Euglycemia DKA – don’t miss it!

EM Reflections May 2021 – Euglycemia DKA

Big thanks to Dr. Paul Page for leading this month’s discussions.

All cases are imaginary but highlight important learning points.

Authored and Copyedited by: Dr. Mandy Peach


A 65 yo female presents with n/v ongoing for 2 days. She feels fatigued and has not been able to keep down fluids. She denies diarrhea. She has no history of abdominal surgeries. She does describe increasing productive cough that preceded the vomitting. She denies fever, but does complain of shortness of breath.

Medications: Atorvastatin, Empagliflozin, Pantoprazole

Vitals: BP 104/66 HR 110 RR 22 O2 96% RA T 37.8 gluc 7.2

On exam there is obvious dehydration, and she seems fatigued with her eyes closed through most of the exam. She does respond to speech. The abdominal exam is unremarkable for focal tenderness. There are expiratory crackles heard at the R lung base.

You order a portable CXR1 and baseline labs including a VBG and lactate.

You suspect pneumonia with dehydration. You initiate a 1L NaCL bolus and order antibiotics.

You continue seeing other patients when you get a call from the nurse – the VBG is back for the patient.
They appear to have a metabolic acidosis with a pH of 7.10 and an anion gap of 14. The lactate appears surprisingly normal. The patient hasn’t made any urine yet for a sample.

What is the differential for anion gap metabolic acidosis2?

Going through the ‘MUDPILES’ mnemonic and revisiting the history nothing seems to fit. But there is a history of DM.

What red flag should trigger you to consider DKA despite the normal glucose?

The patient is on Empagliflozin. This is a SGLT-2 inhibitor. Patient on these medications are at risk of Euglycemic DKA.

In Euglycemia DKA there is a “relative carbohydrate deficiency state with normalization of serum glucose and concomitant elevation of counter-regulatory stress hormones. This leads to free fatty acid catabolism and ketone production.” 3

In any patient on a “zin” consider euglycemic DKA.

You order a serum ketone as well as β-hydroxybutyrate.

Clinically how do patients present with euglycemic DKA3?

Nausea/vomiting, malaise, shortness of breath – the differential is huge for this presentation. Again, look at the medication list for any diabetic patient. If you see a ‘zin’ – consider euglycemic DKA.

Alternatively if you order a gas and incidentally find anion gap metabolic acidosis in a diabetic patient consider ordering ketones/ β-hydroxybutyrate.

What about if this patient was an alcoholic? How would these complicate the diagnosis4?

Alcoholics can also present quite similarly with alcoholic ketoacidosis – nausea/vomiting, malaise, and similar lab findings. Other than the history one distinguishing characteristic is that alcoholic ketoacidosis tends to have frankly low blood glucose.

Are the triggers for euglycemic DKA any different3?

No, triggers for DKA are the same. Essentially any physiological stress.

A quick way to remember is the 5 I’s

Infection pneumonia, UTI, skin, abdominal
Infarction MI, CVA, bowel infarction
Infant on board pregnancy
Indiscretion dietary nonadherence
Insulin deficient insulin pump failure or non-adherence

Infection and insulin deficient secondary to non-adherence or inappropriate dosing are the most common causes.

I would also consider adding a 6th I – iatrogenic meaning drugs

What drugs commonly trigger DKA3?
– Glucocorticoids
– Diuretics
– Atypical antipyschotics

Are there any patients at risk of euglycemic DKA other than those taking the ‘zins’3?

  • Pregnant patients -due to high placental glucose use they can have a relative euglycemia
  • Chronic pancreatitis
  • Bariatric surgery patients – absorption issues

Your patient was straight cathed for a small amount of urine which shows ketones. The beta-hydroxybuterate is also now back and is positive. You confirm euglycemic DKA.

You grab your nearest DKA algorithm to review with the nurses and begin treatment.

Besides ease of use, what are the clinical reasons for using a standardized DKA order set?
Standardized, evidence based DKA order sets have been shown to decrease time to closure of anion gap, reduce length of hospital admission and minimize complications during treatment3.

You get started with the treatment as per the order set. While treatment is commenced you sit down with your medical student and review the goals of DKA3.

Correct fluid deficits – patients in DKA get a osmotic diuresis from hyperglycemia, or dehydration from underlying illness. You want to restore volume before initiating insulin. This improves organ perfusion, renal function and lowers lactate formation.

What fluid to use? Initially NS or RL, but after initial resuscitation consider switching to RL to avoid hyperchloric acidosis associated with large volume resuscitation.

Normal or high corrected sodium? Switch to 0.45% NaCL

1 bag vs 2 bag? Having 2 bags of half NS (one with D10W) both adjusted to maintain maintenance of 250cc/hr and keep euglycemic has been shown to have better outcomes: less hypoglycemia, faster closure of anion gap and less IV insulin required.

Replacement of potassium – patients in DKA have large total body potassium deficits, however due to volume contraction and acidosis the potassium is often read as normal or high.

