Artificial Intelligence for Older Adult Health: Opportunities for Advancing Gerontological Nursing Practice

Full text

Introduction

Artificial intelligence (AI) is an emerging technological trend that encompasses a range of advanced computational techniques, mainly machine learning and natural language processing (NLP). These techniques can be applied to health care datasets in many ways to try to improve the prediction of patient, health service, and other outcomes, which can be used to inform clinical decision making and care delivery (O'Connor et al., 2022). These predictive algorithms have many applications in the physical and virtual world, such as clinical decision support systems, robotics, remote monitoring systems, mobile applications, wearable devices, virtual reality, and gaming technologies. Hence, AI is starting to be used to support the care of older adults, which presents a new opportunity for gerontological nursing practice.

Applications of AI in Older Adult Health

Due to biological changes that accompany aging, physical health conditions that older adults may experience include hearing loss, poor vision, joint and muscle pain, diabetes, and dementia, among others. Older adults are also more prone to developing frailty and experiencing incontinence, falls, delirium, and pressure ulcers—areas of older adult health where AI techniques could possibly help improve patient and other outcomes (O'Connor et al., in press). For example, a team that included a researcher from Penn State College of Nursing developed a machine learning system to identify additional preferences for everyday living of nursing home residents and proposed this “recommender system” could enhance delivery of person-centered care (Gannod et al., 2019). A group of nurses at Columbia University School of Nursing used NLP to identify patients in critical care who lacked surrogates and advanced directives (Song et al., 2022), a strategy that could be applied in gerontological nursing in a range of hospital and community settings to support older adults and their families. Robots often employ AI techniques in their internal software systems, particularly a branch of machine learning called reinforcement learning, to enable them to interact with and adapt to the world around them. Robotics is another area in which gerontological nurses can and are breaking new ground, as robopets are being deployed to provide comfort and psychological support to older adults living in care homes to help address the loneliness and depression they sometimes experience (Abbott et al., 2019).

Although many research studies examining how AI can improve older adult health have not included gerontological nurses, the approach used could be adopted by the profession. For instance, Bayen et al. (2021) developed and tested an AI–based video monitoring system among older adults with Alzheimer's disease and found it helped reduce the time they spent on the ground after a fall, as care staff were notified in real-time and able to respond quickly. This type of AI–based system could support gerontological nurses in their daily practice to help decrease secondary complications from falls and improve the prognosis of older adults' post-fall, while reducing health care costs. In another study, Al-Hameed et al. (2019) tested a proof-of-concept AI–based speech recognition system with older adults at home at risk of developing dementia to determine if linguistic changes indicative of the early stages of this neurodegenerative syndrome could be identified. These novel approaches to preventive care could enhance the care and support nurses provide for older adults and their families.

Limitations and Risks of AI

Like any technology, the various computational approaches that comprise AI have limitations and may introduce some risks. Algorithmic bias is one potential risk, as digital health datasets used to train and test AI algorithms can be missing information from certain populations. Poor quality datasets used to develop AI could skew the predictive models and lead to inappropriate clinical decision making, which could negatively impact older adult care and reinforce existing inequalities in health care (Chu et al., 2022). The retrospective nature of many health datasets that AI techniques are developed on may also reduce the ability to forecast future events, as probabilistic models might be missing key variables that could impact older adult health as seen during the coronavirus disease 2019 pandemic (Chin et al., 2020). Hence, gerontological nurses need to be aware of the limitations of AI–based systems and continue to use their clinical expertise to support the care of older adults.

In addition, Stokes and Palmer (2020) highlight a number of ethical issues when AI is integrated into robotics, as there is some concern that robotic technologies may replace gerontological nurses or automate aspects of their caring roles. As robots lack human emotions, such as empathy, this could lead to less personalized care and poorer therapeutic relationships with older adults, which may compromise their physical, mental, and social health. Personal privacy is another worry when AI is introduced in robotic and home/remote monitoring technologies, as it could lead to increased surveillance and possible inappropriate disclosure and use of personal information (Hasal et al., 2021), which may impact older adults' autonomy and well-being. Interestingly, Galambos et al. (2019) evaluated perceptions of older adults and their families about the use of intelligent sensors and found they appreciated there may be a trade-off between the benefits of assisted living versus personal privacy. This finding suggests that as people age, the advantages and disadvantages of using AI–based technologies could be something older adults and their carers consider.

Conclusion

Given the rapid pace of AI in older adult care, gerontological nurses need to be more aware of and knowledgeable about this technological trend as it will impact their practice and so they can provide guidance to patients about using AI–based tools. Therefore, nurses need more educational opportunities to learn about the range of AI techniques that are available and how to apply them to older adult datasets (Ronquillo et al., 2021). This knowledge will enable nurses to conduct AI research and assess if these predictive algorithms can benefit clinical decision making and patient care, and if it is worthwhile to introduce AI–based technologies into gerontological nursing practice. Nurses could collaborate with colleagues in computer science, engineering, and the technology industry, while encouraging active participation of patients to help co-design AI–based tools to meet the needs of older adults (Blakey et al., 2020). As the digital age accelerates, the nursing profession, particularly those working in gerontology, should embrace AI to help determine if it can support the health and well-being of older adults.

