广西医学

目的探讨应激性高血糖比值（SHR）与脓毒症患者院内死亡风险的关系。方法从重症医学数据库Ⅳ（MIMIC⁃Ⅳ V2.2）中提取1 986例首次入住ICU的脓毒症患者的临床资料，并根据患者的院内死亡情况将患者分为存活组和死亡组。构建Logistic回归模型探讨脓毒症患者院内死亡的危险因素；通过受试者工作特征（ROC）曲线评估SHR对脓毒症预后的预测能力；根据性别、年龄、合并症等因素，对SHR与院内死亡率之间的关系进行亚组分析，进一步探究SHR与分层变量之间的交互作用。结果共有305例患者发生院内死亡，死亡率为15.4%。死亡组年龄、入院血糖、SHR、未合并高血压患者比例、感染合并序贯性器官衰竭评价（SOFA）评分、简化急性生理功能评估系统Ⅱ（SAPSⅡ）评分、机械通气时间、ICU住院天数高于或长于存活组（P<0.05）；多因素Logistic回归分析结果显示，SHR升高时脓毒症患者院内死亡风险增加（P<0.05）。SHR联合SAPSⅡ评分预测脓毒症患者院内死亡的曲线下面积为0.667，高于SHR单独预测的0.570（P<0.05）。亚组分析结果显示，在男性亚组、女性亚组、年龄<65岁亚组、机械通气亚组、SHR≥1.10亚组、未应用糖皮质激素亚组、合并糖尿病亚组、未合并糖尿病亚组、未合并低血糖亚组、未合并高血压亚组、未合并心衰亚组中，SHR升高与脓毒症患者院内死亡风险增加相关（P<0.05）。交互作用结果显示，在合并糖尿病与未合并糖尿病、合并高血压与未合并高血压的人群中，SHR与脓毒症患者院内死亡率之间的关系差异有统计学意义（P交互作用<0.05）。结论 SHR升高与脓毒症患者院内死亡风险增加密切相关，SHR升高是影响脓毒症患者院内死亡的独立危险因素，SHR与SAPSⅡ评分联合可以更好地评估脓毒症患者的预后，对于脓毒症患者的临床治疗和决策具有重要参考价值。

Objective To explore the relation between stress hyperglycemia ratio (SHR) and in⁃hospital death risk in patients with sepsis. Methods The clinical data of 1986 patients with sepsis who were firstly admitted to ICU were extracted from the critical care medical database Ⅳ (MIMIC⁃Ⅳ V2.2). Patients were assigned to survival group or death group according to patients' in⁃hospital death states. The Logistic regression model was established to explore the risk factors for in⁃hospital death in patients with sepsis. The prediction ability of SHR on prognosis of sepsis was evaluated by employing the receiver operating characteristic (ROC) curve. According to the factors in terms of gender, age, comorbidities, etc., a subgroup analysis was performed on the relation between SHR and in⁃hospital mortality, and then the interaction effect between SHR and stratification variables was further explored. Results A total of 305 patients were dead in hospital, with the incidence rate of death in 15.4%. The death group exhibited higher age, blood glucose at admission, SHR, proportion of patients without concomitant hypertension, infection and concomitant Sequential Organ Failure Assessment (SOFA) score, Simplified Acute Physiology System Ⅱ (SAPSⅡ) score, and longer mechanical ventilation duration, length of ICU stays as compared with the survival group (P<0.05). The results of multivariate Logistic regression analysis revealed that when SHR was elevated, in⁃hospital death risk of patients with sepsis was increased (P<0.05). Area under the curve of SHR combined with SAPSⅡ score for predicting in⁃hospital death of patients with sepsis was 0.667, which was higher than that of SHR single prediction (0.570, P<0.05). The results of subgroup analysis indicated that in the male subgroup, the female subgroup, age<65 years old subgroup, mechanical ventilation subgroup, SHR≥1.10 subgroup, non⁃administration of glucocorticoids subgroup, concomitant diabetes mellitus subgroup, non⁃concomitant diabetes mellitus subgroup, non⁃concomitant hypoglycemia subgroup, non⁃concomitant hypertension subgroup, non⁃concomitant heart failure subgroup, SHR elevation correlated with increased risk of in⁃hospital death in patients with sepsis (P<0.05). The results of interaction effect depicted that in concomitant and non⁃concomitant diabetes mellitus, concomitant and non⁃concomitant hypertension populations, there was a statistically significant difference in relation between SHR and in⁃hospital mortality in patients with sepsis (Pinteraction effect<0.05). Conclusion SHR elevation is closely related to increased risk of in⁃hospital death in patients with sepsis, and SHR elevation is the independent risk factor for affecting in⁃hospital death in patients with sepsis. The combination of SHR and SAPSⅡ score can preferably evaluate prognosis of patients with sepsis, exerting crucial reference values for clinical treatment and decision on patients with sepsis.