Diabetic ketoacidosis: no clear role for bicarbonate.

Clinical bottom line (level 1b-)

  1. Patients with diabetic ketoacidosis who were given sodium bicarbonate, had no clear fall in ketone bodies than those given no intervention.
Hale et al: British Journal of Medicine 1984; 289: 1035-1038
Expires October 2003

The study

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

38 patients (aged range 15 to 80 years; mean 44, 56% female) diabetic ketoacidosis (defined as capillary pH <7.2, bicarbonate <15 mmol/l, ketonuria> ++)

Excluded if
  • fluid was changed to 5% dextrose before the end of the study (five patients)
  • unsatisfactory samples (one patient)


    • Control Group: (n = 19, 16 analysed): no intervention
    Experimental Group: (n = 19, 16 analysed): sodium bicarbonate 1 L 150 mM, given over the middle hour instead of saline
    All patients received 20 U insulin im stat then 1 L saline in one hour iv, then 6 units insulin im after one hour, plus 1 L saline over one hour, repeated again one hour later.
    86% followed for 2 hours

    The evidence

    Outcome Control Group
    (SD)    		 Experimental Group
    (SD)    		 Mean Difference
    (95% CI)
    fall in ketone bodies (mmol/l/hour) 1.4
    (1.6)     		1.6
    (2.0)     		-0.2
    (-1.5 to 1.1)

  • Rates of change of lactate, alanine, pyruvate and glycerol did not differ between groups.

    • Comments

    1. Looked only at biochemical end points; clinical relevance not established.
    2. Process of randomisation not described.

    Citation

    1. Hale PJ, Crase J, Nattrass M: Metabolic effects of bicarbonate in the treatment of diabetic ketoacidosis. British Journal of Medicine 1984; 289: 1035-1038
    Contributor: Richard Hardern and Chris Ball, July 2000
    Reviewer:

    Clinical Question.
    Patient severe DKA
    Intervention or Exposure bicarbonate
    Comparison no intervention
    Outcome hypokalaemia or hypoglycaemia