The lactate-to-albumin ratio as a potential biomarker for short-term mortality risk in critically ill patients with urosepsis: a retrospective study with dual-cohort validation.

Background: The lactate-to-albumin ratio (LAR) comprehensively reflects the nutritional status, inflammatory level, and degree of oxidative stress in critically ill patients and has been demonstrated to be associated with poor prognosis in various critical illnesses. However, its prognostic role in the specific context of urosepsis remains to be fully elucidated. Methods: This study utilized the Medical Information Mart for Intensive Care (MIMIC-IV) database as an internal discovery cohort and included an external validation cohort from Bijie Hospital of Zhejiang Provincial People's Hospital. Multivariable Cox regression, restricted cubic splines (RCS), Kaplan-Meier survival curves, and other methods were employed to analyze the association between LAR and short-term adverse outcomes in urosepsis patients. The robustness of the findings was assessed through subgroup analysis and interaction tests. Furthermore, four machine learning algorithms were combined to screen key variables for constructing a multivariable Cox risk prediction model, the performance of which was evaluated using receiver operating characteristic (ROC) curves. Results: A total of 1,055 urosepsis patients were included. The 28-day ICU mortality and in-hospital mortality rates were 20.6 and 19.1%, respectively. After multivariable adjustment, for each unit increase in LAR as a continuous variable, the hazard ratios (HR) for ICU mortality and in-hospital mortality were 1.35 (95% CI: 1.08-1.68) and 1.50 (95% CI: 1.19-1.90), respectively. When treated as a categorical variable, the high LAR group had a 73% increased risk of ICU mortality and a 71% increased risk of in-hospital mortality compared to the low LAR group. RCS analysis revealed a non-linear positive dose-response relationship between LAR and mortality. ROC analysis demonstrated that a simple risk model incorporating LAR and five other variables had superior predictive performance compared to traditional critical illness scores, and these conclusions were consistently validated in the external cohort. Conclusion: The LAR is a cost-effective supplementary predictor of short-term mortality risk in critically ill urosepsis patients, primarily aiding early identification of high-risk cases to intensify monitoring and treatment. Its clinical value require further prospective validation.