After evaluating this article, participants will be able to:
1. Incorporate strategies into practice for the most effective use of insulin in DKA
2. Develop and implement strategies for earlier detection of DKA
3. Incorporate strategies for the management of DKA which result in less complications
Managing diabetes, in particular DKA, is a staple in the practice of emergency medicine. The pancreas hasn’t changed since we discovered that it secreted insulin. So, the treatment has no reason to change either. Right? Well, this statement would be true if what we do was right in the first place. Passed from one generation to the next are a few concepts that have never had support in the literature. More importantly, these concepts have been disproved, but the undoing of bad teaching is difficult at best. We must first question the teachings of our mentors and then pursue the truth ourselves. Many of the most influential educators in each of our developmental history may have led us a stray, not intentionally, of course. However, despite the fact that the evidence has not supported certain actions, our mentors and educators followed the mantra, “Frequently wrong, but never in doubt.” Although they didn’t know they were wrong, nonetheless, their conviction to their teachings and their opinions, coupled with our desire to follow the words of our respected leaders, has led many of us down a path of dogma without a single bread crumb to lead us back to the evidence and more appropriate management.
Although these misconceptions may not be universal, many EPs have struggled with the issues of sodium bicarbonate use in metabolic acidosis for DKA and what the best option for insulin administration really is.
Despite the fact that many admitting physicians question our clinical acumen when sodium bicarbonate is not used for DKA-associated metabolic acidosis and we often fold to the pressure of a low pH, the evidence has been resoundingly clear for decades. There is no place for bicarb in the management of DKA, particularly in pediatrics. In review of the literature, there are some truly compelling facts that I’d like to share with you.
First, just like any other form of metabolic acidosis, no data exists that has shown better outcomes by administering sodium bicarbonate to patients with DKA. As a matter of fact, they may do worse. Green, et al., reported in 1998 that pediatric patients in DKA have no improvement in metabolic recovery when bicarb was given. In addition, in 147 admissions of 107 children with DKA, 57 were never given sodium bicarbonate, despite the fact that 9 had pH less than 7.0 and one had a pH of 6.73. Per multivariate analysis, the duration of hospitalization was 23% (16 hours) longer than those not receiving sodium bicarbonate. In other words, the patients who received bicarb required longer hospitalizations. However, they didn’t receive bicarb because they were sicker. (Green SM. Failure of adjunctive bicarbonate to improve outcome in severe pediatric diabetic ketoacidosis. Ann Emerg Med. 1998 Jan;31(1):41-8.). Morris reported no benefit or harm in a small retrospective review of 21 patients with DKA and a pH ranging from 6.90-7.10, and Viallon, in Critical Care Medicine in 1999, reported that 39 consecutive DKA patients (24 treated with bicarb and 15 without) with pHs between 6.90 and 7.10 had similar outcomes. Again, there was no benefit to bicarbonate administration. In addition, Dr. Glaser reported that their investigators identified 61 pediatric DKA patients that had developed cerebral edema. Several potential complications from the use of sodium bicarbonate were identified: Hypokalemia, paradoxical CNS acidosis, impaired tissue oxygenation and cerebral edema. Bicarb was the only treatment option that positively correlated with the development of cerebral edema. Even the rate of fluid administration failed to demonstrate such a correlation. (N Engl J Med 2001; 344: 264– 269.). In review of the literature, I came across a very interesting article that actually demonstrated that bicarbonate administration results in increased ketone production ( J Clin Endocrinol Metab. 1996;81: 314 – 320.721 – 727.). Not only does this therapy not offer any therapeutic benefit whatsoever, it is associated with serious complications and actually worsens the pathogenesis of DKA. That’s why we should be calling this drug “Badcarb!”
Bolus or not? When managing DKA, should you just start the drip or give a bolus first? Historically, a bolus was always given. It seems to make sense. Give a larger dose of insulin to get a bigger and earlier bang for your buck. Interestingly, it was commonplace in the 1970s to give as much as 100 units. Well, if a little is good, more must be better, but, the literature has been clear since the early 1980s that a bolus isn’t just unnecessary, it’s actually harmful. DKA is usually a young person’s disease, and thus, much of the data is from the pediatric literature.
However, extrapolating the answer to this question to older patients in DKA seems very reasonable. Fort, et al. reported that in 20 episodes of DKA in 19 children a bolus prior to a low dose infusion (0.1 unit/kg/hr), compared to the infusion alone, showed no additional benefit and was noted to be potentially harmful (The Journal of Pediatrics - January 1980 (Vol. 96, Issue 1, Pages 36-40.). Although this concept was called into question over three decades ago, many practitioners just can’t break the habit. I checked the more recent literature to see what is being published currently. We haven’t listened. So, researchers keep sending the message. An article published by Goyal in 2010 showed no benefit to an insulin bolus compared to the 0.1 unit/kg/hr infusion alone. Although there was no reported harm from those receiving a bolus, there was no benefit as evidenced by no difference in the amount of intravenous fluids required, rate of glucose change, normalization of the anion gap or length of stay ( J Emerg Med. 2010 May;38(4):422-7.). This is a worthless step, and some pediatric investigators are suggesting that an ultra-low dose will work just as well as the standard infusion (0.05 units/kg/hr).
Recently, while reviewing the literature, I came across a concept that is very intriguing, capnography for the prediction of DKA. Capnography should be ubiquitous to every ED. The indications and applications include monitoring for procedural sedation, invasive mechanical ventilation and non-invasive mechanical ventilation (BiPAP). These alone create a cost-benefit argument that is hard to beat. Add another indication, like the prediction of DKA, and this is a true slam-dunk. Nasal capnography was utilized to track 58 type one diabetics. 26% (15) developed DKA. None of the 15 had an end-tidal CO2 greater than 30 (J Paediatr Child Health. 2007 Oct;43(10):677-80.). Although these numbers are small, I find them very compelling. Particularly with a non-invasive device that we should have anyway, this application makes a lot of sense and is an adjunct to diabetes management that I’m anxious to try.
Diabetes is an old disease. Perhaps, becoming too familiar with an old foe leads to complacency in our willingness to challenge the dogma that surrounds its management. The literature is speaking. Are you listening?
Dr. Kevin Klauer is the editor-in-chief of Emergency Physicians Monthly, the CMO of Emergency Medicine Physicians, and the vice speaker of the ACEP Council.