Diabetic ketoacidosis

Rehydration and insulin management

Fluids

The cause of cerebral edema during treatment remains unclear. However, too rapid reduction in intravascular osmolality may aggravate the process. It seems prudent therefore that rehydration should occur more slowly in children with DKA than in other causes of dehydration

Proceed with urgency but with caution

Use either: Fluid calculation (model 1)

Requirements = Deficit + Maintenance

  • Calculate DEFICIT = estimated % dehydration x body weight (kg and equivalent in ml)
  • Calculate MAINTENANCE (ml) (Table 7)

  • Then add DEFICIT to 48 h MAINTENANCE and replace this volume evenly over 48 h as Normal saline 0.9% initially

Or use: Fluid calculation (model 2)

Covers MAINTENANCE + 10% DEFICIT given evenly over 48 h in children of all sizes

  • 6 ml/kg per h for children weighing 3–9 kg
  • 5 ml/kg per h for children weighing 10–19 kg
  • 4 ml/kg per h for children weighing >20 kg (up to maximum of 250 ml/h)
These calculations will usually cover ongoing losses which in most cases do not need additional replacement, but excessive continuing fluid losses might need replacing if the severity of dehydration is not improving

  • When the BG falls to 12–15 mmol/l the infusion should be changed to a fluid containing glucose, the most commonly recommended being saline 0.45% (or 0.9%) with glucose 4–5% (or glucose 5% with added sodium chloride 80 mmol/l or more)

Oral fluids

  • In severe dehydration and acidosis only allow sips of cold water or ice to suck
  • Oral fluids (e.g. fruit juice/oral rehydration solution) should only be offered after substantial clinical improvement and no vomiting
  • Oral fluid volume should be subtracted from the IV calculations

Potassium

  • Total body potassium is always substantially depleted in DKA
  • Serum/plasma potassium may be low, normal or high
  • If serum potassium is not available before the completion of resuscitation, ECG monitoring is recommended before potassium is added to the infusion fluid
  • Start potassium replacement as soon as resuscitation is completed and the ECG does not show elevated T-waves (or if serum potassium is not elevated)
  • Potassium chloride 40 mmol is usually added to each liter of saline infusion
    [Phosphate or acetate salts of potassium may also be used but there is no evidence that they are preferable]

Insulin

DKA is caused by insulin deficiency, either relative or absolute

  • Insulin should not be started until shock has been successfully reversed by emergency resuscitation and a saline/potassium rehydration regimen begun (this avoids sudden influx of potassium from plasma into cells, with danger of cardiac arrhythmia)
  • During the first 60–90 min of initial rehydration the BG may fall substantially even without insulin treatment
  • Insulin by continuous low-dose IV infusion is the optimal method
    [An initial bolus of insulin is not recommended]
  • A solution of soluble insulin 1 unit/ml made up in Normal saline is best by electronic pump
  • Recommended initial insulin dose = 0.1 units/kg per h
    [Some recommend 0.05 units/kg per h particularly for younger patients]
  • When syringe pumps are not available a separate low-dose infusion may be given, e.g. soluble insulin 50 units in 500 ml Normal saline (i.e. 1 unit insulin per 10 ml saline), the bag being changed every 24 h to avoid inactivation of insulin
  • When insulin infusion methods are not available the use of hourly IM/SC injections of soluble or rapid-acting insulin 0.1 units/kg has been shown to be effective
  • After resuscitation the typical rate of fall of BG is 4–5 mmol/h
  • When BG falls to 12–15 mmol/l, change to the glucose saline infusion (as above) to maintain BG in the desired range of 8–12 mmol/l
  • If BG rises again above 15 mmol/l, increase the insulin infusion by 25%
  • If BG falls to <8 mmol/l, or falls too rapidly, increase the concentration of glucose to 10% (or more) with added saline
  • The insulin infusion rate should only be decreased if the BG level remains below the target range despite glucose supplementation
  • Do not stop insulin infusion or decrease below 0.05 units/kg per h
  • because a continuous supply of both insulin and glucose substrate is needed to promote anabolism and reduce ketosis

Bicarbonate

There is no evidence that bicarbonate is either necessary or safe in DKA. Bicarbonate should not be used in the initial resuscitation

  • Potential hazards of bicarbonate therapy
    • exacerbation of CNS acidosis
    • hypokalemia and altered calcium ionization
    • excessive osmolar load
    • tissue hypoxia
  • Persistent acidosis is likely to be caused by inadequate resuscitation, inadequate insulin effect or sepsis
  • Bicarbonate may be considered for treatment of impaired cardiac contractility in persistent severe shock
    [If bicarbonate is considered, proceed with caution giving 1–2 mmol/kg bicarbonate over 60 min]
 
Monitoring progress

 
 
Consensus Guidelines 2000
Click here for 'Index'
  
  
  
  
  
  
  
  
  
IDF Type 1 Guidelines
IDF Type 2 Guidelines