ignorance. I am only an RN, but did my time in ICU enough to understand the conversation. And yes, my experience is limited(I am no CCRN),but I also was a chemist before I was a nurse. So when I can apply my core science to my trade.. it brightens my day - beats the hell out of passing so many freakin pills all day!

anions - negatively charged ions.
cations - positively charged ions.

Acid/base also fall on this- not just the 0.9% NS they may have been receiving.... or any other salt.

rules; ROME. DKA= metabolic acidosis. I have metabolic acidosis, my respiratorations will be rapid and shallow do "breathe off" the carbon or rather... 'create bicarb.' Fruity breath.

Of course.... they will be urinating like no tomorrow, thirsty, and hungry.

Im going to presume we already have ABG's going as a matter of rule and are giving insulin, per order, watching blood PH, as well as anion gap.

(If I recall correctly the CCRN I was working with was pointing directly to the HCO3 levels for guidance on insulin therapy DKA)

HCO3- as DKA is an acidic state, I will expect it to be low. Since the kidneys may take time ... vs the respiratory... I expect a high RR.

Remember the thing that caused the DKA to begin with.

CAUSES FIRST.

If DKA was caused by an infection rather than just noncompliance... this could be an entirely different beast. the osmolality is also a concern.

As if High Blood sugar was the problem - as we know... it isnt a life threatening problem unless it triggers DKA. Thing of it is... 99.9% of the time we are gonna see those high sugars on our Type I- the type II... we arent monitoring if they are asymptomatic - but that depends on the level of care as well.

Now imagine my concern knowing all this... when another RN tells me just to keep on giving insulin- and not even send the patient to the ICU???!

Plus "more insulin" can turn into problem - it will not always be the answer. The entire reason we are watching the HCO3 levels is to ensure we d/c insulin at the proper time (IV being a little more direct than SC). Probably another reason we should be doubly sure to check for any kind of infection as a cause in the first place.
 
^^
be7.jpg
 
  • Like
Reactions: 1 user
Nice necro bump but yes that post was horrendous

Sent from my LG-D801 using Tapatalk
 
  • Like
Reactions: 1 user
Forgive my ignorance. I am only an RN, but did my time in ICU enough to understand the conversation. And yes, my experience is limited(I am no CCRN),but I also was a chemist before I was a nurse. So when I can apply my core science to my trade.. it brightens my day - beats the hell out of passing so many freakin pills all day!

anions - negatively charged ions.
cations - positively charged ions.

Acid/base also fall on this- not just the 0.9% NS they may have been receiving.... or any other salt.

rules; ROME. DKA= metabolic acidosis. I have metabolic acidosis, my respiratorations will be rapid and shallow do "breathe off" the carbon or rather... 'create bicarb.' Fruity breath.

Of course.... they will be urinating like no tomorrow, thirsty, and hungry.

Im going to presume we already have ABG's going as a matter of rule and are giving insulin, per order, watching blood PH, as well as anion gap.

(If I recall correctly the CCRN I was working with was pointing directly to the HCO3 levels for guidance on insulin therapy DKA)

HCO3- as DKA is an acidic state, I will expect it to be low. Since the kidneys may take time ... vs the respiratory... I expect a high RR.

Remember the thing that caused the DKA to begin with.

CAUSES FIRST.

If DKA was caused by an infection rather than just noncompliance... this could be an entirely different beast. the osmolality is also a concern.

As if High Blood sugar was the problem - as we know... it isnt a life threatening problem unless it triggers DKA. Thing of it is... 99.9% of the time we are gonna see those high sugars on our Type I- the type II... we arent monitoring if they are asymptomatic - but that depends on the level of care as well.

Now imagine my concern knowing all this... when another RN tells me just to keep on giving insulin- and not even send the patient to the ICU???!

Plus "more insulin" can turn into problem - it will not always be the answer. The entire reason we are watching the HCO3 levels is to ensure we d/c insulin at the proper time (IV being a little more direct than SC). Probably another reason we should be doubly sure to check for any kind of infection as a cause in the first place.

Uhhhhh......what?
 
Anybody still using insulin bolus for their DKA?

Someone correct me if I'm wrong, but I've never been able to find an actual original article evaluating this.

There is an article saying large amount of insulin early is a bad idea, and as far as I can tell, that's the origin of the "no bolus in peds" came from; but I don't think anyone has actually studied this.
 
Someone correct me if I'm wrong, but I've never been able to find an actual original article evaluating this.

