EXAMINING DISCREPANCIES IN SOCIAL ROBOT VERSUS HUMAN ASSESSMENTS OF GERIATRIC WELL-BEING
EXAMINING DISCREPANCIES IN SOCIAL ROBOT VERSUS HUMAN ASSESSMENTS OF GERIATRIC WELL-BEING
复制标题
DOI：
10.1093/geroni/igz038.1199
复制
发表时间：
2019-11-08
影响因子：
7
通讯作者：
Sheng AW
中科院分区：
文献类型：
--
作者：
Harrington E;Do H;Bishop AJ;Reese-Melancon C;Sheng AW
社会辅助机器人技术是监测老年人安全和健康的可行工具。然而，目前尚不清楚SAR是否能像训练有素的人类临床医生一样有效地全面筛查老年人的健康状况。本研究的目的是比较SAR与人类评估老年人的福祉。参与者包括30名老年人（法师= 73.40，SD = 7.88），他们完成了机器人管理的健康评估会议期间，同时进行人类管理的评估。使用了老年护理中常见的标准化临床筛选评估工具（例如，短祝福测试（SBT）、UCLA孤独量表、老年抑郁量表、PHQ-4、爱荷华州疲劳量表、跌倒风险）。使用多个相关样本t检验探索SAR和人体评价之间评估评分的变异性。评估分数在几个指标上存在显著差异，包括SBT（t（29）=-9.33，p <0.001），UCLA Longland量表（t（19）= 2.37，p <0.001）。05）和跌倒风险评估（t（29）= 3.03，p <0.01）。具体来说，SAR表明，与人类管理员相比，老年人的认知受损程度更高，更不孤独，更容易摔倒。其他观察到的差异和假设的解释将进行更详细的讨论。目前的研究表明，有一个分歧的老年评估结果的基础上，人类与SAR管理。研究结果对进一步开发SAR技术以与基于人类的评估保持一致以增强认知健康、社会联系和福尔斯预防具有影响。
Socially assistive robotic (SAR) technologies represent a viable tool for monitoring the safety and health of older adults. However, it is unclear whether SARs can comprehensively screen geriatric well-being as effectively as trained human clinicians. The purpose of this study was to compare SAR versus human assessment of geriatric well-being. Participants included 30 older adults (Mage = 73.40, SD = 7.88) who completed a robot-administered well-being assessment session during which human-administered evaluation was simultaneously performed. Standardized clinical screening assessment tools common in geriatric care were administered (e.g., Short Blessed Test (SBT), UCLA Loneliness Scale, Geriatric Depression Scale, PHQ-4, Iowa Fatigue Scale, Fall Risk). Multiple dependent sample t-tests were used to explore variability in assessment scores between SAR and human evaluation. Assessment scores significantly differed on several measures, including the SBT (t(29) = -9.33, p < .001), UCLA Loneliness scale (t(19) = 2.37, p < . 05), and fall risk assessment (t(29) = 3.03, p < .01). Specifically, the SAR indicated that older adults were significantly more cognitively impaired, less lonely, and more likely to fall compared to the human administrator. Other observed differences and hypothesized explanations will be discussed in greater detail. The current study indicates that there is a divergence in geriatric assessment outcomes based on human versus SAR administration. Findings have implications relative to further developing SAR technology to align with human-based evaluations to enhance cognitive well-being, social connectedness, and falls prevention.