Risk factors associated with resolution of diabetic ketoacidosis in pediatric critical care units

Abstract

Objective: Diabetic ketoacidosis (DKA) is the main cause of morbidity and mortality in children with type-I Diabetes Mellitus. The goals of therapy are to correct dehydration, resolution of acidosis and fading of ketosis. Such serious complications necessitate closed monitoring of DKA patients with delicate, balanced therapy, probably at an intensive care facility. Regarding the fact that, each facility shoul determine the clinical profile of their own patient population, we aimed to investigate the risk factors for consequences and determine the timing of DKA resolution by analyzing the demographic and epidemiologic data, clinical outcome and the prognosis of diabetic ketoacidotic children admitted to PICU. Method: This descriptive, retrospective study was conducted in 105 children admitted to PICU with the complaints of DKA between January 2014 and December 2108. Demograhic data including age, gender, weight, height, body mass index (BMI), initial compliants with clinical findings and level of consciousness were recorded. Children were categorized into two groups depending on the timing of DM diagnosis (new onset of diabetes and established diabetes mellitus). DKA severity was determined by the degree of metabolic acidosis (mild, moderate, severe). SPSS-23 was used for statictics. Descriptive analyses were expressed as percentages, mean±standart deviation (SD), median with minimum and maximum values. Chi square and Fischer exact test were used for comparison of categorical variables. Student’s t-test, Mann Whitney U test and Wilcoxon rank sum test were assessed for continous variables. Pearson correlation coefficient and logistic regressions were used for correlations and to determine the risk factors. P-value < 0.05 was considered significant. Results: The patient demographics presented the mean age as 11.31±4.18 years, female/male ratio 1/1.4 and body mass index 18.48±4.48. Children were classified as mild DKA (29.5%), moderate DKA (35.2%) and severe DKA (35.2%) based on the acidosis severity. 48.6% of the patients had Kusmaull respiration; 30.5% had manifested altered consciousness. One patient had tomography-proven brain edema and had required mechanical ventilation due to neurological incapability to sustain airway Children with new onset of diabetes accounted for 51.4% of the study population. The mean age was 9.70±4.47 years; this group constituted a younger population compared the established DM patients (p<0.001). Altered mental state and kusmaull respiration also occurred at a higher rate and the major complaint seemed ae weight loss within two weeks (p=0.006, p=0.002, p<0.001 respectively). Children with established diabetes mellitus presented significant biochemical abnormalities in terms of elevated BUN and serum potassium levels (p<0.001, p<0.001); infections occurred as the major triggering factor for DKA at a rate of 80.4% at this group. We observed a positive correlation with DKA resolution with serum creatinine, calculated osmolality, anion gap (r=0.242, r=0.215, r=0.302) and a negative correlation with blood gas pH and HCO3 (r= -0.704, r= -0.694). In the multivariable regression model including age, gender, body mass index, PRISM-3 score, BUN, serum potassium, phosphate and chloride, only blood gas pH and new onset of diabetes appeared to be the independent risk factors for DKA resolution. 0.1 unit decrement in blood gas pH elongated the resolution by 3.76 hours (p<0.001, adjusted ratio: 0.743). New onset of diabetes mellitus also increased the length of resolution by 5.30 hours (p<0.001). Conclusions: Inıtial blood gas pH and presence of new onset of diabetes are the major risk factors in resolution of ketoacidosis.