【摘要】 目的：评估炎症生物标志物中性粒细胞—淋巴细胞比值（neutrophil-to-lymphocyte ratio,NLR）、单核细胞—淋巴细胞比值（monocyte-to-lymphocyte ratio,MLR）和全身炎症反应指数（systemic inflammatory response index,SIRI）对脓毒症相关急性肾损伤（sepsis-associated acute kidney injury,SA-AKI）患者28 d全因死亡率的预测能力。方法：基于重症监护医学信息集市Ⅳ（medical information mart for intensive care-Ⅳ，MIMIC-Ⅳ）数据库，筛选符合脓毒症3.0标准和改善全球肾脏病预后组织（Kidney Disease:Improving Global Outcomes,KDIGO）急性肾损伤标准的成年ICU住院患者。主要结局为28 d全因死亡率。采用Cox比例风险模型分析炎症生物标志物与死亡率的关系，通过受试者工作特征（receiver operating characteristic,ROC）曲线评估NLR、MLR、SIRI单独及联合序贯器官衰竭评分（sequential organ failure assessment,SOFA）的预测效能，并利用限制性立方样条（restricted cubic spline,RCS）曲线分析上述指标与死亡风险的非线性关系。结果：在5 739例SA-AKI患者中，与存活组相比，死亡组NLR、MLR和SIRI升高（P<0.001）。校正后的Cox比例风险模型显示，NLR(HR=1.52)、MLR(HR=1.60)和SIRI(HR=1.60)为28 d死亡结局的独立预测因子。NLR与SOFA联合预测的ROC曲线下面积（area under the curve,AUC）为0.631，高于SOFA单独预测（AUC=0.572）。RCS分析显示，NLR、MLR、SIRI与死亡率相关的临界值分别为8.902、0.511、5.006(P<0.001)。亚组分析证实，上述指标与死亡率的相关性在不同年龄、性别和合并症患者中保持稳定，无显著的交互作用（P>0.05）。结论：入院时NLR、MLR和SIRI为SA-AKI患者28 d全因死亡率的独立预测指标，这些生物标志物联合SOFA可提高预后评估能力，为ICU中的风险分层提供了一种经济、有效的工具。更多还原

【Abstract】 Objective: To evaluate the ability of the neutrophil-to-lymphocyte ratio(NLR), monocyte-tolymphocyte ratio(MLR), and systemic inflammatory response index(SIRI) for predicting 28-day all-cause mortality in sepsis-associated acute kidney injury(SA-AKI) patients. Methods: This study utilized the MIMIC-Ⅳ(medical information mart for intensive care-Ⅳ) database to identify adult ICU patients meeting sepsis-3 and Kidney Disease: Improving Global Outcomes(KDIGO) criteria for SA-AKI. The primary outcome was defined as all-cause mortality at 28 days. Cox proportional hazards models were used to analyze the associations between inflammatory biomarkers and mortality. The predictive performance of NLR, MLR, and SIRI alone and in combination with the sequential organ failure assessment(SOFA) score was evaluated using receiver operating characteristic(ROC) curves. Restricted cubic spline(RCS) curves were applied to explore the nonlinear relationship between the above indicators and mortality risk. Results: Among 5,739 SA-AKI patients, survivors exhibited significantly higher NLR, MLR, and SIRI levels than non-survivors(all P<0.001). Adjusted Cox models identified NLR [HR(hazard ratios)=1.52], MLR(HR=1.60), and SIRI(HR=1.60) as independent mortality predictors. The area under the ROC curve(AUC) for the combined prediction of NLR and SOFA was 0.631, which was higher than that for SOFA alone(AUC=0.572). Restricted cubic spline(RCS) analysis revealed that the cutoff values of NLR, MLR, and SIRI associated with mortality were 8.902, 0.511, and 5.006, respectively(P<0.001). Subgroup analysis confirmed that the associations of the above indicators with mortality remained stable across patients of different ages, genders, and comorbidities, with no significant interaction observed(P>0.05). Conclusion: Admission NLR, MLR, and SIRI independently predict 28-day all-cause mortality in SA-AKI patients. Integrating these biomarkers with SOFA enhances prognostication, providing a cost-effective tool for risk stratification in the ICU.更多还原