Kron M, Loy S, Sturm E, Nikolaus Th, Becker C
Department of Biometry and Medical Documentation, Universität Ulm, Ulm, Germany.
Am J Epidemiol. 2003 Oct 1;158(7):645-53. doi: 10.1093/aje/kwg203.
The aim of this study was to identify individual predisposing risk indicators for falls in a sample of institutionalized frail elderly in southern Germany. The design was a prospective observational study with a 1-year follow-up (October 1998-September 1999). The study population included 472 long-term-care residents whose mean age was 84 years; 77% were female. Risk indicators for accidental falls were analyzed by using logistic regression. Residents were found to have an incidence density rate of falls of 2,558 per 1,000 resident-years. Multiple logistic regression analysis revealed short-term memory loss, transfer assistance, urinary incontinence, positive fall history, and use of trunk restraints as predictors of falls. In a further logistic regression analysis, depressive symptoms, transfer assistance, urinary incontinence, and positive fall history were associated with frequent falls. Using these risk indicators as a screening procedure to identify fallers would be easy to administer and could be accomplished by nursing staff. Study results encourage specifically addressing urinary incontinence, cognitive impairment, use of restraints, depression, and transfer difficulties as modifiable predisposing risk factors for falls. Fall history represents an important nonmodifiable marker to identify residents at high risk.
本研究的目的是在德国南部机构养老的体弱老年人样本中确定个体跌倒的易患风险指标。研究设计为前瞻性观察性研究，随访1年（1998年10月至1999年9月）。研究人群包括472名长期护理居民，平均年龄84岁；77%为女性。采用逻辑回归分析意外跌倒的风险指标。居民的跌倒发生率为每1000居民年2558次。多元逻辑回归分析显示，短期记忆丧失、转移协助、尿失禁、跌倒史阳性和使用躯干约束是跌倒的预测因素。在进一步的逻辑回归分析中，抑郁症状、转移协助、尿失禁和跌倒史阳性与频繁跌倒有关。使用这些风险指标作为筛查程序来识别跌倒者易于实施，护理人员即可完成。研究结果鼓励特别关注尿失禁、认知障碍、约束使用、抑郁和转移困难等可改变的跌倒易患风险因素。跌倒史是识别高危居民的重要不可改变指标。
