LAUSR

Background: We aimed to evaluate the impact of metformin use on lactate kinetics in patients with severe sepsis and septic shock. Methods: We analyzed data from a registry that included patients who presented to the emergency department and met criteria for severe sepsis or septic shock. Patients were divided into two groups based on metformin use. We compared lactate concentrations, lactate clearance (LC), and normalization at 6 h (H6) and 24 h (H24) after the initial (H0) measurement. Propensity score matching, multiple logistic, and linear regression analysis via a generalized estimating equations method were used. Results: Of 1,318 patients, 71 patients were in the metformin use group and all 71 were selected in a one to two propensity matching. Metformin users showed significantly higher lactate levels at H0 (5.3 vs. 4.4 mmol/L) and H6 (3.8 vs. 2.9 mmol/L) in all patients, although in the matched subset, the effect was marginal (H0, 5.3 vs. 4.9 mmol/L; H6, 3.8 vs. 3.2 mmol/L; H24, 2.7 vs. 2.4 mmol/L). Mean LC (H6, 29% vs. 34%; H24, 43% vs. 49%) and normalization rate (H6, 27% vs. 28%; H24, 49% vs. 52%) were also not significantly different. Although metformin use appeared to be associated with higher lactate levels before using the propensity score method, no significant association was found between metformin use and lactate kinetics variables in the balanced matched subset data. Conclusions: Lactate levels in metformin users were initially elevated in the early phase of resuscitation from severe sepsis and septic shock. However, there was no significant difference in lactate levels, LC, and normalization over the initial 24 h period based on metformin use.