Liu Longjian, Sullivan Dennis H
Geriatric Research, Education and Clinical Center, Central Arkansas Veterans Healthcare System, Little Rock, Arkansas 72205, USA.
J Epidemiol. 2003 Jul;13(4):203-10. doi: 10.2188/jea.13.203.
Little is known about the impact of social demographic factors on post-discharge mortality among the hospitalized elderly.
A one-year prospective study was conducted in a random sample of 646 male patients aged 65 or older who were discharged from a university affiliated Veterans Administration Hospital at Little Rock, AR, USA. Within 48 hours of admission, each subject completed a standardized diagnostic evaluation. Mortality was recorded for all causes. Associations between patient characteristics at hospital discharge and mortality were identified utilizing univariable and multivariable (Cox proportional hazard regression) statistical techniques.
The mean (SD) age was 73(+/- 6) years. Within one year of hospital discharge, 83 patients (13%) died. Multiple social demographic factors were significantly associated with mortality in univariable analysis. After controlling for age, Katz index of ADL score, Charlson co-morbidity index and length of hospitalization, risk of one-year post-discharge mortality remained significantly associated with marital status, race, education, and occupational class. When all of the social demographic factors were included in a stepwise procedure, marital status, education and occupational class were selected as the strongest predictors of mortality. The adjusted hazard ratios (95% confidence interval) of mortality associated with non-married status, education < 6 years, and history of having a blue-collar occupation were 2.01 (1.29-3.15), 1.86 (1.05-3.32), and 2.16 (1.03-4.54) respectively.
The results suggest that social demographic factors are independent determinants of mortality among elderly patients. These factors should be used as important indices in identifying patients at higher risk of death in clinical assessments and in prevention programs for elderly patients after hospital discharge.
关于社会人口统计学因素对住院老年人出院后死亡率的影响，人们了解甚少。
对646名年龄在65岁及以上的男性患者进行了为期一年的前瞻性研究，这些患者来自美国阿肯色州小石城一家大学附属退伍军人管理局医院。入院后48小时内，每位受试者完成了标准化诊断评估。记录所有原因导致的死亡率。利用单变量和多变量（Cox比例风险回归）统计技术确定出院时患者特征与死亡率之间的关联。
平均（标准差）年龄为73（±6）岁。出院后一年内，83名患者（13%）死亡。在单变量分析中，多个社会人口统计学因素与死亡率显著相关。在控制了年龄、Katz日常生活活动能力评分指数、Charlson合并症指数和住院时间后，出院后一年的死亡风险仍与婚姻状况、种族、教育程度和职业阶层显著相关。当所有社会人口统计学因素按逐步程序纳入时，婚姻状况、教育程度和职业阶层被选为死亡率的最强预测因素。与未婚状态、教育年限<6年和蓝领职业史相关的死亡率调整后风险比（95%置信区间）分别为2.01（1.29 - 3.15）、1.86（1.05 - 3.32）和2.16（1.03 - 4.54）。
结果表明，社会人口统计学因素是老年患者死亡率的独立决定因素。在临床评估和老年患者出院后预防计划中，这些因素应用作识别死亡风险较高患者的重要指标。
