Case 25 A diabetic girl with a headache
Katie, a 6-year-old girl with a 3-year history of Type 1 diabetes is admitted one afternoon with an 8-hour history of vomiting. Baseline observations are temperature 36.8°C, pulse 106 /min, blood pressure 100/60 mmHg and central capillary refill time (CRT) 4 secs. Katie is seen immediately by the ST1 and ST4 trainees. Her initial venous blood gas shows a pH 7.25, (normal 7.35–7.45) pCO2 3.4kPa (4.7–6.4 kPa), pO2 5.4kPa (venous sample), bicarbonate 11mmol/L (22-29 mmol/L), base excess – 10 mmol/L (+2.5 to -2.5 mmol/L) and lactate 3 mmol/L (<2.0 mmol/L). Her bedside finger prick blood glucose is 25 mmol/L (3.5–6.0 mmol/L). The ST1 and ST4 trainees diagnose diabetic ketoacidosis (DKA) and a 20 ml/kg bolus of 0.9% saline is administered intravenously. Katie is estimated to be 10% dehydrated and, in addition to the bolus, is written up for a fluid regime to replace the deficit plus maintenance over the next 48 hours. The National Paediatric DKA protocol is consulted and an insulin infusion is started an hour later at 0.1 units/kg/hour. Her U and Es results are sodium 127 mmol/L (normal 135–145 mmol/L), potassium 4.7 mmol/L (3.5–5.6 mmol/L), urea 4.2 mmo/L (2.5–6.6 mmol/L), creatinine 56 μmol/L (20–80 μmol/L) and laboratory glucose 23.5 mmol/L.
What do you think of the management so far?
When reviewed 4 hours later, Katie is alert and orientated although tired. Her pulse is 90/min, her BP 90/60 mmHg and her CRT 2 secs. However her fluid balance is negative at –350 mls and a repeat venous gas is similar. The bedside blood glucose is 12 mmol/L. The protocol states that a further bolus of 0.9% saline may be needed if there has been no clinical improvement. Senior review is also recommended. The ST4 decides to give her a further bolus of 20 mls/kg 0.9% saline to correct the dehydration.
Would you have done this?
Further electrolytes are sent and the sodium is phoned back as 128 mmol/L and the blood glucose is 11.5 mmol/L.
Would these results alter your management?
The medical team is phoned during handover 4 hours later to say that Katie is complaining of a bad headache and has begun wetting the bed. The nurse accepts a verbal prescription for paracetamol. Half an hour later the night team call the ward and are reassured that Katie is now sleeping and comfortable. They plan to review her later on their night round.
Would you be reassured?
Two hours later the nurses phone to say that Katie is unrousable and has a sluggish pupillary reaction to light. Her blood pressure is 140/100 mmHg and her pulse 56/minute. On examination these findings are confirmed and she is hyperreflexic. A clinical diagnosis of cerebral oedema is made. Hypertonic saline (2.7%) 5 mls/kg is administered iv and the duty consultant and PICU team are summoned. She is ventilated on PICU for 10 days and is left with profound motor and learning disabilities. The parents obtain expert advice and file a lawsuit complaining that the DKA protocol was not followed and that Katie received excess fluids at the outset and during her rehydration.
Expert opinion
Guidelines and protocols are written for a purpose – those for DKA are nationally agreed and have been honed from years of experience. Children and young adults with DKA do not die from shock but usually from cerebral oedema which has an incidence of 0.3–1% in DKA. It is unpredictable, more frequent in younger children and newly diagnosed diabetes and has a mortality of approximately 25%. Significant neurological morbidity occurs in 10–26% of survivors. Case-controlled studies have shown that cerebral oedema is probably linked to the overestimation of dehydration and shock, rapid administration of excess crystalloid fluids, too rapid a reduction in blood glucose and, possibly, early use of insulin. The protocol aims to minimize the risks by a slow correction of the metabolic abnormalities.
Katie had mild DKA with no evidence of shock at the outset – her pulse and BP were normal and CRT is a poor sign in DKA as hypocarbia leads to peripheral vasoconstriction. The initial fluid bolus was unnecessary. If used in DKA, boluses should be given in 10 ml/kg aliquots and resuscitation fluids should be subtracted from the 48 hr fluid total. Katie’s dehydration was also overestimated – most patients with DKA are only 3–5% dehydrated and no more than 8% should be used in calculations.
Once any shock has been treated it is never urgent to restore physiological normality. At 4 hours Katie was clinically improving and a negative fluid balance is not unusual at this stage. It is crucial to treat the patient and not the figures and Katie did not need the second bolus. The recommended senior review was not sought.
Regular neurological observations are essential in monitoring DKA. Headache and other changes in neurological status such as incontinence should trigger an urgent clinical review because they suggest cerebral oedema and early recognition and management improves outcome. Hypertonic saline or mannitol should be given immediately, maintenance fluids halved and the deficit replaced over 72 hours with close monitoring in a PICU.
Initial dilutional hyponatraemia is common in DKA as the high serum osmolarity drives water from the intracellular to the extracellular space. There is also hypernatruria during the osmotic diuresis. The glucose-corrected serum sodium (Serum Na + 0.4 × (Glucose − 5.5)) may prove a useful early warning sign of the development of cerebral oedema in DKA. It should rise by about 5 mmol/L over the first 8 hours of therapy. Typically it falls in children who develop late cerebral oedema. This child’s initial corrected sodium of 134.2 mmol/L fell to 130.4 mmol/L at 4 hrs.
The final common denominator for all these factors is understanding the risk and minimizing it by paying close attention to detail, making frequent reviews and following the protocol. Any deviation from it should be very carefully considered and discussed with a consultant experienced in DKA. It should also be clearly documented in the notes.
Legal comment
This is a high-value claim. Katie will require care and support all her life. It will be difficult to defend the actions of the hospital staff. Nonetheless, the lawyers for the NHSLA will want to investigate how the mistake occurred just in case something emerges which might raise an arguable defence.
The crucial question is whether the ST4’s decision to administer the second bolus can be justified. If he thought the circumstances justified a departure from the protocol, what were they? Such deviations and the reasons for them should be recorded in the notes at the time. Why did he not get a senior to review the case as the protocol recommended?
Even if the ST4 can give a factual account of events that justifies his decision, the standard of care in the next phase of treatment is highly questionable. It seems that an instructed expert will conclude that the wrong treatment was given (breach of duty) and that, combined with a failure to regularly review, has caused the damage (causation). The claim will probably be settled out of court. The value of the claim will depend on a number of factors, not least of which are Katie’s life expectancy and the level of care that she will require. It could easily be several million pounds.
