Abstract
Background: To determine the effect of chloride on the acid-base status, four approaches are currently used: 1) accepted ranges of serum chloride values; 2) chloride corrections, such as chloride deficiency/excess and chloride modification; 3) the Cl/Na ratio; and 4) the sodium- chloride difference, such as base-excess chloride. However, these approaches are governed by different concepts, and they can evaluate the effects of chloride on the acid-base status differently. Our aim is to investigate which approach to the evaluation of the effect of chloride is the best.Methods: In this retrospective cohort study, 2529 critically ill patients who were admitted to the tertiary care unit were evaluated between 2011 and 2018. Patient characteristics and blood gas parameters at the ICU admission and outcomes were recorded. The effects of chloride on the acid-base status according to each evaluative approach were validated by the standard base excess and apparent strong ion difference. To compare approaches, kappa and Bland-Altman tests and a linear regression model were used. Results: In the linear regression model for all patients, only base-excess chloride in all the chloride evaluation approaches was significantly related to the standard base excess. In the subgroup, the correlation and limits of agreement between base-excess chloride and the standard base excess were the strongest (r2=0.92 p<0.001 bias: 0.5mmol/L). Conclusions: For the evaluation of the effect of chloride on the acid-base status, base-excess chloride is a better approach than accepted ranges of serum chloride values, chloride corrections and the Cl/Na ratio.
Partitioning evolutive standard base excess determinants in septic shock patients