The strong ion gap does not have prognostic value in critically ill patients in a mixed medical/surgical adult ICU.

OBJECTIVE: To examine whether the strong ion gap (SIG) or standard base excess corrected for abnormalities of serum chloride and albumin (BE(UA)) can predict outcome and to compare the prognostic abilities of these variables with standard base excess (SBE), anion gap (AG), pH, and lactate, the more traditional markers of acid-base disturbance.

DESIGN: Prospective, observational study.

SETTING: University teaching hospital, general adult ICU.

PATIENTS: One hundred consecutive patients on admission to the ICU.

MEASUREMENTS AND RESULTS: The anion gap (AG) was calculated and corrected for abnormal serum albumin (AG(corrected)). Serum lactate was measured and SBE, BE(UA), SIG, and APACHE II scores calculated for each patient. 28-day survival was recorded. There was a significant difference between the mean APACHE II (P < 0.001), SBE (P < 0.001), lactate (P = 0.008), AG (P = 0.007), pH (P < 0.001), and BE(UA) (P = 0.009) of survivors and non-survivors. There was no significant difference between the mean SIG (P = 0.088), SIDeff (P = 0.025), and SID app (P = 0.254) between survivors and non-survivors. The pH and SBE demonstrated the best ability of the acid-base variables to predict outcome (AUROC curves 0.72 and 0.71, respectively). Neither of these were as good as the APACHE II score (AUROC 0.76)

CONCLUSION: Traditional indices of SBE, BE(UA,) lactate, pH, AG, and APACHE II all discriminated well between survivors and non-survivors. In this group of patients the SIG, SIDeff, and SIGapp appear to offer no advantage in prediction of outcome and their use as prognostic markers can therefore not be advocated.