The Influence of Chloride for the Interpretation of Plasma Bicarbonate During the Treatment of Diabetic Ketoacidosis.

Abstract

Hyperchloremic metabolic acidosis can occur in diabetic ketoacidosis (DKA) and may affect the acid-base interpretation during treatment. This study aims to describe the prevalence of hyperchloremia during the treatment of DKA and its effect on the interpretation of bicarbonate value. A cross-sectional study, including all cases of DKA in patients aged 1 to 18 years old admitted from 2010 to 2015, was performed. Laboratory tests were performed on admission (baseline), 2 and 6 hours after admission, and when resolution of DKA was achieved. Adjusted bicarbonate value was calculated using regression equations. Seventy-nine DKA episodes were included. The average age was 13.3 ± 3.8 years. Baseline levels were as follows: plasma glucose, 479 ± 133 mg/dL; pH 7.1 ± 0.083; bicarbonate, 9.65 ± 2.9; and anion gap, 23.9 ± 7.5. The time to achieve resolution of DKA was 12.2 ± 4.4 hours, and the decrease in capillary glucose was 25.5 (19.7-38.2) mg/dL per hour. After 6 hours of treatment, the proportion of patients presenting hyperchloremia increased from 23% to 77%. By using adjusted bicarbonate, the percentage of patients achieving resolution of DKA after 6 hours of treatment would have been 35.4% (confidence interval 95%, 28-49), in comparison with 24.1% (confidence interval 95%, 18-37) using observed bicarbonate (P = 0.004). The hyperchloremia developed during the treatment of DKA could modify the value of measured plasma bicarbonate concentration and unnecessarily prolong the initial phase of treatment.