Bentzen Hege, Bergland Astrid, Forsén Lisa
Norwegian Institute of Public Health, Oslo, Norway.
Oslo University College, Oslo, Norway.
Eur J Ageing. 2008 Aug 22;5(3):253-263. doi: 10.1007/s10433-008-0088-2. eCollection 2008 Sep.
The aim of the present study was to identify predictors for initial uptake and adherence with the use of hip protectors when offering hip protectors free of charge to nursing-home residents. An 18 months prospective follow up study was carried out in 18 Norwegian nursing homes. One thousand two hundred and thirty-six residents were included in the study of which 604 started to use a hip protector. A multivariate logistic regression model was used to identify predictors for the initial uptake. A Cox proportional hazard model was used to identify predictors for adherence. A stepwise backward strategy was used in both the logistic and in the Cox regression. The effect of nursing homes as clusters was adjusted for in the analysis. The uptake rate among all residents was 46% and the adherence was approximately 75% after 3 months, and approximately 60% after 18 months. Female gender [odds ratio (OR): 1.54, 95% CI: 1.06-2.24, = 0.022], previous fractures (OR: 1.67, 95% CI: 1.02-2.75, = 0.043), previous falls (OR: 2.08, 95% CI: 1.35-3.19, < 0.001) and memory (not able to memorise: OR: 3.71, 95% CI: 2.09-6.59, < 0.001, large problems with memorising: OR: 2.85, 95% CI: 1.81-4.49, < 0.001, medium problems with memorising: OR: 2.45, 95% CI: 1.39-4.33, = 0.002, some problems with memorising: OR: 1.99, 95% CI: 1.14-3.48, = 0.016) seemed to be important predictors for uptake. Among those who took up the offer male gender (HR: 1.71, 95% CI: 1.00-2.91, = 0.049), memory (not able to memorise: HR: 0.26, 95% CI: 0.14-0.50, < 0.001, large problems with memorising: HR: 0.32, 95% CI: 0.22-0.45, < 0.001, medium problems with memorising: HR: 0.46, 95% CI: 0.30-0.73, < 0.001, some problems with memorising: HR: 0.49, 95% CI: 0.32-0.73, = 0.001) and bowel incontinence (HR: 0.41, 95% CI: 0.25-0.66, < 0.001) were predictors for a lower probability of ending hip protector use. Factors related to a high risk of falling were important predictors for both uptake and adherence. The fact that neither memory impairments nor incontinence (bowel) seemed to be barriers to hip protector use is important since these characteristics are common among nursing-home residents and tertiary prevention such as the use of hip protectors is probably the most feasible intervention to prevent hip fractures in this group.
本研究的目的是确定在向养老院居民免费提供髋部保护器时，初始使用和坚持使用髋部保护器的预测因素。在挪威的18家养老院进行了一项为期18个月的前瞻性随访研究。1236名居民被纳入研究，其中604人开始使用髋部保护器。使用多变量逻辑回归模型确定初始使用的预测因素。使用Cox比例风险模型确定坚持使用的预测因素。在逻辑回归和Cox回归中均采用逐步向后策略。在分析中对养老院作为聚类的效应进行了调整。所有居民的使用率为46%，3个月后的坚持率约为75%，18个月后约为60%。女性[比值比（OR）：1.54，95%置信区间：1.06 - 2.24，P = 0.022]、既往骨折（OR：1.67，95%置信区间：1.02 - 2.75，P = 0.043）、既往跌倒（OR：2.08，95%置信区间：1.35 - 3.19，P < 0.001）和记忆力（无法记忆：OR：3.71，95%置信区间：2.09 - 6.59，P < 0.001，记忆有大问题：OR：2.85，95%置信区间：1.81 - 4.49，P < 0.001，记忆有中等问题：OR：2.45，95%置信区间：1.39 - 4.33，P = 0.002，记忆有一些问题：OR：1.99，95%置信区间：1.14 - 3.48，P = 0.016）似乎是使用的重要预测因素。在接受提供的人群中，男性（风险比：1.71，95%置信区间：1.00 - 2.91，P = 0.049）、记忆力（无法记忆：风险比：0.26，95%置信区间：0.14 - 0.50，P < 0.001，记忆有大问题：风险比：0.32，95%置信区间：0.22 - 0.45，P < 0.001，记忆有中等问题：风险比：0.46，95%置信区间：0.30 - 0.73，P < 0.001，记忆有一些问题：风险比：0.49，95%置信区间：0.32 - (此处原文有误，应为0.75)，P = 0.001）和大便失禁（风险比：0.41，95%置信区间：0.25 - 0.66，P < 0.001）是停止使用髋部保护器可能性较低的预测因素。与高跌倒风险相关的因素是使用和坚持使用的重要预测因素。记忆力受损和失禁（大便）似乎都不是使用髋部保护器的障碍，这一事实很重要，因为这些特征在养老院居民中很常见，而三级预防如使用髋部保护器可能是预防该群体髋部骨折最可行的干预措施。
