RESUMO
BACKGROUND: Applying peripheral venous lactate instead of arterial lactate in clinical practice is questionable because of deviation between both values. We aimed to find the relationship between the arterial lactate and the peripheral venous lactate before reasoned that the venous lactate could be used in substitution to the arterial lactate in sepsis. METHODS: We conducted a prospective, cross-sectional study at a university hospital. The patients with sepsis in ICU who required lactate level monitoring were enrolled in this research. The correlation and agreement between arterial lactate (A-LACT) and peripheral venous lactate (V-LACT) were the primary outcomes. RESULTS: A total of 63 paired samples were collected. The A-LACT and V-LACT were strongly correlated ( r = .934, P < .0001, r2 = .873). The regression equation was A-LACT = (0.934 × V-LACT) - 0.236. The mean difference between V-LACT and A-LACT was 0.66 ± 1.53 mmol/L. The 95% limits of agreement were between -3.66 and 2.33 mmol/L. The V-LACT ≥ 4 mmol/L can predict A-LACT level ≥ 4 mmol/L with 87.5% sensitivity and 91.5% specificity, and the area under receiver operating characteristic curve was 0.948. CONCLUSION: The present study demonstrated a strong correlation between A-LACT and V-LACT, but an agreement between both parameters was poor. We suggest not to use the V-LACT in substitution to the A-LACT in sepsis regarding the absolute value and clearance rate, but the V-LACT ≥ 4.5 mmol/L may be used for predicting the A-LACT ≥ 4 mmol/L.
