The Correlation Between Serum Glucose-Potassium Ratio and Mortality Risk in Sepsis Patients: Analysis based on the MIMIC-IV database

BackgroundSepsis is a systemic inflammatory response syndrome triggered by infection, often leading to multiple organ dysfunction and high mortality rates. Despite advancements in critical care, predicting the prognosis of sepsis remains challenging. This study explores the predictive value of the Glucose-Potassium Ratio (GPR) for all-cause mortality (ACM) in sepsis patients. MethodsThis retrospective cohort study utilized data from the MIMIC-IV 3.0 database, which included adult sepsis patients admitted to the ICU between 2008 and 2022. A total of 9,360 patients were analyzed based on GPR stratification (Q1 [&le;] 25.35, Q2 = 25.35-34.69, Q3 [&ge;] 34.69). Kaplan-Meier survival curves, Cox regression models, and restricted cubic splines (RCS) were employed to assess GPRs association with mortality at 28 days, 90 days, 180 days, and 1 year. Subgroup analyses based on age, gender, race, and comorbidities were also performed. ResultsHigher GPR was associated with significantly reduced mortality at all time points (28 days, 90 days, 180 days, and 1 year), with hazard ratios for Q3 ranging from 0.71 to 0.79 (P < 0.001). Kaplan-Meier curves showed better survival for patients with higher GPR. Subgroup analysis revealed stronger protective effects in older adults ([&ge;]60 years), females, and Asians. GPR outperformed glucose and potassium alone in predicting mortality, with higher AUC values for GPR across different time points. ConclusionGPR is an independent predictor of sepsis mortality, with a threshold of [&ge;]34.69 associated with reduced risks of mortality at multiple time points. It offers a superior predictive capacity compared to glucose or potassium levels alone, supporting its potential use as a clinical tool for early risk stratification and individualized treatment of sepsis patients.