Siobhán O'Connor, PhD, RGN, BSc

Senior Lecturer

Division of Nursing, Midwifery and

Social Work

University of Manchester

Manchester, United Kingdom

siobhan.oconnor@manchester.ac.uk

全文翻译（仅供参考）

导言

Artificial intelligence (AI)是一种新兴的技术趋势，涵盖了一系列先进的计算技术，主要是机器学习和自然语言处理（NLP）。这些技术可以多种方式应用于医疗保健数据集，以尝试改善对患者、医疗服务和其他结果的预测，这些预测可用于为临床决策和护理提供信息（奥康纳等人，2022年）。这些预测算法在物理和虚拟世界中有许多应用，如临床决策支持系统、机器人、远程监控系统、移动的应用、可穿戴设备、虚拟现实和游戏技术。因此，人工智能开始用于支持老年人的护理，这为老年护理实践带来了新的机遇。

人工智能在老年健康中的应用

由于衰老带来的生物学变化，老年人可能会经历的身体健康状况包括听力丧失、视力下降、关节和肌肉疼痛、糖尿病和痴呆等。老年人也更容易出现虚弱、失禁、跌倒、谵妄和压疮--在老年人健康领域，人工智能技术可能有助于改善患者和其他方面的结果（O'Connor等人，出版中）。例如，一个包括宾夕法尼亚州立大学护理学院研究人员在内的团队开发了一个机器学习系统，以识别养老院居民对日常生活的额外偏好，并提出这种"推荐系统"可以增强以人为本的护理服务（Gannod等人，2019年）。哥伦比亚大学护理学院的一组护士使用NLP来识别缺乏替代者和预先指示的重症监护病人（Song等人，2022），这项策略可应用于医院和社区的老人护理工作，以支援老人及其家人。机器人通常会在其内部软件系统中采用人工智能技术，特别是机器学习的一个分支，称为reinforcement learning，让他们能与周围的世界互动，并适应环境。机器人技术是另一个老人科护士可以而且正在开拓的新领域，因为机器人被用来为安老院舍的长者提供舒适和心理支持，以帮助他们解决有时会感到的孤独和抑郁（Abbott等人，2019年）.

虽然许多研究人工智能如何改善老年人健康的研究并没有包括老年科护士，但所采用的方法可为专业人士所采用。例如Bayen et al.（2021）开发了一种基于人工智能的视频监控系统，并在患有阿尔茨海默氏症的老年人中进行了测试，发现该系统有助于减少他们跌倒后在地上的时间。护理人员可即时得到通知，并能迅速作出反应。这类以人工智能为基础的系统可支援老人科护士的日常工作，以协助减少跌倒后的继发性并发症，并改善老人跌倒后的预后。在另一项研究中，Al-Hameed等人（2019）测试了一个概念验证的人工智能语音识别系统，测试对象是家中有痴呆风险的老年人，以确定是否可以识别出这种神经退行性综合征早期阶段的语言变化。这些预防性护理的新方法可以增强护士为老年人及其家人提供的护理和支持。

人工智能的局限性和风险

与任何技术一样，构成AI的各种计算方法都有局限性，可能会引入一些风险。算法偏倚是一个潜在风险，因为用于训练和测试AI算法的数字健康数据集可能会遗漏某些人群的信息。用于开发AI的低质量数据集可能会扭曲预测模型，并导致不适当的临床决策。这可能会对老年人护理产生负面影响，并加剧现有的卫生保健不平等（Chu等人，2022年）。人工智能技术开发所基于的许多健康数据集的回顾性也可能降低预测未来事件的能力，因为概率模型可能会遗漏可能影响老年人健康的关键变量，就像2019年冠状病毒病大流行期间所看到的那样（Chin等人，2020年）。因此，老年科护士需要意识到基于人工智能的系统的局限性，并继续使用他们的临床专业知识来支持老年人的护理。

此外，斯托克斯和帕尔默（2020）强调了人工智能融入机器人技术时存在的诸多伦理问题，因为有人担心机器人技术可能会取代老年科护士或将其护理角色的某些方面自动化，由于机器人缺乏人类情感，如同理心，这可能会导致个性化护理减少，与老年人的治疗关系变差，可能会损害他们的身体、精神、和社会健康。当人工智能引入机器人和家庭/远程监控技术时，个人隐私是另一个令人担忧的问题，因为它可能导致监控增加，并可能导致个人信息的不当披露和使用（Hasal等人，2021年），这可能会影响老年人的自主性和幸福感。有趣的是，Galambos等（2019）评估了老年人及其家人对使用智能传感器的看法，发现他们意识到辅助生活的好处与个人隐私之间可能存在权衡。这一发现表明，随着人们年龄的增长，使用基于人工智能的技术的利弊可能是老年人及其照顾者需要考虑的问题。

结语

鉴于人工智能在老年人护理领域的发展速度很快，老年科护士需要更多地了解这一技术趋势，因为这将影响他们的实践，因此他们可以为患者提供使用人工智能工具的指导。因此，护士需要更多的教育机会，以了解可用的人工智能技术的范围，以及如何将其应用于老年人数据集（Ronquillo等人，2021年）。这些知识将使护士能够进行人工智能研究，并评估这些预测算法是否有利于临床决策和患者护理，以及是否值得将基于人工智能的技术引入老年护理实践。护士可以与计算机科学，工程和技术行业的同事合作，同时鼓励病人积极参与，协助共同设计以人工智能为基础的工具，以满足老年人的需要（Blakey等人，2020年）。随着数字时代的加速发展，护理行业，尤其是从事老年医学工作的护理行业，应该采用人工智能来帮助确定人工智能是否能够支持老年人的健康和福祉。

THE

END

百度浏览   来源 : NursingResearch护理研究前沿   

版权声明：本网站所有注明来源“医微客”的文字、图片和音视频资料，版权均属于医微客所有，非经授权，任何媒体、网站或个人不得转载，授权转载时须注明来源：”医微客”。本网所有转载文章系出于传递更多信息之目的，且明确注明来源和作者，转载仅作观点分享，版权归原作者所有。不希望被转载的媒体或个人可与我们联系，我们将立即进行删除处理。 本站拥有对此声明的最终解释权。