社区获得性肺炎患者入院血糖水平与预后的相关性：一项多中心、回顾性研究

The impact of blood glucose level on the clinical outcomes of patients hospitalized with community-acquired pneumonia: A multicenter, retrospective study

目的探讨社区获得性肺炎(CAP)患者入院血糖水平对其临床结局的影响。方法回顾性分析2014年1月-2015年12月,北京积水潭医院、北京朝阳医院、青岛市立医院以及云南省第二人民医院出院诊断中包含“肺炎”或“肺部感染”患者的病例资料,按照纳入及排除标准选取获得性肺炎(CAP)患者作为研究对象。收集研究对象的人口学资料、临床和影像学特征以及临床结局。依据患者既往是否合并糖尿病,将患者分为合并糖尿病组(DM-CAP组)与不合并糖尿病组(非DM-CAP组)。采用单因素回归和Cox多因素回归模型分析CAP患者入院血糖与ICU住院率及30 d病死率的关系。结果共纳入3366例CAP患者,其中DM-CAP患者占15.8%(531/3366),非DM-CAP患者占84.2%(2835/3366),经性别、年龄、基础病和CURB-65评分调整的Cox多因素回归结果显示:相比于非DM-CAP组患者,DM-CAP组患者入住ICU和入院后30 d死亡的HR(95%CI)分别为0.991(0.644~1.525)(P>0.05)和1.109(1.084~2.796)(P<0.05)。在非DM-CAP组,相比于正常血糖水平(4.0~5.9 mmol/L)患者,血糖为6.0~13.9 mmol/L时,ICU住院风险HR(95%CI)为1.804(1.036~2.862)(P<0.05),30 d死亡风险HR(95%CI)为2.539(1.483~4.348)(P<0.01);血糖为14.0~19.9 mmol/L时,ICU住院风险HR(95%CI)为4.542(2.302~8.959)(P<0.001),30 d死亡风险HR(95%CI)为3.172(1.113~4.265)(P<0.001)。在DM-CAP组,仅血糖≥20.0 mmol/L时,30 d死亡风险较正常血糖水平患者增高,其HR(95%CI)为5.516(1.077~28.250)(P<0.05);未发现血糖水平与ICU住院风险相关。结论糖尿病可增高CAP患者30 d病死率。入院血糖水平对DM-CAP与非DM-CAP患者预后的影响不同。

Objective To identify the blood glucose level at admission on the clinical outcomes of patients hospitalized with community-acquired pneumonia(CAP). Methods This was a multicenter retrospective study. Data on patients with CAP in four hospitals from Beijing, Shandong and Yunnan Province during January 1, 2014 through December 31, 2015 were reviewed, including demographic, clinical and radiologic characteristics and outcomes. Cox regression model was performed to evaluate the impact of the blood glucose level on ICU admission rate and 30-day mortality of CAP patients. Results Overall, 3366 CAP patients were entered into the final analysis. Diabetic patients(DM-CAP) accounted for 15.8%(531/3366). A Cox multivariable regression model(adjusted for sex, age, comorbidities and CURB-65 score) confirmed diabetes mellitus was associated with increased risk for 30-day mortality of CAP patients(HR=1.109, 95% CI 1.084-2.796, P<0.05). In non-DM-CAP patients, compared to patients with normal blood glucose level [(4.0-5.9) mmol/L], patients with blood glucose of(6.0-13.9) mmol/L had increased risk for ICU admission(HR=1.804, 95% CI 1.036-2.862, P<0.05) and 30-day mortality(HR=2.539, 95% CI 1.483-4.348, P<0.01);in those with blood glucose of(6.0-13.9) mmol/L, the risk for ICU admission increased to HR=4.542(95% CI 2.302-8.959, P<0.001), and the risk for 30-day mortality was HR=3.172(95% CI 1.113-4.265, P<0.001). In DM-CAP patients, only blood glucose level ≥20.0 mmol/L increased the risk for 30-day mortality(HR=5.516, 95% CI 1.077-28.250, P<0.05);blood glucose level has no impact on ICU admission. Conclusions Hyperglycaemia can increase the 30-day mortality of CAP patients. The effect of blood glucose level at admission on prognosis is different in DM-CAP and non DM-CAP patients.