@文章{信息:doi/10.2196/19732，作者=“Kim, Ben和McKay, Sandra M和Lee, Joon”，标题=“消费级可穿戴设备预测加拿大家庭护理服务客户的脆弱性:前瞻性观察性概念验证研究”，期刊=“J Med Internet Res”，年=“2020”，月=“9”，日=“3”，卷=“22”，数=“9”，页=“e19732”，关键词=“脆弱性”;移动健康;衣物;身体活动;家庭护理;预测;预测建模，老年人;背景:虚弱对家庭护理老年人的健康有不利影响，并与住院和长期护理住院率的增加有关。家庭护理客户中虚弱的患病率了解甚少，范围从4.0%到59.1%。虽然存在虚弱筛查工具，但在实践中使用不一致，需要更多创新和更易于使用的工具。 Owing to increases in the capacity of wearable devices, as well as in technology literacy and adoption in Canadian older adults, wearable devices are emerging as a viable tool to assess frailty in this population. Objective: The objective of this study was to prove that using a wearable device for assessing frailty in older home care clients could be possible. Methods: From June 2018 to September 2019, we recruited home care clients aged 55 years and older to be monitored over a minimum of 8 days using a wearable device. Detailed sociodemographic information and patient assessments including degree of comorbidity and activities of daily living were collected. Frailty was measured using the Fried Frailty Index. Data collected from the wearable device were used to derive variables including daily step count, total sleep time, deep sleep time, light sleep time, awake time, sleep quality, heart rate, and heart rate standard deviation. Using both wearable and conventional assessment data, multiple logistic regression models were fitted via a sequential stepwise feature selection to predict frailty. Results: A total of 37 older home care clients completed the study. The mean age was 82.27 (SD 10.84) years, and 76{\%} (28/37) were female; 13 participants were frail, significantly older (P<.01), utilized more home care service (P=.01), walked less (P=.04), slept longer (P=.01), and had longer deep sleep time (P<.01). Total sleep time (r=0.41, P=.01) and deep sleep time (r=0.53, P<.01) were moderately correlated with frailty. The logistic regression model fitted with deep sleep time, step count, age, and education level yielded the best predictive performance with an area under the receiver operating characteristics curve value of 0.90 (Hosmer-Lemeshow P=.88). Conclusions: We proved that a wearable device could be used to assess frailty for older home care clients. Wearable data complemented the existing assessments and enhanced predictive power. Wearable technology can be used to identify vulnerable older adults who may benefit from additional home care services. ", issn="1438-8871", doi="10.2196/19732", url="//www.mybigtv.com/2020/9/e19732", url="https://doi.org/10.2196/19732", url="http://www.ncbi.nlm.nih.gov/pubmed/32880582" }