Diabetic Ketoacidosis (DKA) Myths

Recently, I was asked to give a lecture to both my residents and nurses at the University of Texas Health Science Center at San Antonio (UTHSCSA) on some common DKA myths. Now this topic was originally covered by my good friend Anand Swaminathan on multiple platforms and I did ask his permission to create this blogpost with the idea of improving patient care and wanted to express full disclosure of that fact. I specifically covered four common myths that I still see people doing in regards to DKA management:

  1. We should get ABGs instead of VBGs
  2. After Intravenous Fluids (IVF), Insulin is the Next Step
  3. Once pH <7.1, Patients Need Bicarbonate Therapy
  4. We Should Bolus Insulin before starting the infusion

DKA Myths

Case: 25 y/o female with PMH of Type I DM who presents via EMS with AMS.  Per EMS report, the patient ran out of her insulin 3 days ago…..

  • Vital Signs: BP 86/52 HR 136 RR 30 O2Sat 97% on room air  Temp 99.1
  • Accucheck: CRITICAL HIGH
  • EMS was not able to establish IV access, so decided to just bring her to the ED due to how sick she looks.
  • Your nurses are on point today and get you two large bore 18G IVs and start to draw blood work to send to the lab.  You state I need a blood gas, and the nurse turns to you and asks do you need an ABG or VBG?

Myth #1: We should get ABGs instead of VBGs in DKA


So you do a literature review and come across two studies that specifically look at ABG vs VBG in an ED population:

Study #1: Kelly AM et al. Review Article – Can Venous Blood Gas Analysis Replace Arterial in Emergency Medical Care. Emery Med Australas 2010; 22: 493 – 498. PMID: 21143397

  • For pH, 3 studies of patients with DKA (265 patients) were reviewed showing a weighted mean difference of 0.02 pH units.  Only one study, which was the largest study (200 patients) reported 95% limits of agreement of pH that were -0.009 – 0.02 pH units.
  • For bicarbonate, 7 studies with just over 900 patients were reviewed showing a weighted mean difference of -1.41 mmol/L.   Only 2 studies (246 patients and 95 patients) reported 95% limits of agreement which ranged from -5.8 to +5.3 mmol/L.

Study #2: Ma OJ et al. Arterial Blood Gas Results Rarely Influence Emergency Physician Management of Patients with Suspected Diabetic Ketoacidosis. Acad Emerg Med Aug 2003; 10(8): 836 – 41. PMID: 12896883

  • ABG analysis changed ED physicians diagnosis in 1% (95% CI 0.3 – 3.6%) of patients
  • ABG analysis changed ED physicians treatment in 3.5% (95% CI 0.3 – 3.6%) of patients (Change from SQ to IV insulin or vice versa)
  • ABG analysis changed patient disposition in 2.5% (95% CI 1.1% – 5.7%) of patients
  • Venous pH correlated well with arterial pH with difference of -0.015 +/- 0.006 pH units

Other Reasons ABGs may not be preferred

  • Although rare, ABGs can cause radial artery spasm, infarct, and/or aneurysms
  • ABGs are painful to patients, even more so than IV access
  • By the way, when is the last time you checked a Modified Allen’s Test before doing a radial ABG?
    1. Block off Radial and Ulnar Artery for 30 seconds
    2. Have pt make fist
    3. Let go of ulnar artery and should have color return in <10seconds
    4. If no color return or delayed…then pt DOES NOT have collateral blood flow (i.e. – Allen’s Test)

Myth#1 Busted: VBG can be used in place of ABGs

Your VBG comes back and shows the following results:


You start IVFs and want to initiate insulin therapy next….

Myth #2: After Intravenous Fluids (IVF), Insulin is the Next Step

Patients with DKA will have total body potassium depletion from osmotic diuresis and electrolyte losses. Measured levels of serum potassium may be falsely normal or elevated due to extracellular shifts of potassium from secondary acidosis. The American Diabetes Association (ADA) actually recommends obtaining a a serum potassium level before initiating insulin, but this is based on anecdotal evidence. That being said, according to Arora et al, approximately 5% of patients with DKA will have hypokalemia.

