Diabetic ketoacidosis: hypertonic dextrose did not speed biochemical resolution.

Clinical bottom line (level 1b-)

  1. Patients who were recovering from diabetic ketoacidosis and had a glucose <14 mmol/l did not clearly have faster improvement in biochemical markers following 10% glucose and insulin rather than 5% glucose and insulin.
Krentz et al: Diabetic Medicine 1989; 6: 31-36
Expires October 2003

The study

Unblinded concealed randomised trial without intention-to-treat
Setting: teaching hospital, UK

17 patients (aged range 16 to 36 years, ?% male) severe diabetic ketoacidosis (defined as a clinical diagnosis of uncontrolled diabetes requiring emergency hospital admission, capillary pH <7.2, venous bicarbonate <15 mmol/l and ketonuria > 2+)

Excluded if
  • initial total ketone body concentration <5.0 mmol/l (14 patients)


    • Control Group: (n = 9, 9 analysed): 5% dextrose with 10 units/l insulin infused at 250 ml/h
    Experimental Group: (n = 8, 8 analysed): 10% dextrose with 40 units/l insulin infused at 250 ml/h
    All patients received insulin 20 U bolus im then 6 U hourly im with iv 0.9% saline, until glucose <14 mmol/l, then they were randomised. Potassium was given in both groups in order to maintain normokalaemia.
    100% followed for 6 hours

    The evidence

    Outcome Control Group
    (SD)    		 Experimental Group
    (SD)    		 Mean Difference
    (95% CI)
    pH 7.31
    (0.03)     		7.32
    (0.06)     		-0.010
    (-0.058 to 0.038)
    capillary bicarbonate (mmol/l) 16.2
    (2.7)     		16.0
    (4.5)     		0.20
    (-3.6 to 4.0)
    glucose (mmol/l) 11.5
    (2.7)     		15.7
    (3.7)     		-4.2
    (-7.5 to 0.88)
    total ketone bodies 5.18
    (3.42)     		7.34
    (1.61)     		-2.2
    (-5.0 to 0.67)

    Comments

    1. The small numbers and lack of clinical endpoints make these results less certain.
    2. Not clear why patients with total ketone bodies <5.0 mM were excluded from analysis nor is it clear if this was a decision made a priori.
    3. No episodes of hypokalaemia were observed through the potassium dropped in both groups during the dextrose infusion.

    Citation

    1. Krentz AJ, Hale PJ, Singh BM: The effect of glucose and insulin infusion on the fall of ketone bodies during treatment of diabetic ketoacidosis. Diabetic Medicine 1989; 6: 31-36
    Contributor: Richard Hardern and Chris Ball, July 2000
    Reviewer:

    Clinical Question.
    Patient DKA
    Intervention or Exposure 10% glucose
    Comparison 5% glucose
    Outcome biochemical markers