The base excess gap is not a valid tool for the quantification of unmeasured ions in cardiac surgical patients: a retrospective observational study.

Background: The base excess gap (BE(gap)) method is commonly used for the quantification of unmeasured ions in critically ill patients. However, it has never been validated against the standard quantitative acid-base approach.
Objective: To compare the BE(gap) as a tool for the prediction of the excess of unmeasured ions with the offset of strong ion gap (SIG) from its reference value.
Design: A retrospective observational study.
Setting: Adult ICU in a tertiary hospital.
Patients: One hundred and thirty-five cardiac surgical patients admitted for postoperative care.
Interventions: None.
Main Outcome Measures: BE(gap) was calculated as BE(gap) = SBE - BE(si) - BE(wa), where SBE is the standard base excess, BE(si) is the partition due to strong ions ([Na+]-[Cl-]-[lactate-] - 30.5) and BE(wa) is the partition due to weak acids [0.25×{42 - (albumin)}]. The deviation of the observed SIG (SIG(ob)) from its reference value was calculated as deltaSIG = 2.85 - SIG(ob). We used Bland-Altman and concordance correlation analysis to compare BE(gap) with deltaSIG. A bias of ±1 meq l(-1) with limits of agreement of ±2 meq l(-1) and a concordant correlation coefficient of more than 0.9 were considered to indicate a strong agreement.
Results: The concordant correlation coefficient between BE(gap) and deltaSIG was 0.702. The mean bias between the two variables was 1.8 meq l(-1), with a lower limit of agreement of -0.9 meq l(-1) and an upper limit of agreement of 4.4 meq l(-1).
Conclusion: The BE gap method cannot reliably quantify the unmeasured ion excess in cardiac surgical patients. Clinicians should use the full Stewart-Figge model for quantitative acid-base assessments.