Serum bicarbonate and anion gap cannot reliably predict elevated serum lactate levels

Abstract

Study objectives: New treatment strategies for severe sepsis and septic shock have made it important to rapidly identify patients with serum lactate levels greater than 4.0 mmol/L. The measurement of serum electrolyte levels is generally considered routine for patients with significant medical illness; the measurement of lactate levels is not as universal. We ask how well a reduced serum bicarbonate level ( 3 -Cl >20) predict a lactate level of 4.0 mmol/L or greater. Methods: This was a retrospective medical record review at a university hospital. Laboratory results were reviewed for patients 18 years of age or older who, during a 2-month period, had time-matched determinations of serum lactate and electrolyte levels. The correlation of lactate with both bicarbonate and anion gap was determined, and the test properties of each as a screening test for a lactate greater than 4.0 mmol/L were determined. Results: One hundred twenty-five patients (58% female patients, age range 2 months to 97 years) had lactate and electrolyte levels measured concurrently during the study period. Lactate level was greater than 4.0 mmol/L in 17% of patients. The sensitivity and specificity of a reduced bicarbonate were 38% (95% confidence interval [CI] 18% to 62%) and 78% (95% CI 69% to 85%), respectively. The sensitivity and specificity of an elevated anion gap were 38% (95% CI 18% to 62%) and 89% (95% CI 82% to 95%), respectively. There was only minimal correlation between bicarbonate or anion gap and lactate level ( r =0.06 and 0.11, respectively). Conclusion: A lactate level greater than 4.0 mmol/L cannot be predicted from either reduced bicarbonate or an elevated anion gap. Measuring bicarbonate or calculating the anion gap cannot be considered appropriate screens when attempting to identify patients who may benefit from aggressive sepsis management.