【摘要】 目的 探究老年重症监护病房（intensive care unit, ICU）患者短期和长期全因死亡率与老年营养风险指数（geriatric nutritional risk index,GNRI）之间的关系。方法 以重症医学MIMIC-Ⅳ(medical information mart for intensive careⅣ)数据库中符合纳排标准的患者为研究对象，以患者入ICU后28、90和365 d是否发生全因死亡为结局变量，以人口学特征、实验室指标及合并症等为协变量，将患者分为高GNRI组（GNRI≥98）和低GNRI组（GNRI<98），并采用倾向评分匹配法将两组进行匹配，利用Cox回归及限制性立方样条探究GNRI和患者全因死亡率之间的关联，采用受试者工作特征曲线下面积(area under curve, AUC)评价GNRI对患者365 d死亡风险的预测效能，同时采用分层分析探究不同亚组人群中GNRI对患者全因死亡风险的影响。结果 共纳入4 959例患者，28、90和365 d患者死亡率分别为23.0%、30.3%和38.7%。倾向性评分匹配后，低GNRI组患者28、90和365 d死亡风险分别为高GNRI组患者死亡风险的1.373(95%CI:1.186～1.589)、1.409(95%CI:1.240～1.600)和1.457(95%CI:1.301～1.631)倍（P<0.05），且限制性立方样条表明两者呈现非线性关系，GNRI对患者1年死亡风险的预测效能AUC为0.653 (95%CI:0.638～0.669)，亚组分析表明低GNRI对男性老年患者影响更大。结论 GNRI与老年ICU患者28、90和365 d全因死亡风险密切相关，GNRI可为老年ICU患者短期和长期预后提供临床决策参考。更多还原

【Abstract】 Objective To investigate the relationship between short-term and long-term all-cause death rate and Geriatric Nutritional Risk Index(GNRI) in geriatric intensive care unit(ICU). Methods The patients in MIMIC-IV(Medical Information Mart for Intensive Care IV) database who met the criteria of inclusion were selected as the study subjects. Whether allcause death occurred at the 28, 90, and 365 days after admission of ICU was taken as the outcome variable, and the demographic characteristics, laboratory indexes, and complications were taken as covariables. The patients were divided into high GNRI group(GNRI ≥ 98) and low GNRI group(GNRI < 98). The propensity score matching method was used to match the two groups, and Cox regression and restricted cube splines were used to explore the relationship between GNRI and all-cause mortality. The area under the receiver working characteristic curve(AUC) was used to evaluate the predictive efficacy of GN-RI on the risk of all-cause death. Stratified analysis was used to explore the effect of GNRI on the risk of all-cause death in different subgroups. Results A total of 4 959 patients were enrolled, and the mortality rates of 28 days, 90 days, and 365days were 23.0%, 30.3%, and 38.7%, respectively. After matching the propensity score, the death risk of patients with low GNRI at 28, 90, and 365 days was 1.373(95%CI: 1.186-1.589), 1.409(95%CI: 1.240-1.600), and 1.457(95%CI: 1.301-1.631) times higher than those of patients with high GNRI, respectively, and the restricted cubic spline showed that there was a non-linear relationship between them. The predictive efficiency of GNRI for one-year death risk was 0.653(95%CI: 0.638-0.669). Subgroup analysis showed that low GNRI had a greater impact on male elderly patients. Conclusion GNRI is closely related to the risk of all-cause death in elderly patients with ICU at 28, 90, and 365 days. GNRI can provide clinical reference for short-term and long-term prognosis of elderly patients with ICU.更多还原