There is an article saying large amount of insulin early is a bad idea, and as far as I can tell, that's the origin of the "no bolus in peds" came from; but I don't think anyone has actually studied this.

http://rebelem.com/benefit-initial-insulin-bolus-diabetic-ketoacidosis/#ITEM-1414-1

http://www.emdocs.net/myths-dka-management/

http://www.ncbi.nlm.nih.gov/pubmed/18514472

http://www.ncbi.nlm.nih.gov/pubmed/7587857

http://www.ncbi.nlm.nih.gov/pubmed/18694978

http://www.ncbi.nlm.nih.gov/pubmed/19564476

I was trained to not bolus and only CII. I think we need even more studies but I've yet to see any persuasive data to change my practice.
 
Someone correct me if I'm wrong, but I've never been able to find an actual original article evaluating this.

There is an article saying large amount of insulin early is a bad idea, and as far as I can tell, that's the origin of the "no bolus in peds" came from; but I don't think anyone has actually studied this.
Independently correlated with cerebral edema in peds. No data that shows a decrease in time to resolution.
Remember, the problem isn't the glucose, it's the acid. It doesn't matter (until it is low) what the glucose is, so stop trying to bolus it down. You need steady reduction of ketosis. There are case reports of intractable asystole from insulin boluses, so you can wait until the K before giving it anyway.
Also, use LR, for the love of god.
 
I'll bolus insuli. if there are EKG changes of hyper kalmia. If the qrs is wide, I treat w/ bicarbonate as well (in adults, probably wouldn't on kids unless frankly in cardiovascular collapse)--obv in addition to calcium. Otherwise never bolus insulin, just no point and leads to higher incidence of hypokalemia. In general my concerns in DKA are electrolytes > precipitating condition > acidosis > glucose.

Im really unimpressed with the correlation of fluid blouses with cerebral edema, all the data is terrible, and I think the largest of the rectrospective reviews didn't find an independant correlation between fluid blouses and CE. However, the peds thought leaders tend to really argue against it, so while in the past I've usually given a 10 ml/kg bolus I'm now going to basically just use 1.5mivf in the absence of de compensated shock.
 
I'll bolus insuli. if there are EKG changes of hyper kalmia. If the qrs is wide, I treat w/ bicarbonate as well (in adults, probably wouldn't on kids unless frankly in cardiovascular collapse)--obv in addition to calcium. Otherwise never bolus insulin, just no point and leads to higher incidence of hypokalemia. In general my concerns in DKA are electrolytes > precipitating condition > acidosis > glucose.

Im really unimpressed with the correlation of fluid blouses with cerebral edema, all the data is terrible, and I think the largest of the rectrospective reviews didn't find an independant correlation between fluid blouses and CE. However, the peds thought leaders tend to really argue against it, so while in the past I've usually given a 10 ml/kg bolus I'm now going to basically just use 1.5mivf in the absence of de compensated shock.
Eh, fluid boluses aren't strongly correlated with cerebral edema. It can even happen before treatment.
 
I got my answer from Duke Internal Medicine, saying use the uncorrected.

We need an authoritative answer - not consensus.

Correct. The data and most recent endocrinology guidelines are to use the uncorrected gap.

My understanding from an endocrine Lecture at a hospitalist conference I recently attended is The sodium level is a concentration, whereas the gap is a calculated number of measured cations and anions. That number does not change when the circulating plasma volume does, but the total amount of diluent does, and thus the concentration of the cations and anions change. the absolute number of positive and negative charges present (Na, Cl, HCO3) used to calculate the gap is the same.
 
I got my answer from Duke Internal Medicine, saying use the uncorrected.

We need an authoritative answer - not consensus.

Correct. The data and most recent endocrinology guidelines are to use the uncorrected gap.

My understanding from an endocrine Lecture at a hospitalist conference I recently attended is The sodium level is a concentration, whereas the gap is a calculated number of measured cations and anions. That number does not change when the circulating plasma volume does, but the total amount of diluent does, and thus the concentration of the cations and anions change. the absolute number of positive and negative charges present (Na, Cl, HCO3) used to calculate the gap is the same.

You use the uncorrected, but not for the reason above.

Bicarb and chloride are diluted by the glucose just as much as the sodium, so if you were to use the corrected sodium you'd have to also use the corrected chloride... and they'd pretty much exactly cancel it out.

And yes... (except when the potassium gets too low) the answer for someone who is in DKA/HHS (or recently was) is always to continue the insulin. The IV insulin can't be stopped until 1-2 hours after they receive subq (to allow for adequate absorption)
 
So my answer is correct, the anions and cations are equally diluted as I said, they're concentration goes down, but there total numbers and thus the numbers use to calculate the gap, are the same. You are saying the same thing I am
 
You guys are killin' the acid base stuff. Rock on pH rebels.
 
  • Like
Reactions: 1 user
Top