Lactic acid is a marker of life‐threatening illness, and any delay between recognizing an increased level and initiating treatment may be dangerous. The presence or absence of an anion gap (AG) has classically been used as a screening tool for lactic acidosis. Prior studies of the reliability of AG as a screening tool in the emergency department (ED) setting have produced conflicting results. We sought to determine the value of the AG for detecting the presence of lactic acidosis in adults presenting to the ED.
The authors conducted a retrospective cohort study over a 24‐month period on subjects 18 years or older presenting to the ED. The AG was calculated from initial basic chemistry panel with lac‐tate levels drawn within 1 hour of the electrolytes. Receiver operating characteristic (ROC) curves were created using AG versus lactate level with abnormal lactate variously defined as 2.5 and 4.0 mmol/L. The area under the curve (AUC) was calculated, and Youden's index was used to determine the optimal AG cut point for detection of an abnormal lactate level. Finally, 2 × 2 tables using the optimal AG were used to calculate sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for 2 levels of abnormal lactate.
A total of 13,456 patients were included in the study. AUC was 0.82 for a lactate cutoff of 2.5 mmol/L and 0.90 for a lactate cutoff of 4.0 mmol/L. The optimal cut point for AG appeared to be 12 for both definitions of abnormal lactate. Using an AG > 12, the sensitivity for detecting lactic acid > 2.5 mg/dL was 66% (64%‐67%), specificity 85% (85%‐86%), PPV 53% (52%‐54%), and NPV 91% (90%‐91%). Similarly, for a lactate > 4.0 mmol/L, the sensitivity was 86% (84%‐88%), specificity 81% (81%‐81%), PPV 28% (27%‐28%), and NPV 99% (98%‐99%).
We found high sensitivity and very high NPV for abnormal lactate with an AG cut point of 12. The AG can be rapidly calculated from a screening electrolyte profile done shortly after ED presentation. Perhaps the greatest utility of the AG in the ED is in excluding hyperlactemia from consideration. Patients with an anion gap ≤ 12 are unlikely to have clinically significant lactic acidosis.
S. Namo ‐ none; K. Olarte‐Meridia ‐ none; W. Oureshi ‐ none; N. Hector ‐none; G. Buran ‐ none