Starting the insulin infusion will also shift potassium intracellularly. Therefore potassium should be replaced before starting insulin therapy. See the table below for guidance3.

Closure of the anion gap to stop ketone production – the issue with DKA is not necessarily the hyperglycemia, it is the ketoacidosis from low circulating insulin. After fluid resuscitation and potassium replacement, the goal is to treat the excess of serum ketones by providing insulin. This corrects the metabolic acidosis.

Avoiding hypoglycemia secondary to insulin as you correct the acidosis is pertinent. Goal is 12-14mmol/L. Once glucose drops before 14 add D5 infusion to avoid hypoglycemia as you continue the insulin infusion.

Do not stop the insulin infusion if glucose drops! It is needed to correct the ketoacidosis. If it is stopped ketone production will quickly increase again.

Gluc really low? Decrease the insulin infusion by 50%, give an amp of D50 and switch to D10.

Treat underlying precipitant.


It’s been a couple of hours. The medicine team is busy with unwell patients on the floor and you are still managing the DKA patient. You have been reassessing gases and the anion gap is not closing.

What could be going on3?
– Inadequate fluid resuscitation
– Inadequate insulin dose
– Malfunction of insulin infusion
– Underlying diagnosis contributing to anion gap hasn’t been addressed.

You reevaluate fluid status and the patient has not made any additional urine other than the small amount attained on straight cath.
You decide to repeat a 500cc bolus to address dehydration as well as increase the insulin infusion.

Could this patient be at risk of cerebral edema3?
Certainly, over-resuscitating too quickly can put patients at risk of cerebral edema. However, our patient has clinical and laboratory signs that they are still fluid deplete.

When replacing fluids consider isotonic fluids ie. D5 RL to decrease the risk.

Avoid lowering serum osmolality too quickly (ie. No more than 3mmol/kg/hr) or decreasing sodium by > 10mmol/L in 24 hours.

The sodium will often increase initially due to glucose moving intracellularly – this is not actually a measure of serum sodium – do not treat.

Admissions are backed up in the ED and you’re still caring for the patient at the end of your shift. You handover to the senior resident working with the incoming staff.

What are your goals for resolution? 3

Glucose < 11.1 AND 2 of:
– Normalization of anion gap
– Venous pH > 7.3
– Serum bicarbonate ≥15 mEq/L

At this point the patient should be mentally alert and able to eat. At this point, switch to their subcutaneous insulin dose at home. Ensure their basal insulin is also administered.

There should be an overlap of 2-4 hours before stopping the insulin infusion – if insulin infusion is abruptly stopped before administering subcutaneous insulin the patient can quickly return to an acidotic state.

What if this is the first presentation of DM and they are not on any treatment at home5?

“In patients with new-onset type 1 diabetes who have presented with DKA, an initial total daily dose (TDD) of 0.5 to 0.8 units/kg units of insulin per day is reasonable, until an optimal dose is established.

Approximately 40 to 50 percent of the TDD should be given as a basal insulin, either as once- or twice-daily U-100 glargine or detemir, or as twice-daily intermediate-acting insulin (NPH).

The long-acting insulin can be given either at bedtime or in the morning; the NPH is usually given as approximately two-thirds of the dose in the morning and one-third at bedtime. The remainder of the TDD is given as short-acting or rapid-acting insulin, divided before meals.”

The resident astutely asks about respiratory status, and if they were to decompensate what would be suggested management3?
Bottom line – avoid intubation DKA patients if possible

  • These patients hyperventilate to try and correct the acidosis, so the ventilator must also match this large volume and RR. This puts them at risk of ventilator injury and ARDS

  • Because they need to compensate with hyperventilation if there is a prolonged period of apnea from complicated intubation the acidosis can significantly worsen, putting them at high risk for circulatory collapse

But if you have to intubate, some pointers:

  • Like any patient, resuscitate first
  • If you paralyze – bag the patient throughout.
  • Consider anti-emetic
  • If the serum bicarb is < 10, considering giving an amp of bicarb
  • Once tubed the vent settings should have a high tidal volume (8cc/kg) and high respiratory rate (24-28)

How about alternative therapies if the patient is tiring, like Bipap?

DKA patients often have gastroparesis so are high risk of aspiration and emesis. Ideally, BiPap should be avoided.
If there are oxygenation issues consider high-flow nasal cannula.

The patient has resolution of their DKA within the ED and is finally admitted for treatment of the underlying cause – community acquired pneumonia.


References and further reading

  3. Helman, A. Baimel, M. Sommer, L. Tillmann, B. Episode 146 – DKA Recognition and ED Management. Emergency Medicine Cases. September, 2020. Accessed [July 16, 2021
  4. Helman, A. Himmel W. Best Case Ever 58 Euglycemic DKA. Emergency Medicine Cases. June 2017. Assessed July 19, 2021.
  5. Hirsch I, Emmett M. 2020. Diabetic ketoacidosis and hyperosmolar hyperglycemic state in adults: Treatment.
Print Friendly, PDF & Email