Clinical utility of serum biochemical variables for predicting acid–base balance in critically ill horses

Abstract

BackgroundProfiles from serum biochemical analyzers include the concentration of strong electrolytes (includingl‐lactate), total carbon dioxide (tCO2), and total protein. These variables are associated with changes in acid–base balance. Application of physicochemical principles may allow predicting acid–base balance from serum biochemistry without measuring whole bloodpHandpCO2.ObjectivesThe purpose of the study was to determine if the acid–base status of critically ill horses could be accurately predicted using variables included in standard serum biochemical profiles.MethodsTwo jugular venous blood samples were prospectively obtained from critically ill horses and foals. Samples were analyzed using a whole blood gas andpHanalyzer (BG) and a serum biochemistry multi analyzer system (AMAS). Linear regression, Deming regression, and Bland–Altman plots were used for method comparison andP < .05 was considered significant.ResultsValues from 70 horses and foals for Na, K, Cl, and total protein concentrations, and consequently the calculated variables used for acid base interpretation, were different between the AMAS and BG analyzer. Using physicochemical principles, BG results accurately predictedpH, whereas the AMAS results did not when a fixed value forpCO2was used.ConclusionsMeasurement ofpCO2is required in critically ill horses for accurate prediction of whole bloodpH. Differences in the measured values of Na and Cl concentration exist when measured in serum by the AMAS and in whole blood or plasma by BG, indicating that the accurate prediction of whole bloodpHis analyzer‐dependent. Application of physicochemical principles to plasma or serum provides a practical method to evaluate analyzer accuracy.