Abstract
BACKGROUND: Hyperchloremic metabolic acidosis that develops during the treatment of diabetic ketoacidosis is usually attributed to the chloride content of resuscitation fluids. We explored an alternative explanation, namely that fluid-induced plasma volume expansion alters the absolute differences in the concentrations of sodium and chloride (the Na-Cl gap) enough to affect the acid-base balance. We analyzed data from a prospective single-center cohort study of 14 patients treated for diabetic ketoacidosis. All patients received 1 L of 0.9% saline over 30 min on two consecutive days. Blood gases were sampled before and after the infusions. RESULTS: The initial plasma volume was estimated to be 25 ± 13% (mean ± SD) below normal on admission to the intensive care unit. At that time, most patients had an increased actual Na-Cl gap, which counteracts acidosis. However, the correction of the plasma volume deficit revealed that these patients would have had a decreased Na-Cl gap upon admission if they had been normovolemic at that time; the estimated "virtual Na-Cl gap" of 29 ± 5 mmol/L was significantly lower than the uncorrected value, which was 39 ± 5 mmol/L (P < 0.001). On Day 2, most patients had a decreased actual Na-Cl gap (33 ± 5 mmol/L), approaching the corrected value on Day 1. CONCLUSIONS: The hyperchloremic acidosis commonly seen in diabetic ketoacidosis may not be primarily caused by the chloride content of resuscitation fluids but, rather, by the restoration of plasma volume, which reveals the hidden metabolic acidosis caused by a decreased Na-Cl gap. Trial registration Clinical Trials Identifier NCT02172092, registered June 24, 2014, https://www. CLINICALTRIALS: gov/NCT02172092.