Should I give bicarbonate in DKA?

Should I give bicarbonate to DKA patients with severe acidemia? I’ve certainly been admonished for NOT doing it. The reason for withholding bicarb has been that I’ve heard that it doesn’t help and may actually be a bad idea. I can’t say the action (or inaction) was based on a deep understanding.


Direct Download


How could bicarb in DKA be a bad idea if even the American Diabetes Association (ADA) recommends we give a bicarb to DKA patients with pH under 6.9? The argument in favor of giving bicarb is that the more acidemic the patient, the higher the risk of circulatory collapse and cardiac arrest. Even though there is no evidence of benefit, the ADA gives a very specific set of steps to take in the low pH patient..

  • Because severe acidosis may lead to numerous adverse vascular effects, it is recommended that adult patients with a pH less than 6.9 should receive bicarbonate. Specially 100 mmol sodium bicarbonate, two ampules, in 400 mL sterile water with 20 mEq KCL admitted at a rate of 200ml/hr for 2 hours until the venous pH is over 7. If the ph isn’t over 7 at that point, they say repeat the bicarb infusion every 2 hours until the ph is over 7.0

With that sort of exact guidance, you’d think there would be evidence to back it up, but here is the sentence that precedes the above recommendation.

  • No prospective randomized studies concerning the use of bicarbonate in DKA with pH values <6.9 have been reported.

Because of the lack of evidence, the UK guidelines say this

  • Adequate fluid and insulin therapy will resolve the acidosis in diabetic ketoacidosis and the use of bicarbonate is not indicated 

But as the saying goes, “absence of evidence is not evidence of absence”, so is there a downside to giving bicarb to DKA patients? It turns out there there may be. Several FOAMed bloggers have tackled this in great detail: Pulm Crit, REBEL EM, Life in the Fast Lane, emdocs, and Jacobi EM. (just to name a few)


Here are just of few of the problems with bicarb in DKA patients


Giving bicarb drives potassium into the intracellular space.

  • DKA patients are total body potassium depleted.
  • Once the IV fluid and insulin get going the potassium is likely to drop quickly. In a patient already at risk for hypokalemia, administration of bicarb can drop the serum potassium even faster.

Does bicarbonate infusion in DKA improve outcome?

  • The preponderance of evidence, albeit small numbers of patients, suggests that bicarb does not improve outcome, even in those with low pH.
  • The most widely cited article on this is a 2011 systematic review from  Annals of Intensive Care that found no evidence of benefit for either neurologic or hemodynamic outcome. There was some evidence of a transient improvement in acidosis with the first 2 hours but no evidence of clinical efficacy.

Bicarb slows ketone clearance.

  •  A 1996 study found that giving bicarb slowed the clearance of ketones and AND  transiently increased acetoacetate and beta hydroxybutyrate levels. 

Bicarb may cause CSF acidosis. This goes back to a 1967 study by Posner and Plum.

  • A series of 7 severely acidotic patients.
  • Some were obtunded and some weren’t 
  • The authors postulated that it’s the degree of CSF acidosis that determines coma more than peripheral acidosis.
  • To study this, whenever blood ph was studied, they did a neurologic exam and a lumbar puncture.
  • Lower CSF pH correlated with a lower level of consciousness.
  • In 2 patients with DKA, they found that giving IV bicarbonate infusion, while it improved serum pH, was associated with more acidotic CSF.
  • Other studies have called the importance or even validity of bicarb infusion causing CSF acidosis into question and found treating DKA how we regularly do can itself cause the CSF pH to transiently drop.

George Willis, ED doc and DKA expert, uses bicarb in DKA in three scenarios

  • DKA with cardiac arrest
  • Persistent hypotension despite vasopressors
  • Hyperkalemia with arrhythmia

So should ANY DKA patients get bicarbonate? I think there are several choices

  • You can follow the US/ADA guidelines and use bicarb if the pH is under 6.9. This is not based on solid evidence, more-so the worry that severe acidemia can lead to circulatory collapse (and bicarb may mitigate that)
  • You can follow the UK guidelines and just not give bicarb at all
  • I like the Willis rule of 3. Hyper K with arrhythmia, severe hypotension despite pressors, cardiac arrest -because these are patients who are about to die. With hyper-K, you might push just enough potassium into the cells to make a difference. In cardiac arrest, you might give a quick bump up in pH to improve the cardiovascular situation, then again, you might not.

Mentioned in the intro



Kitabchi, Abbas E., et al. “Hyperglycemic crises in adult patients with diabetes.” Diabetes care 32.7 (2009): 1335-1343.

Dyer, P. H., and M. S. Hamersley. “Diabetes UK Position Statements and Care Recommendations Joint British Diabetes Societies guideline for the management of diabetic ketoacidosis.” (2011).

Chua, Horng Ruey, Antoine Schneider, and Rinaldo Bellomo. “Bicarbonate in diabetic ketoacidosis-a systematic review.” Annals of intensive care 1.1 (2011): 23.

Savage, M. W., et al. “Joint British Diabetes Societies guideline for the management of diabetic ketoacidosis.” Diabetic Medicine 28.5 (2011): 508-515.

Okuda, Y. U. K. I. C. H. I., et al. “Counterproductive effects of sodium bicarbonate in diabetic ketoacidosis.” The Journal of Clinical Endocrinology & Metabolism 81.1 (1996): 314-320.

Posner, Jerome B., and Fred Plum. “Spinal-fluid pH and neurologic symptoms in systemic acidosis.” New England Journal of Medicine 277.12 (1967): 605-613.



  1. aimwendels

    I don’t even check gases in DKA anymore. I you have the glucose and you and the BMP why ask questions you don’t want answer to? When I was a CC fellow I always taught my residents that bicarb in DKA was poison.

    1. Post
      Rob Orman

      Agreed. I have never seen pH influence management (except it pushes people to give bicarbonate!) The place where I’ve found vbg useful in the emergency Dept is when it’s not clear what’s going on with the patient and a vbg will give serum bicarbonate level in minutes, compared to the 45-60 minutes it takes with central lab.

    2. David P

      It can help you rule out more complex acid/base disorders as well as determine appropriate respiratory compensation.

  2. Charles

    On third bullet point on bottom (Should any DKA get bicarb), second sentence, I think you mean hyper-K, not hypo-K- correct?

    1. Post
  3. David P.

    Rob -- Big fan of your work. Great podcast. Thanks for everything you do.

    This brings up an bigger issue in my mind: Why do ER docs and intensivists defer management to specialists because the problem technically falls within the purview of their speciality? Endocrinologists may give us recommendations and guidelines regarding DKA, but they don’t actually care for sick patients with DKA; and I’m not sure that taking care of ambulatory patients in the clinic makes them the expert for an ED/ICU topic. Same goes for a lot of other topics that are primarily managed by the ED and ICU.

  4. Chris

    Recently had a young DKA patient with a nonhemolyzed K+ of 8.4 (and glucose of 1970), QRS widened slightly, peaked Ts, started insulin drip and gave some calcium. Hours after admission (pt had left the department) I found out the patient had suffered a cardiac arrest. The documentation was not clear as to initial rhythm, etc but I did wonder whether or not bicarbonate should have been administered in the ED given the extent of hyperkalemia and EKG changes, albeit with no documented arrhythmias as mentioned above. That being said, this has been the only time I’ve thought twice about administering bicarb to a DKA patient.

Awesome article, I know - please share your erudite thoughts...