A previously well 2-year-old child presented to an Accident and Emergency department at 3.30 in the morning with disordered consciousness, hallucinations, screaming episodes and abdominal pain. She was seen by the paediatric team and thought to have an intussusception and thus was referred to a regional paediatric surgical unit at another hospital. During this initial assessment no measurement of blood glucose level was made. Four hours later she was reviewed by the surgical team and an intussusception was excluded on clinical and radiological grounds. She was then referred on to the paediatric medical team within the same hospital. At this time she was still hallucinating but otherwise her examination was unremarkable. The child's mother was questioned about any medications that the child might have been exposed to, but mentioned only the possibility of rat poison. A urine specimen was sent for toxicology. Only during this third assessment was a blood sample drawn for measurement of blood glucose by a reagent strip with a glucometer (BM Stix) and the child noted to be profoundly hypoglycaemic with a BM Stix reading of 0.8 mmol/l (14 mg/dl). The result was not con®rmed by laboratory measurement. She was treated with a bolus of 10% dextrose given intravenously and two subsequent BM Stix readings showed values of 6.1 mmol/l (100 mg/dl) and 6.4 mmol/l (115 mg/dl). Her full blood count, routine chemistry, and capillary blood gas values were normal. Her urine did not show the presence of any ketone bodies. She was subjected to an EEG, (which did not suggest any epileptiform activity) and a CT scan of her brain, which was normal.Sixteen hours after her admission she became unresponsive, began convulsing and was unable to maintain her airway. She required intubation and ventilation and transfer to a paediatric intensive care unit was arranged. Her blood glucose value on the blood gas printout was 1.4 mmol/l (25 mg/dl). She was treated with a further bolus of 50% dextrose. After transfer and initial stabilisation, her glucose requirement was calculated to be 14.5 mg/kg per min in order to maintain a blood glucose concentration >2.6 mmol/l (normal glucose requirements for her age 4±6 mg/kg per min). Despite continuous intravenous infusion of glucose she had two further hypoglycaemic episodes and appropriate endocrine and metabolic blood samples were taken at the time of the hypoglycaemia (Table1). These further two episodes of hypoglycaemia may have been due to the miscalculation of glucose requirements as there was confusion about the percentage of dextrose being infused and the glucose requirements in mg/kg per min. This resulted in the child receiving less glucose than required to maintain normoglycaemia. Urine and blood specimens were sent for routine toxicology. She was extubated after 24 h with no complications and subsequently transferred to another tertiary centre for further assessment of the cause of the hypoglycaemia.At this centre, speci®c questioning revealed a family history of non-insulin dependent diabetes...