If you look back to your VBG you will see your K+ is 2.8.  If you start insulin therapy before electrolyte replacement you you will worsen hypokalemia which is a very real cause of morbidity and mortality from cardiac arrhythmias and respiratory muscle weakness. Don’t forget that insulin will activate your Sodium-Potassium ATPase which will shift potassium intracellular and worsen your hypokalemia.


So the next real question is how should K+ be replaced? An oversimplified and general rule of thumb is that for every 10mEq of K+ given the serum K+ should rise by 0.1mEq, but at extremely low potassium levels this rule is not exactly true. I would also remember a few other things:

  • Place ALL patients on monitor for continuous ECG monitoring
  • Maximum Concentration for Peripheral IV administration = 10mEq/50mL
  • Maximum Concentration for Central IV administration = 20mEq/50mL

Every institution will have its own protocols for what is the acceptable rate of IV potassium repletion, but the first thing I would say is that if the Oral or Enteral route is available, use it.  Below I will try and describe how I replace potassium, but certainly other ways are acceptable, and every patient is different, but in general….

Clinical Bottom Line: After starting IVF, the next step in DKA management is electrolyte replacement, NOT Insulin.

Your nurse now approaches you and says that the patients pH is 6.9 so we need to give the patient sodium bicarbonate…

Myth #3: Once pH <7.1, Patients Need Bicarbonate Therapy

Consensus guidelines for the management of DKA recommended administering sodium bicarbonate to DKA patients who present with an initial blood gas pH of < 7.0. That recommendation was updated and changed in 2009 to limit sodium bicarbonate use to DKA patients with blood gas pH of < 6.9.

Screen Shot 2015-10-04 at 2.21.25 PM

Guidelines are nice, but they don’t keep up with research….Since the 2009 DKA guideline….2 more studies have come out evaluating the usefulness of sodium bicarbonate in DKA. There has been a systematic review published in 2011 and the largest retrospective analysis ever published in 2013.  We have covered both of these before on REBEL EM: Any Benefit to Sodium Bicarbonate in DKA?. To summarize:

Study #1: Chua et al. Bicarbonate in Diabetic Ketoacidosis – A Systematic Review. Ann Intensive Care 2011; 1 (23). PMID: 21906367

  • 44 studies of DKA patients reviewed which showed:
    • Transient improvement in metabolic acidosis
    • No improved glycemic control
    • Risk of cerebral edema in pediatric patients
    • No studies with pH <6.85

Study #2: Duhon et al. Intravenous Sodium Bicarbonate Therapy in Severely Acidotic Diabetic Ketoacidosis. Ann Pharmacother 2013. 47(7 – 8): 970 – 5. PMID: 23737516

  • Retrospective study of 86 patients with DKA which showed:
    • No difference in: Time to resolution of acidemia, time to hospital discharge, time on IV insulin, potassium requirement in 1st 24hrs
    • Subgroup Analysis of pH < 6.9 (n = 20) showed no statistical difference in time to resolution of acidemia

Clinical Bottom Line: Intravenous bicarbonate therapy may transiently make acidemia better, but there is no improvement of glycemic control, time on insulin, time to hospital discharge, and in kids can worsen cerebral edema.

So after potassium replacement your potassium is finally at an acceptable level so that you may start insulin therapy.  Your nurse asks you would you like to bolus the insulin before you start the infusion?

Myth #4: We Should Bolus Insulin Before Starting the Infusion

Again we have covered this on REBEL EM: Is There Any Benefit to an Initial Insulin Bolus in Diabetic Ketoacidosis? Again, I will summarize the key study here:

Goyal et al. Utility of Initial Bolus Insulin in the Treatment of Diabetic Ketoacidosis. J Emerg Med 2010; 38(4): 422 – 7. PMID: 18514472

  • Prospective, Observational Study of 157 patients with DKA:
    • Insulin bolus at the start of an insulin infusion IS EQUIVALENT to no insulin bolus at the start of an insulin infusion in several endpoints including:
      • Decrease normalization of glucose
      • Affect the rate of change of anion gap
      • Reduce ED or hospital length of stay
    • Insulin bolus at the start of an insulin infusion DOES:
      • Increase hypoglycemic events by 6 fold (6% vs 1%) [NOT Statistically Significant]


I generally just start my insulin drips at 0.05 – 0.14 U/kg/hr, only after electrolytes have been replaced.

