[Skip to Content]
[Skip to Content Landing]
February 1990

The Fall of the Serum Anion Gap

Author Affiliations

From the Clinical Laboratory (Dr Winter) and the Department of Medicine (Dr Gabow), Denver General Hospital, Colo; the Departments of Pathology (Drs Winter and Schultz), Medicine (Drs Pearson and Gabow), and Clinical Pathology and Laboratory Medicine (Dr Lepoff), University of Colorado School of Medicine; and the Laboratory Service, Veterans Administration Medical Center (Dr Schultz), Denver.

Arch Intern Med. 1990;150(2):311-313. doi:10.1001/archinte.1990.00390140057012

• Using modern electrode technology (Beckman ASTRA analyzer), we evaluated the reference range for the anion gap (calculated as sodium minus chloride minus bicarbonate concentrations) in serum to determine whether the 8 to 16 mmol/L reference range in common use is still valid. After measurement of electrolytes in (1) serum from 29 healthy volunteers, (2) aqueous standards verified against National Bureau of Standards reference material, and (3) serum from 120 blood donors, we drew the following conclusions. (1) The reference range for the anion gap has shifted downward (to 3 to 11 mmol/L in one of our laboratories), primarily because of an upward shift in chloride values. (2) Using the ASTRA analyzer, a majority of normal individuals can be expected to have serum anion gaps of 6 mmol/L or less unless chloride calibration is deliberately altered. (3) If the anion gap is to remain an effective tool in diagnosing acid-base disorders, clinicians need to be aware that the traditional reference range may not be appropriate with new instrumentation.

(Arch Intern Med. 1990;150:311-313)