How NOT to Mess Up DKA - Metabolic Acidosis Part 1


It's super easy to screw up your DKA patient. Whether you are the treating physician or the transporting paramedic, you need to understand all the steps in DKA treatment. In DKA, proper timing of treatments and careful observation in a must. Let's break down this process step-by-step.

Insulin should probably not be your first treatment

Giving insulin as first line treatment in the DKA patient can seriously harm them. Why?

Insulin brings water into the cells. Your DKA patient is already likely very dehydrated. When glucose builds up in our blood, our kidneys want to waste it in an attempt to offload glucose and normalize BGL values. When the kidneys waste glucose, what follows? Water. Polyuria (production of abnormally large volumes of dilute urine) is a sign of increased blood glucose and glucose wasting by the kidneys. Polyuria = dehydration. Therefore, giving insulin as a first line treatment will further deplete intravascularly available fluid, and may lead to hypovolemic shock and cardiovascular compromise or collapse.

Insulin will lower K+ levels. Insulin moves potassium intracellular. The patient may have a normal potassium right now, but as acidosis -which we will talk about in a minute- starts to correct, so does the potassium. This means that your patient with a potassium of 3.5 might go into a hypokalemic arrest within the hour if acidosis and blood glucose levels are corrected without prophylactic K+ replacement. Bad form! Always take renal function into consideration when adding potassium replacement as well.

This hypokalemia is usually a later sign of the disease process. You may catch the patient early when they are in an initial hypERKalemia. Watch the potassium closely in this case, since the patient most likely has a whole body potassium depletion. If the initially hyperkalemic patient is getting down to a borderline low potassium level, it's time to strongly consider prophylactic potassium replacement.

Take home point: Resist the urge to treat DKA with insulin until the need for appropriate fluid and K+ replacement has been evaluated.

Remember why your patient is acidotic

Why is your DKA patient acidotic in the first place? Ketones, and possibly lactic acid. Where do ketones come from? Mostly lipolysis (the breakdown of fat for energy). This process produces a byproduct called a ketone. Ketones are acidic, so the buildup of ketones cause ketoacidosis.

Why did these ketones start to form? The patients cell normally run on acetyl-coA entering the Krebs cycle. But now, due to a lack of insulin, fatty acids have entered the Krebs cycle and are being broken down into ketone bodies. These ketone bodies include:

- Acetone (which can sometimes be smelled on the patient's breath)

- 3-hydroxybuyyrate (which can be tested for by blood draw)

- And acetoacetate

These ketone bodies can be evaluated by direct lab draw or simply by evaluating the patient for an anion gap metabolic acidosis (DKA is a cause of a high anion gap metabolic acidosis).

Stopping ketone production is the reversible cause for this acidosis. This will prevent us from giving unneeded sodium bicarbonate. How do we stop ketone production? By allowing the cells to feed on glucose via insulin administration AFTER administration of fluids and K+ to prevent cardiovascular and hypokalemic collapse.

The stopping point of our therapy depends heavily on the presence of these ketones in the blood, not the blood glucose level. More on that in a little bit.

Take home point: Correcting the blood glucose is not our main concern. Stopping the process of ketone production (and therefore acidosis) is our goal.

Correcting hyperglycemia is easy to oversimplify

Lets check our progress so far.

1. We have identified our patient has absolute hypovolemia and started appropriate fluid replacement.

2. We have anticipated the drop in serum K+ that results from raising the pH (potential hydrogen) and giving insulin, and have started prophylactic K+ replacement if needed.

3. We have remembered why our patient is acidotic in the first place and not given any unnecessary treatments.

4. We have identified the end point of our therapy as ketone elimination, not glucose reduction.

Now we are on the home stretch. Our fluids are going in, our K+ is infusing, the insulin is infusing, all we have to do now is wait.......

2 pm

BGL: 350

3 pm

BGL: 175

3:01 pm

BGL: 174, and the patient is having a full blown seizure. 'Ah crap, oops.'

What went wrong??

Don't mess up this late in the game. Lowering BGL lowers your serum osmolality. I always remember it like this... "SOLUTES SUCK!" Alright... What does that mean? Solutes are like magnets for water, and these water magnets hold the water in your blood (and therefore in your veins and arteries). There are a few different magnets at work all the time (like sodium, glucose, and blood urea nitrogen/BUN). When the patient's blood glucose was high, they had a lot of 'magnets' holding water in their blood vessels. When we drop the BGL too fast (like more than 75-100/hour) we can cause a rapid shift of water out of the blood vessels. This would be fine, if our patient didn't have a skull. Unfortunately, lots of our patients have skulls. Our brain is in a fixed container (think of the monro-kellie doctrine). When we add fluid to that container too rapidly, it does not allow time for other fluid or matter to move out of the way. The result? Cerebral edema and seizures to follow. Giving the 13 year old diabetic kid a seizure in front of his parents is a sure way to look and feel incompetent.

When the BGL is falling too fast, slow down your insulin infusion, give dextrose, or both if necessary.

Once the patient gets down to a BGL under 300, it is likely time to start a dextrose infusion along with your insulin. Sounds stupid, right? Not if you've been following along. These last couple hundred of BGL you need to lower is critical, and you can't drop too fast at this stage because the serum osmolality is pretty normal at this point. There is a high risk of cerebral edema. Also, giving dextrose at this point is not stupid if you think about why you're giving the insulin in the first place. You are giving the insulin to stop ketone production (insulin stops the fatty acids from entering the cells). Insulin needs to continue until ketone bodies clear. How can we tell when the ketone bodies clear? Check out the next section.

Take home point: Dropping the blood glucose level more than 75-100 / hour can cause cerebral edema and seizures.

The importance of the anion gap in DKA

Checking the anion gap is the main way we will be able to tell if the ketone bodies have cleared. Once the ketone bodies have cleared, the metabolic acidosis may be almost resolved. It may take time for their bicarbonate levels to return to normal, but the anion gap is what you are after at this point in treatment.

How do we assess the anion gap? Check out the chart below.

Take home point: Closing the anion gap is the main endpoint in our treatment, not normalization of blood glucose. We may even need to continue insulin therapy along with a dextrose infusion to accomplish this.

Other Great FOAMed Resources:

EMCRIT has lots of stuff on acid-base including a 4 part lecture series and a flow chart.

https://emcrit.org/wp-content/uploads/acid_base_sheet_2-2011.pdf

PharmacyJoe has great information on dosing and important treatment timing considerations.

https://www.pharmacyjoe.com/diabetic-ketoacidosis-dka-and-hyperosmolar-hyperglycemic-state-hhs/

MedCram lectures give a great visual of DKA

https://www.youtube.com/watch?v=yIc2XFNLhm8

Rosen's Emergency Medicine summary of DKA:

Here is a summery box from chapter 126 in Rosen's Emergency Medicine. This is for reference only, and in no way replaces the required reading of the entire chapter.

#DKA #FOAMfrat #MEtabolic #Acidosis #Diabetic #ketoacidosis #aniongap #hyperglycemia