Engle V F, Graney M J
College of Nursing, University of Tennessee, Memphis.
J Gerontol. 1993 Sep;48(5):S269-75. doi: 10.1093/geronj/48.5.s269.
The purpose of this study was to document prevalence, identify leading indicators, and develop and assess prediction models for nursing home resident outcomes (death, discharge home). New residents (N = 647) in eight nursing homes were evaluated on admission and 3 and 6 months later. Predictor variables tested included 14 health status variables, 5 demographic variables, and 6 other resident and nursing home variables. A cancer diagnosis was the most important risk factor for death at 3 months (OR = 8.0) and 6 months (OR = 10.8). A first nursing home admission was the most important risk factor for discharge to home at 3 months (OR = 9.5) and 6 months (OR = 13.9). Not having a dementia diagnosis was the second most important risk factor for discharge to home at 3 months (OR = 5.8) and 6 months (OR = 4.2). All prediction models demonstrated useful potential for clinical application.
本研究的目的是记录患病率，确定主要指标，并开发和评估针对养老院居民结局（死亡、出院回家）的预测模型。对八所养老院的新入住居民（N = 647）在入院时、3个月后和6个月后进行了评估。测试的预测变量包括14个健康状况变量、5个人口统计学变量以及6个其他居民和养老院变量。癌症诊断是3个月时（OR = 8.0）和6个月时（OR = 10.8）死亡的最重要风险因素。首次入住养老院是3个月时（OR = 9.5）和6个月时（OR = 13.9）出院回家的最重要风险因素。未被诊断为痴呆是3个月时（OR = 5.8）和6个月时（OR = 4.2）出院回家的第二重要风险因素。所有预测模型均显示出在临床应用中的有用潜力。
