[ ] Labs: BMP Ca Mag Phos, VBG, lactate, lipase, UA Uket Ucg
+/- trop, tox, osms, serum ketones (beta-hydroxybutyrate BHOB or Acetoacetate AcAc)
+/- EKG CXR RUQ US blood cxs etc.
[ ] IVF to start, +/- K+
[ ] FS q1-2 hour
[ ] lytes q2-4
[ ] insulin only after you've seen the potassium is > 3.3!
Definition of DKA
Glucose > 200 mg/dL
pH < 7.3
AG > 10-12* or HCO3 < 18
* Na+ - [Cl + HCO3]
(Metabolic acidosis is non-specific. Kussmaul breathing of respiratory compensation, alongside stomach acid lost in the nausea & vomiting of DKA, can cause a mixed acid-base picture. With serum BHOB levels now widely available, it has been argued a serum BHOB > 3 mmol/L is a better diagnostic criteria.)
DKA vs HHS?
Hyperosmolar Hyperglycemic State
glc > 600
osms > 320
pH > 7.3
Altered MS common
Overall diuresis and mortality rates are much higher in HHS.
Treat as DKA — but with about half of the rates of fluid and insulin.
Many hospitals have exacting protocols, know and follow yours.
IV fluid is the front-line treatment.
Give isotonic (LR or NS) 1L per hour over first 2 hours.
(Or 20 cc/kg for first hour).
Then 0.5-1L per hour next 3 hours.
Once patient voids, add K+ to IVF
Not an exact science, just add 20-30 mEQ. Or:
|< 3.5||40 mEQ/L|
|> 5.5||hold K+|
If hyperchloremic acidosis: Can give KPhos (esp if phos < 2.5). 30 mmol in 500 cc.
Insulin is not emergent here. Could be started upstairs even.
Historically given regular insulin bolus followed by 0.1 U/kg/h gtt. Recent studies have skipped the bolus but used a slightly higher gtt rate of regular insulin 0.14 U/kg/h. Roughly 10 U/hr in 70 kg patient.
Blood sugar should fall 50-70 mg/dL/h.
If not, double rate and reassess. (Sepsis? MI? Dead gut? Pancreatitis?)
Tinkering / Boarding DKA
When glc dropping fast and/or gap closing
- Halve the gtt
- Add D5 to IVF (Dropping glc a lot may be a/w cerebral edema. The goal of the insulin is not to plummet sugar to normal, but to stop fatty acid metabolism -> ketones.)
- If glc still dropping very fast, switch to D10
When glc <250 and gap closed/closing
- Change to D5 1/2NS @ 250/h
- Feed patient
- NPH or Regular Insulin 0.2 U/kg SQ
0.6 U/kg/day divided by 2-3 doses per day
- Keep gtt on 1-2 more hours, as low as 0.02-0.05 U/kg/h, then off
- When all criteria normal, DC everything and switch to sliding scale insulin
SubQ Insulin instead IV?
Two RCTs have shown giving fast-acting insulins SQ — one with Aspart (Novolog), another with Lispro (Humalog) — every 1 or 2 hours is safe and effective
But severely ill DKA should get a gtt and ICU.
Dosing regimens for SQ Aspart (Novolog) or Lispro (Humalog)
- Give bolus 0.3 U/kg SQ
- Follow with either 0.1 U/kg SQ every 1 hour, or 0.2 U/kg SQ every 2 hours
- Check fingerstick glucose either q1H if dosing q1H, or q2H if dosing q2H.
- When glc levels fall below 250 mg/dL: Reduce to 0.1 U/kg q2H until DKA resolves
Bicarb gtts in severe acidosis?
Bicarbonate never been shown to be useful, but in severe acidosis (pH < 6.9) gtt is recommended.
Put 100 mmols (2 amps) of bicarb in 400 mL of sterile water with 20 mEQ KCl, run over 2 hours. It's isotonic. Repeat until pH greater than 7.
What About Phosphate?
No benefit shown in phosphate repletion; aggressive repletion has caused hypocalcemia.
But replete for phosphate < 1 mg/dL, or between 1 and 2 mg/dL if any cardiopulmonary issues
Give 0.1-0.2 mmol/kg over 6 hours
10 mL of KPhos IV contains 44 mmol of K and 30 mmol of phos
Obtunded DKA Exam
[ ] EKG, CXR, abdominal imaging
[ ] palpate sinuses (rhinocerebral mucomycosis)
[ ] ears, throat
[ ] supple neck
[ ] skin head -> toe (cellulitis, Fournier's)
[ ] rectal (abscess, GI bleed, prostatitis)
[ ] bones & joints
Death from DKA?
1) "The underlying cause," i.e. from their sepsis or MI
2) Potassium derangements
- acidosis-driven hyperK+ early (even if overall potassium depleted);
- hypokalemia late (after fluids & insulin with forgotten K+ supplementation)
Underdeveloped world: Dehydration, cerebral edema
Normally BOHB and AcAc are present in serum in equimolar amounts.
In DKA, however, the average ratio is 3:1 (but varies all over, from 1:3 to 7:1).
None of the urine dips (which use a nitroprusside rxn and turn purple in presence of AcAc) detect BOHB.
So it is possible to be "urine ketone negative" and still have DKA.
According to the American Diabetes Association, 10% of the DKA population presents with glucose levels below 250 mg/dL. The ADA theorizes that this is sometimes from the patient taking insulin just prior to arriving at hospital, or possibly from antecedent food restriction, or possibly from something inhibiting gluconeogenesis.