A 14-year-old female with symptoms of nausea, vomiting, and abdominal pain for 3 days was admitted to the local clinic with the diagnosis of acute viral gastroenteritis. Laboratory results showing hyperglycemia (glucose, 500 mg/dL) and ketonuria strongly indicated DKA. The patient was transferred to our emergency unit because of worsening mental status despite initiation of therapy with intravenous fluid and shortacting insulin. The patient's medical history was unremarkable (her mother had type 2 diabetes mellitus). On admission, the patient was lethargic, severely dehydrated, and drowsy (Glasgow Coma Scale [GCS, eye-verbal-motor]=14 [
3-
5-
6]). Vital signs indicated tachycardia (124 beats/min), tachypnea (40 breaths/min), and hypothermia (35℃). Blood pressure was in the normal range (106/56 mmHg), body weight had decreased by 1.7 kg in the previous month to 48 kg (25th–50th percentile), and body mass index was 20 kg/m
2 (50th–75th percentile). Chest and abdominal x-rays showed no specific abnormalities. Initial venous blood gas analysis indicated metabolic acidosis (pH, 6.92; PCO
2, 27 mmHg; PO
2, 57 mmHg; HCO
3, 5.6 mmol/L). Serum chemistry and electrolyte tests revealed 569 mg/dL glucose, 133 mmol/L corrected sodium, 3.2 mmol/L potassium, 8.5 mg/dL corrected calcium, 2.1 mg/dL phosphate, 25.9 mg/dL blood urea nitrogen, 0.57 mg/dL creatinine, and 0.6 mmol/L lactate. Urinalysis indicated the presence of ketones and glucose. Endocrinologic assessments showed 0.26 ng/mL C-peptide (normal range, 0.6–2.3 ng/mL) and 15.7% glycosylated hemoglobin. Antiglutamic acid decarboxylase was positive (1.72 U/mL; range, 0–1 U/mL), whereas anti-islet cell and anti-insulin antibodies were negative. These findings indicated a diagnosis of type 1 diabetes and severe DKA. The main metabolic data during hospitalization are shown in
Table 1. Fluid (0.45% saline) and continuous intravenous insulin infusion therapy were initiated at a rate of 0.05 unit/kg/hr immediately according to DKA treatment protocol [
5]. Hourly neurological evaluations were performed for signs of cerebral edema. Although the patient remained drowsy, signs of cerebral edema were not evident. The patient suffered a <1 min generalized tonic seizure at 4 hours postadmission that was followed by bradycardia and cardiopulmonary arrest. External cardiac massage was initiated, followed by one epinephrine injection. Spontaneous circulation resumed after 6 minutes of cardiopulmonary resuscitation. Laboratory data after return of spontaneous circulation showed metabolic acidosis and hypophosphatemia (pH, 7.05; PCO
2, 23.6 mmHg; PO
2, 181 mmHg; HCO
3, 6.7 mmol/L; glucose, 517 mg/dL; corrected sodium, 136 mmol/L; potassium, 3.2 mmol/L; corrected calcium, 8.8 mg/dL; phosphate, 1.2 mg/dL). The patient received K
2HPO
4 and KCl to achieve 1.9 mEq/kg/day of phosphorus after the cardiac arrest. A brain computed tomographic (CT) scan showed no brain edema or intracranial pathology. Dexamethasone and mannitol were prescribed for prevention of brain edema and the patient was admitted to the intensive care unit. Although the patient remained drowsy (GCS=14 [
3-
5-
6]), vital signs were stable. At 16 hours after initiation of DKA treatment, the patient became stuporous (GCS=9 [
2-
2-
5]) with shallow, weak, and slow respiration that necessitated endotracheal intubation. Blood tests at this time revealed respiratory acidosis and severe hypophosphatemia (pH, 6.80; PCO
2, 92.9 mmHg; PO
2, 160.3 mmHg; HCO
3, 14.6 mmol/L; glucose, 375 mg/dL; corrected sodium, 142 mmol/L; potassium, 2.8 mmol/L; phosphate, 0.4 mg/dL; lactate, 0.6 mmol/L). Phosphate replacement was increased to achieve a 3.8 mEq/kg/day phosphorus rate. A repeat brain CT imaging study showed no specific findings. The patient's electrolyte and pH values began normalizing within 2 days but mechanical ventilation was still required because hypophosphatemia was not alleviated (phosphate, 1.4 mg/dL). Phosphate replacement was increased to achieve a 5.7 mEq/kg/day phosphorus rate. Fluid replacement with potassium and phosphate, and continuous intravenous insulin administration was maintained until laboratory findings were normal. On the third day, the patient’s vital signs were stable, self-respiration was recovered, arterial blood gas levels were normal (pH, 7.39; PCO
2, 25.2 mmHg; PO
2, 139.3 mmHg; HCO
3, 15.6 mmol/L), and plasma phosphate had risen (3.5 mg/dL). Weaning from inotropic and ventilator supports was possible. The patient was transferred to the general ward the next day. On the fourth day of hospitalization, oral feeding and multiple subcutaneous insulin injections were started and phosphate supplementation was stopped. An electroencephalogram showed no epileptiform discharges or slow waves. The patient was discharged on day 10 with no neurologic complications. The patient was followed up regularly. No neurologic complications or respiratory problems were noted, and the patient did not require further phosphate supplementation.
This study approval and informed consent were waived by the Institutional Review Board of Severance Hospital.