ESPE Abstracts (2019) 92 P3-54

Relationship Between Chloride Infusion and Base Excess in Initial Treatment of Pediatric Diabetic Ketoacidosis

Kentaro Sawano1,2, Fusa Nagamatsu1, Kazuhiro Shimura1, Yuki Abe3, Yukie Izumita3, Yohei Ogawa2, Nagisa Komatsu4, Shigeru Takishima5, Akie Nakamura6, Hiromi Nyuzuki2, Takeshi Yamaguchi6, Koji Muroya7, Sayaka Watanabe-Yamamoto6, Keisuke Nagasaki2, Akihiko Saitoh2, Yukihiro Hasegawa1


1Department of Endocrinology and Metabolism, Tokyo Metropolitan Children's Medical Center, Tokyo, Japan. 2Division of Pediatrics, Department of Homeostatic Regulation and Development, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan. 3Department of Pediatrics, Niigata City General Hospital, Niigata, Japan. 4Department of Pediatrics, Japanese Red Cross Hospital Kumamoto, Kumamoto, Japan. 5Department of Pediatrics, Soka Municipal Hospital, Soka, Japan. 6Department of Pediatrics, Hokkaido University Graduate School of Medicine, Sapporo, Japan. 7Department of Endocrinology and Metabolism, Kanagawa Children's Medical Center, Yokohama, Japan


Background: In initial treatment of DKA, volume expansion should begin with 0.9% saline to restore the peripheral circulation. The use of large amounts of chloride-rich/bicarbonate-free fluids may cause the rapid development of hyperchloremic metabolic acidosis, which is described in ISPAD Clinical Practice Consensus Guidelines 2018.

The severity of DKA, defined by pH, HCO3- Base Excess (BE), is one of the factors in the prognosis. Advanced acidosis is a risk factor of cerebral edema, which is known to the major event in term of the life prognosis. We aimed to investigate the relationship between acidosis and Chloride infusionin our early treatment of pediatric DKA before administration of insulin.

Methods and Results: (Subjects)

A total of 45 children with DKA admitted to seven institutions between 2010 and 2018 were retrospectively analyzed.Children in whom insulin or bicarbonate was administered before visiting our hospitals were excluded.

Methods: As an indicator of acidosis, BE was used to remove respiratory factors. We analyzed the average Chloride infusion rate (mEq/kg/h) and BE change rate (BE change/hour) from the start of infusion to before insulin administration. Anion gap and lactic acid were calculated and measured, respectively, as one of the factors of BE decrease.

We also divided the children into two groups according to average Chloride infusion rate, less than 10ml/kg/h (group S; N=21), and more (group R; N=24).

Results: The mean of BE change rate was -0.57 and its 95% confidence interval is -0.831 to -0.309, which was a negative value. There was no relationship among BE change rate and anion gap/lactic acid change rate. No significant difference in BE change rate between S group and R group were observed. The results were similar when analyzed separately for either moderate or severe cases.

Conclusions: The 95% confidence interval of the mean of BE rate of change in this study was in the negative range below zero, suggesting that the BE decreased in initial treatment of DKA with chloride-rich fluids.