The effect of dehydration, hyperchloremia and volume of fluid resuscitation on acute kidney injury in children admitted to hospital with diabetic ketoacidosis.

Background: Acute kidney injury (AKI) is a recognized comorbidity in pediatric diabetic ketoacidosis (DKA), although the exact etiology is unclear. The unique physiology of DKA makes dehydration assessments challenging, and these patients potentially receive excessive amounts of intravenous fluids (IVF). We hypothesized that dehydration is over-estimated in pediatric DKA, leading to over-administration of IVF and hyperchloremia that worsens AKI.

Methods: Retrospective cohort of all DKA inpatients at a tertiary pediatric hospital from 2014 to 2019. A total of 145 children were included; reasons for exclusion were pre-existing kidney disease or incomplete medical records. AKI was determined by change in creatinine during admission, and comparison to a calculated baseline value. Linear regression multivariable analysis was used to identify factors associated with AKI. True dehydration was calculated from patients' change in weight, as previously validated. Fluid over-resuscitation was defined as total fluids given above the true dehydration.

Results: A total of 19% of patients met KDIGO serum creatinine criteria for AKI on admission. Only 2% had AKI on hospital discharge. True dehydration and high serum urea levels were associated with high serum creatinine levels on admission (p = 0.042; p < 0.001, respectively). Fluid over-resuscitation and hyperchloremia were associated with delayed kidney recovery (p < 0.001). Severity of initial AKI was associated with cerebral edema (p = 0.018).

Conclusions: Dehydration was associated with initial AKI in children with DKA. Persistent AKI and delay to recovery was associated with hyperchloremia and over-resuscitation with IVF, potentially modifiable clinical variables for earlier AKI recovery and reduction in long-term morbidity. This highlights the need to re-address fluid protocols in pediatric DKA.