Anion gap corrected for albumin, phosphate and lactate is a good predictor of strong ion gap in critically ill patients: a nested cohort study.

Objective: Corrected anion gap and strong ion gap are commonly used to estimate unmeasured anions. We evaluated the performance of the anion gap corrected for albumin, phosphate and lactate in predicting strong ion gap in a mixed population of critically ill patients. We hypothesized that anion gap corrected for albumin, phosphate and lactate would be a good predictor of strong ion gap, independent of the presence of metabolic acidosis. In addition, we evaluated the impact of strong ion gap at admission on hospital mortality.

Methods: We included 84 critically ill patients. Correlation and agreement between the anion gap corrected for albumin, phosphate and lactate and strong ion gap was evaluated by the Pearson correlation test, linear regression, a Bland-Altman plot and calculating interclass correlation coefficient. Two subgroup analyses were performed: one in patients with base-excess <-2 mEq/L (low BE group - lBE) and the other in patients with base-excess >-2 mEq/L (high BE group - hBE). A logistic regression was performed to evaluate the association between admission strong ion gap levels and hospital mortality.

Results: There was a very strong correlation and a good agreement between anion gap corrected for albumin, phosphate and lactate and strong ion gap in the general population (r2=0.94; bias 1.40; limits of agreement -0.75 to 3.57). Correlation was also high in the lBE group (r2=0.94) and in the hBE group (r2=0.92). High levels of strong ion gap were present in 66% of the whole population and 42% of the cases in the hBE group. Strong ion gap was not associated with hospital mortality by logistic regression.

Conclusion: Anion gap corrected for albumin, phosphate and lactate and strong ion gap have an excellent correlation. Unmeasured anions are frequently elevated in critically ill patients with normal base-excess. However, there was no association between unmeasured anions and hospital mortality.