住院老年患者谵妄的发生率及预测因素：一项回顾性队列研究。

Incidence and predictors for delirium in hospitalized elderly patients: a retrospective cohort study.

作者信息

Isfandiaty Ratih, Harimurti Kuntjoro, Setiati Siti, Roosheroe Arya G

机构信息

Department of Internal Medicine, Faculty of Medicine, University of Indonesia-Cipto Mangunkusumo Hospital. Jl. Diponegoro no. 71, Jakarta Pusat 10430, Indonesia.

出版信息

Acta Med Indones. 2012 Oct;44(4):290-7.

PMID:23314969

Abstract

AIM

to determine the incidence and predictors for delirium and to develop a prediction model for delirium in hospitalized elderly patient in Indonesia.

METHODS

a retrospective cohort study was conducted in elderly patients (aged 60 years and older) who were hospitalized in Internal Medicine Ward and Acute Geriatric Ward Cipto Mangunkusumo Hospital from January 2008 until December 2010. Patients were not delirious on admission. Twelve predefined predictors for development of delirium during hospitalization were identified on admission. Independent predictors for delirium were identified by Cox's proportional hazard regression analysis and each independent predictor was quantified to develop delirium prediction model. The calibration performance of the model was tested by Hosmer-Lameshow test and its discrimination ability was determined by calculating area under the receiver operating characteristic curve (AUC).

RESULTS

subjects consist of 457 patients, predominantly male (52.5%) and were in 60-69 age group (55.8%), with mean age of 69.6 (SD 7.09) years old. Delirium developed in 86 patients (cumulative incidence 18.8%, incidence density 0.021 per person-days) during first fourteen-days of hospitalization. Three independent predictors for delirium were identified, including: infection (without sepsis, adjusted HR1.83 (95% CI 0.82-4.10); with sepsis, adjusted HR 4.86, 95% CI 2.14-11.04), cognitive impairment (adjusted HR 3.12; 95%CI 1.89-5.13) and decrease of functional status (adjusted HR 1.74; 95% CI 1.07-2.82). Predictive model was performed using the final model of multivariate analysis and stratified into three levels: low- (rate of delirium 4.4%), intermediate- (32.8%), and high-risk (54.7%) groups.The Hosmer-Lemeshow test revealed good precision (p-value 0,066) and the AUC showed good discrimination ability (0.82, 95% CI 0.78-0.88).

CONCLUSION

incidence of delirium is 18.8% in hospitalized elderly patients, with incidence density of 0.021 per person days. Infections, cognitive impairment, and decrease of functional status on admission are independent predictors for the development of delirium during hospitalization.

摘要

目的

确定印度尼西亚住院老年患者谵妄的发病率及预测因素，并建立谵妄预测模型。

方法

对2008年1月至2010年12月在西爪哇省茂物市西托·曼古库苏莫医院内科病房和急性老年病房住院的60岁及以上老年患者进行回顾性队列研究。患者入院时无谵妄。入院时确定了12个预先定义的住院期间谵妄发生的预测因素。通过Cox比例风险回归分析确定谵妄的独立预测因素，并对每个独立预测因素进行量化以建立谵妄预测模型。通过Hosmer-Lameshow检验测试模型的校准性能，并通过计算受试者工作特征曲线下面积（AUC）确定其辨别能力。

结果

研究对象包括457例患者，以男性为主（52.5%），年龄在60-69岁组（55.8%），平均年龄69.6（标准差7.09）岁。86例患者（累积发病率18.8%，发病密度0.021人·日）在住院的前14天内发生谵妄。确定了谵妄的三个独立预测因素，包括：感染（无脓毒症，调整后HR 1.83（95%CI 0.82-4.10）；有脓毒症，调整后HR 4.86，95%CI 2.14-11.04）、认知障碍（调整后HR 3.12；95%CI 1.89-5.13）和功能状态下降（调整后HR 1.74；95%CI 1.07-2.82）。使用多变量分析的最终模型进行预测模型分析，并分为三个级别：低风险（谵妄发生率4.4%）、中风险（32.8%）和高风险（54.7%）组。Hosmer-Lemeshow检验显示精度良好（p值0.066），AUC显示辨别能力良好（0.82，95%CI 0.78-0.88）。