基于危险因素分层的重症患者谵妄发生预测模型建立及其实用性研究

发布时间：2018-04-29 16:41

本文选题：谵妄 + 预测模型　；参考：《兰州大学》2017年硕士论文

【摘要】：目的:通过建立一个基于危险因素分层的模型,来预测成年重症患者发生谵妄的几率,以便于对高危患者进行早期干预,从而降低谵妄的发生率。方法:前瞻性、观察性、单中心研究。选取年龄、APACHE II评分、昏迷、急诊手术、机械通气、多发伤、代谢性酸中毒、高血压病史、谵妄病史、痴呆病史以及右美托咪定注射液11个相关因素,使用CAM-ICU评估量变进行谵妄评估。通过多元Logistic回归建立谵妄预测模型,AUROC曲线评估模型的预测能力。结果:2016年5月17日到2016年9月25日,共筛查681名连续入住患者,其中61名因不符合相关纳入标准而排除,排除最多的原因是:30例ICU住院期间持续昏迷;其次,18例入ICU时间小于24小时,最后,13例入ICU前已出现谵妄症状,最终研究纳入620例病人,其中包括162例神经系统疾病患者,160例患者发生谵妄,发生谵妄的患者中神经系统疾病患者有64例。入选患者平均年龄55±18岁,49.2%为男性,平均APACHE-II评分16±4分。对神经系统疾病和谵妄发生行Spearman相关性分析显示:相关系数为0.186(P0.01)。AUROC为0.78(95%CI 0.72 to 0.83)结论:通过11个相关因素建立了重症患者谵妄发生预测模型,并进一步明确了右美托咪定注射液对于重症患者谵妄发生的预防作用。同时,神经系统疾病患者发生谵妄风险高,应使用相应的预防措施。
[Abstract]:Objective: to establish a risk factor stratified model to predict the incidence of delirium in adult severe patients in order to reduce the incidence of delirium. Methods: prospective, observational, monocentric study. 11 related factors including age Apache II score, coma, emergency surgery, mechanical ventilation, multiple injuries, metabolic acidosis, hypertension, delirium, dementia and dexmetomidine injection were selected. Delirium was evaluated with CAM-ICU. Multivariate Logistic regression was used to establish the prediction model of delirium and the predictive ability of the model was evaluated by AUROC curve. Results: from May 17, 2016 to September 25, 2016, a total of 681 consecutive admitted patients were screened, of which 61 were excluded because they did not meet the relevant inclusion criteria. The most common reason for exclusion was 30 cases of ICU persistent coma while in hospital. Then 18 cases were admitted to ICU for less than 24 hours, and finally 13 cases developed delirium symptoms before entering ICU. The final study included 620 cases, including 162 cases of nervous system diseases, 160 cases of delirium. Among the patients with delirium, there were 64 patients with neurological diseases. The average age of the patients was 55 卤18 years old and 49.2% were male. The average APACHE-II score was 16 卤4. Spearman correlation analysis of nervous system diseases and delirium showed that the correlation coefficient was 0.78(95%CI 0.72 to 0.83. Conclusion: a predictive model of delirium in severe patients was established by 11 related factors. The preventive effect of dexmetomidine injection on the occurrence of delirium in severe patients was further clarified. At the same time, the risk of delirium in patients with neurological diseases is high.
【学位授予单位】：兰州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R459.7

【相似文献】