Risk factors for cerebral edema in diabetic ketoacidosis in a developing country

Abstract

To study the clinical profile and risk factors of cerebral edema in children with diabetic ketoacidosis with specific reference to fluid refractory shock. Retrospective review of medical records. Twelve-bed pediatric intensive care unit of a teaching hospital. Seventy-seven patients admitted to pediatric intensive care unit with a diagnosis of diabetic ketoacidosis over 5 yrs. Medical records were reviewed, and data with respect to patients' age, clinical features, biochemical profile (blood glucose, osmolality, electrolytes, urea, creatinine, arterial pH, PaCO(2), and HCO(3) at admission, 6-12 hrs, 24 hrs, and beyond 24 hrs), cerebral edema, presence of sepsis and shock, treatment details, and primary outcome in terms of survival or death were retrieved. Patients with and without cerebral edema were compared. Variables that were significant on univariate analysis were entered in a multiple logistic regression analysis to determine the predictors for cerebral edema. Odds ratio and 95% confidence interval were calculated using SPSS version 15. Mean age of the patients was 5.6 (standard deviation, 3.8) years. Fifty-five (71.4%) patients had new-onset diabetes mellitus. Cerebral edema was seen in 20 patients (26%). Blood glucose, serum osmolality, and CO(2) values at admission and rate of decline in glucose and osmolality during the first 12 hrs were similar in the cerebral edema and noncerebral edema groups. On multiple logistic regression analysis, fluid refractory shock (odds ratio, 7.3; 95% confidence interval, 1.3-41; p = .025) and presence of azotemia (odds ratio, 4.3; 95% confidence interval, 1.1-16; p = .034) at admission were predictors for development of cerebral edema. Mortality in cerebral edema group was 25% as compared to 3% in the noncerebral edema group. Patients with fluid refractory shock and azotemia at admission had higher odds for development of cerebral edema. Initial blood glucose, effective osmolality, or decline in glucose and osmolality had no association with cerebral edema.