摘要
目的探讨真实世界ICU常见疾病的血糖控制与患者短期预后的关系。方法本研究是一项回顾性队列研究,数据来源为公开数据库MIMIC-Ⅲ。选择首次入院疾病诊断相关分组(DRG)诊断为心脏疾病、肾脏疾病、呼吸系统疾病、颅脑疾病、脓毒症&传染病的成人ICU患者9794例作为研究对象,以院内死亡、30 d死亡作为研究终点。计算时间加权平均血糖(TWAG),将所有患者按TWAG水平分为4组:0~6.1 mmol/L、6.1~7.8 mmol/L、7.8~10.0 mmol/L、≥10.0 mmol/L,比较各组患者临床特点、终点事件发生情况。建立终点事件的Logistic回归模型,进行亚组分析、交互作用分析探讨协变量的影响,曲线拟合分析阈值效应。结果TWAG 0~6.1 mmol/L组患者的院内病死率、30 d病死率最低(P<0.05)。调整协变量后,TWAG≥6.1 mmol/L各组的院内死亡、30 d死亡风险均较0~6.1 mmol/L组显著升高(P<0.05)。其中TWAG≥10.0 mmol/L组患者的院内死亡风险升高3.99(95%CI 3.03~5.25)倍,30 d死亡风险升高3.11(95%CI 2.41~4.03)倍。亚组分析显示,是否合并糖尿病、是否出现低血糖事件、不同DRG诊断的各亚组患者存在显著的交互效应(P<0.05)。非糖尿病亚组和非低血糖亚组患者的TWAG与短期预后的关系呈J型曲线,将折点设置在6 mmol/L和10 mmol/L时存在阈值效应(P<0.05)。除肾脏疾病、脓毒症&传染病外,TWAG>7.8 mmol/L是ICU常见疾病短期预后的危险因素。结论TWAG与ICU常见疾病患者的短期预后显著相关。是否合并糖尿病、是否出现低血糖事件、不同DRG诊断患者的TWAG与ICU患者短期预后的关系差异较大。
Objective To determine the association between glucose control and short-term outcomes of the patients with common diseases in intensive care unit(ICU).Methods This was a retrospective cohort study,the data source was the public database MIMIC-Ⅲ.9794 adult patients admitted to ICU with heart disease,renal disease,respiratory disease,craniocerebral disease,sepsis and disseminated infections by diagnosis related proups(DRG)for the first time were selected as the research objects.The primary outcomes were in-hospital mortality and 30-day mortality.Time-weighted average glucose(TWAG)was calculated,and all the patients were divided into 4 groups accordingly:0-6.1 mmol/L,6.1-7.8 mmol/L,7.8-10.0 mmol/L,≥10.0 mmol/L.Clinical characteristics and outcome incidence were compared among different TWAG groups.Logistic regression models predicting the outcomes were established,subgroup analysis and interaction analysis were also performed.Threshold analysis was conducted based on the curve fit model.Results The lowest in-hospital mortality and 30-day mortality appeared in the TWAG 0-6.1 mmol/L group(P<0.05).After adjusting covariates,TWAG≥6.1 mmol/L groups showed significantly increased in-hospital mortality and 30-day mortality compared with the group of TWAG 0-6.1 mmol/L(P<0.05).Impressively,the patients of TWAG≥10.0 mmol/L had a dramatically increased risk of 3.99(95%CI 3.03-5.25)and 3.11(95%CI 2.41-4.03)for in-hospital mortality and 30-day mortality,respectively.In subgroup analysis,whether combined with diabetes or hypoglycemia,and different DRG diagnosis were identified as the factors with significant interaction effects(P<0.05).There was a J-shape relationship between TWAG and short-term outcomes of the patients with and without diabetes,threshold effects were shown when the inflection points of TWAG were set at 6 mmol/L and 10 mmol/L(P<0.05).Except for renal disease,sepsis and disseminated infections,TWAG>7.8 mmol/L seemed a risk factor for short-term outcomes in ICU patients with common diseases.Conclusions There is significant relationship between TWAG and short-term outcomes of ICU patients with common diseases.There are significant differences in the relationship between TWAG of different subgroups of diabetes,hypoglycemia and DRG diagnosis and short-term outcomes in ICU patients.
作者
田伊茗
王翠娟
张萌萌
王锐
Tian Yi-ming;Wang Cui-juan;Zhang Meng-meng;Wang Rui(Department of Endocrinology,the First Hospital of Qinhuangdao,Qinhuangdao 066000,China)
出处
《中国急救医学》
CAS
CSCD
2022年第3期227-233,共7页
Chinese Journal of Critical Care Medicine
关键词
重症
血糖控制
预后
真实世界研究
Intensive care unit
Glucose control
Prognosis
Real-world study