Elevated osmol gaps in patients with alcoholic ketoacidosis

The use of the osmol gap as a surrogate marker of toxic alcohol poisoning is common. Unfortunately, many patients with alcoholic ketoacidosis have elevated osmol gaps and are misdiagnosed with toxic alcohol poisoning. We aimed to characterize the range of osmol gaps in patients with alcoholic ketoacidosis.

This was a retrospective poison center study. Data from 24 years were reviewed using the following case definition of alcoholic ketoacidosis: (1) documented alcohol use disorder; (2) presence of urine or serum ketones or an elevated blood beta-hydroxybutyrate concentration; (3) an anion gap ≥14 mmol/L. Potential cases of alcoholic ketoacidosis that failed to fulfill all three criteria were adjudicated by three toxicologists. Exclusion criteria included (1) detectable toxic alcohol concentration, (2) hemodialysis and/or multiple doses of fomepizole, (3) no osmol gap documented, (4) other diagnoses that lead to a metabolic acidosis. Demographics, pH, anion gap, lactate concentration, and osmol gap were extracted.

Of 1,493 patients screened, 55 met criteria for alcoholic ketoacidosis. Sixty-four percent were male, and their median age was 52 years. The median osmol gap was 27 [IQR 18-36]. The largest anion gap was 57 mmol/L, and the lowest pH was 6.8. Forty-five (82%) of the patients with alcoholic ketoacidosis had osmol gaps >10; 38 (69%) had osmol gaps >20; 24 (44%) had osmol gaps >30; 11 (20%) had osmol gaps > 40.

The large range of osmol gaps in patients with alcoholic ketoacidosis often reaches values associated with toxic alcohol poisoning. The study is limited by the potential for transcribing errors and the inability to identify the cause of the osmol gap.

In this retrospective study, patients with alcoholic ketoacidosis had a median osmol gap of 27. Given that alcoholic ketoacidosis is easily and inexpensively treated, proper identification may prevent costly and invasive treatment directed at toxic alcohol poisoning.