Which one is a better predictor of ICU mortality in septic patients? Comparison between serial serum lactate concentrations and its removal rate

Purpose: To predict 28-day mortality with serum lactate and oxygenation profile in sepsis. Methods: 82 patients were admitted to the ICU with sepsis. Comorbid disease, hemodynamic and oxygenation parameters were recorded. Serum lactate was measured at T0, T6, T12 and T24 hours of admission. Arterial and venous oxygen saturation levels were also measured. Regression and ROC analyses were used to predict death within 28 days. Results: Out of 82 patients, 32 died within 28 days of ICU admission. Non-survivors differed from survivors in having higher serum lactate concentrations on admission (0.6 mmol/L; P = 0.033), requiring more norepinephrine (14 μg/min; P < 0.001), higher frequency of acute kidney injury, prolonged mechanical ventilation (5-days; P < 0.001) and ICU stay (1-day; P = 0.029). Saturation of oxygen in arterial (a), central venous blood (cv) and (a–cv) were similar between the survivors and non-survivors. T24 level of lactate was the best predictor of 28-day mortality with 78% sensitivity and 90% specificity (AUC = 0.912 ± 0.033). Conclusion: Serial measurements of serum lactate with special emphasis on its concentration at 24 hour after admission remains the most predictive of short-term mortality in the ICU. Other predictors of mortality are relatively inferior and must be used collectively in context to better predict the clinical outcome of sepsis.