Coagulation risk in sepsis patients with elevated lactate-to-albumin ratio: a retrospective cohort study

Sepsis-induced coagulopathy (SIC) is commonly observed in patients with sepsis and is correlated with an elevated risk of mortality. The objective of this study is to analyse the association between the lactate-to-albumin ratio (LAR) and the occurrence of SIC, as well as all-cause mortality.

From the MIMIC-IV3.1 database, patients with sepsis who fulfilled the inclusion criteria were selected. The relationship between LAR and SIC was assessed by constructing a multivariate logistic regression model and performing Restricted cubic spline (RCS) analysis. To evaluate the predictive ability of different biomarkers for SIC, we performed ROC curve analysis. Subgroup analysis verifies the robustness of the results. Additionally, Cox regression analysis was conducted to investigate the association between LAR and mortality.

A total of 8,661 patients with sepsis were included in the final analysis, of whom 5,784 developed SIC. In the comprehensive logistic regression model, each additional unit of LAR increased the odds of SIC rather by 1.85 (OR, 2.85, 95% CI, 2.48-3.28; p < 0.001). The odds of SIC rather in the highest LAR group was significantly increased (T3 vs T1: OR, 3.57; 95% CI, 3.06-4.17; p < 0.001). The relationship between LAR levels and SIC incidence was found to be nonlinear and positive, as revealed by RCS analysis. ROC analysis indicated superior predictive performance of LAR (AUC = 0.71) compared to albumin (AUC = 0.61) or lactate (AUC = 0.69). Subgroup analysis indicates a stable relationship between LAR and SIC. Moreover, an increase in LAR was correlated with a heightened risk of mortality at both 28 days and 360 days (28-d HR, 1.15; 95% CI, 1.11-1.19; 360-d HR, 1.14; 95% CI, 1.11-1.18; all p < 0.001).

Observational findings indicate that elevated early LAR levels may be associated with increased rates of SIC and all-cause mortality in patients with sepsis. However, further prospective studies are needed to validate these associations and assess their clinical utility.