Association of lactate to Po2 ratio (LPR) with 28-day mortality in ICU sepsis: A retrospective cohort study based on the MIMIC-IV database

The high mortality rate among patients with sepsis continues to be a major hurdle in critical care environments, and metabolic imbalances are a defining feature of the underlying pathophysiology of this disease. However, the association between the lactate-to-oxygen tension ratio (LPR) and patient outcomes in sepsis remains underexplored, with insufficient research to fully elucidate its clinical significance. In the context of assessing a patient’s metabolic health, lactate alone might not be sufficient to fully capture their physiological condition, as it is influenced by non-hypoxic factors including hepatic dysfunction, catecholamine administration, and renal clearance impairment. The integration of lactate and arterial partial pressure of oxygen (PaO 2 ) into the LPR offers a composite measure that simultaneously reflects tissue metabolic stress and systemic oxygenation status, potentially providing a more comprehensive representation of the metabolic-oxygenation mismatch characteristic of sepsis. This research examines the relationship between LPR and 28-day mortality rates among patients with sepsis admitted to the intensive care unit. According to ICD-9 diagnosis codes, patients diagnosed with sepsis in the Medical Information Mart for Intensive Care IV (v.2.2) database were selected. The receiver operating characteristic (ROC) curve, Kaplan–Meier (K–M) survival curve, logistic regression analyses, sensitivity analysis, and subgroup analysis were used to assess the predictive power of LPR. Based on the MIMIC-IV database, patients with sepsis admitted to the ICU for the first time (n = 14,921) were included. The 28-day mortality was 17.26%. The area under the ROC curve (AUC) for LPR was 0.725 (95% confidence interval [CI], 0.714–0.73), which was comparable to that of OASIS (0.765; 95% CI, 0.755–0.775) and SAPSII (0.772; 95% CI, 0.762–0.781), and superior to SOFA (0.624; 95% CI, 0.612–0.636; P = .500). The cutoff value of LPR was 0.137 based on the ROC curve. The high LPR group showed a poor prognosis in the K–M analysis ( P < .001). Multivariate logistic regression showed that LPR10 was significantly associated with poor outcome (odds ratio, 1.61; 95% CI, 1.42–1.84; P < .001), after adjusting for lactate alone and established severity scores. Net Reclassification Improvement analyses confirmed incremental prognostic value of LPR beyond SOFA (NRI 0.142; P < .001), SAPS II (NRI 0.118; P < .001), and OASIS (NRI 0.095; P < .001). Subgroup analysis showed no significant interaction between LPR and any subgroup ( P for interaction: 0.065–0.867). LPR is a rational and easily accessible marker that is strongly associated with 28-day mortality in ICU patients with sepsis. The integration of lactate and PaO 2 into a single ratio captures prognostic information beyond that provided by either parameter alone or by established severity scores, offering a practical bedside tool for early risk stratification.