早期营养治疗与有创机械通气病人28d病死率的预测因素研究

Early nutrition support and predictors of 28-day mortality of patient with invasive mechanical ventilation in respiratory intensive care unit

目的:研究呼吸重症监护病房(RICU)有创机械通气病人早期营养治疗及其他因素对28 d病死率的影响。方法:将呼吸衰竭行气管插管的64例病人分为存活组和病死组。记录血清清蛋白(ALB)水平、APACHEⅡ评分和并发症情况,营养支持方式(包括完全肠内营养和混合营养),分析影响28 d病死率的因素。结果:混合效应模型分析显示,入住RICU时间是个显著的预测变量(P<0.05),分组不是显著的预测变量(P>0.05)。而入住RICU时间和分组的交互效应是个显著的预测变量(P<0.05)。包含五个变量:第1天实际热量/标准热量、第2天实际热量/标准热量、1周后清蛋白、呼吸机相关性肺炎(VAP)、肾功能衰竭的判别分析显示Wilks's Lambda是显著的(P<0.05),表明该模型能预测第28天病人是否死亡。结论:1早期(48 h内)摄入热量与28 d病死率有关。2五个变量的模型对第28天病死率正确的预测为83.3%,对存活正确的预测为72.7%。

Objective：To observe the relationship of early nutrition feeding with 28-day mortality in critically ill patients in respiratory intensive care.Methods：From January 2011 to December 2012,64 patients due to respiratory failure were admitted to the RICU.The patients were divided into survival group and the death group.The albumin levels,APACHE Ⅱ score at admission,and complications were revorded.The relationship between nutritional support and 28-day mortality was analysed.Results：Mixed-effects model showed that the time in RICU was a significant predictor （P ＜ 0.05）.And the RICU time and packet interaction was a significant predictor （P ＜ 0.05）.The ratios of actual calories to standard calories in the first and second days,albumin after a week,VAP and renal failure showed Wilks＇s Lambda was significant,lambda =0.72,x2 =11.140,P =0.049 ＜ 0.05.Conclusion：1.The calorie intake within 48 hours is associated with 28-day mortality.2.The five variables model can correctlypredict the 28-day mortality with 83.3％,and the 72.7％ of the survival rate can be correctly predicted.