Clinical Bottom Line: Insulin boluses increase hypoglycemic events without other clinical benefits in the treatment of DKA.


Clinical Bottom Line in DKA Management:

  1. VBG can be used in place of ABGs
  2. After starting IVF, the next step in DKA management is electrolyte replacement, NOT Insulin.
  3. Intravenous bicarbonate therapy may transiently make acidemia better, but there is no improvement of glycemic control, time on insulin, time to hospital discharge, and in kids can worsen cerebral edema.
  4. Insulin boluses increase hypoglycemic events without other clinical benefits in the treatment of DKA.

For More on this Topic Checkout:

Cite this article as: Salim Rezaie, "Diabetic Ketoacidosis (DKA) Myths", REBEL EM blog, November 2, 2015. Available at: https://rebelem.com/diabetic-ketoacidosis-dka-myths/.
The following two tabs change content below.

Salim Rezaie

Emergency Physician at Greater San Antonio Emergency Physicians (GSEP)
Creator & Founder of REBEL EM

Like this article?

Share on facebook
Share on Facebook
Share on twitter
Share on Twitter
Share on linkedin
Share on Linkdin
Share on email
Share via Email

Want to support rebelem?

24 thoughts on “Diabetic Ketoacidosis (DKA) Myths”

  1. I love how the managment of DKA changes so dramatically at the age of 16…
    In the UK paeds ED folk tend to avoid ABG’s due to the pain involved. We also always wait for an hour before starting any insulin. Interestingly bicarb has gone out of fashion and only tends to be used by intensivists who may ask for it in resus situatuions (in my experience). As for the insulin- we do it by weight (as we do for all things to do with kids) and don’t give a bolus.
    When I worked in Australia we used to use EtCO2 to predict if kids were in DKA http://www.ncbi.nlm.nih.gov/pubmed/17854452
    can be useful in newly diagnosed to give the ametop/ emla time to work before you cannulate in newly diagnosed diabetic- last thing you want is to increase needle phobia at this vulnerable stage.
    Fab summary- as always.

  2. I understand the low K issue but even though up to 5% may have low K, how many have significant arrythmias and death? Isn’t checking a K based on the ADA guidelines to check K before starting, which I think is just a guidelines recommendation–something you state don’t keep up with research in general. I am aggressive in starting insulin and have not had any issues with low K. I guess this has happened a lot to you?

    • Hello Paulie,
      Appreciate you reading and your question. So I stated 5% may have a low K+, but this does not account for all the patients who may have normal to low normal serum K+, which is a significantly higher proportion than 5%. Once you start the insulin and the K+ becomes low, it is near impossible to catch up. I advocate for IVF, check a K+ (Either VBG with electrolytes or iSTAT), then adjust your fluids accordingly before initiating the insulin drip. These patients are dehydrated and volume down, so starting with IVF and waiting 30 min max for K+ before starting the insulin drip is not going to kill the patient, but making them hypokalemic has the potential to kill them. Its more of a risk/benefit issue than it is an actual percentage of death. Why start something that could kill a patient, when you can safely wait 30 minutes to adjust your fluids and avoid that risk all together. Hope this helps.


  3. Salim, congratulations for your post, they are excellent and I follow you of course. I’m Luis Vargas Emergency Physician from Bogotá, Colombia. First promotion of the first Residency program in my city. I have a Podcast channel and I wanted to ask you for permission to translate and comment on your publication on my podcast channel in Spanish, thank you very much.
    Luis Vargas MD.


Leave a Comment

Time limit is exhausted. Please reload CAPTCHA.