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

全文翻譯（僅供參考）

導言

Artificial intelligence (AI)是一種新興的技術趨勢，涵蓋了一系列先進的計算技術，主要是機器學習和自然語言處理（NLP）。這些技術可以多種方式應用于醫(yī)療保健數(shù)據集，以嘗試改善對患者、醫(yī)療服務和其他結果的預測，這些預測可用于為臨床決策和護理提供信息（奧康納等人，2022年）。這些預測算法在物理和虛擬世界中有許多應用，如臨床決策支持系統(tǒng)、機器人、遠程監(jiān)控系統(tǒng)、移動的應用、可穿戴設備、虛擬現(xiàn)實和游戲技術。因此，人工智能開始用于支持老年人的護理，這為老年護理實踐帶來了新的機遇。

人工智能在老年健康中的應用

由于衰老帶來的生物學變化，老年人可能會經歷的身體健康狀況包括聽力喪失、視力下降、關節(jié)和肌肉疼痛、糖尿病和癡呆等。老年人也更容易出現(xiàn)虛弱、失禁、跌倒、譫妄和壓瘡--在老年人健康領域，人工智能技術可能有助于改善患者和其他方面的結果（O'Connor等人，出版中）。例如，一個包括賓夕法尼亞州立大學護理學院研究人員在內的團隊開發(fā)了一個機器學習系統(tǒng)，以識別養(yǎng)老院居民對日常生活的額外偏好，并提出這種"推薦系統(tǒng)"可以增強以人為本的護理服務（Gannod等人，2019年）。哥倫比亞大學護理學院的一組護士使用NLP來識別缺乏替代者和預先指示的重癥監(jiān)護病人（Song等人，2022），這項策略可應用于醫(yī)院和社區(qū)的老人護理工作，以支援老人及其家人。機器人通常會在其內部軟件系統(tǒng)中采用人工智能技術，特別是機器學習的一個分支，稱為reinforcement learning，讓他們能與周圍的世界互動，并適應環(huán)境。機器人技術是另一個老人科護士可以而且正在開拓的新領域，因為機器人被用來為安老院舍的長者提供舒適和心理支持，以幫助他們解決有時會感到的孤獨和抑郁（Abbott等人，2019年）.

雖然許多研究人工智能如何改善老年人健康的研究并沒有包括老年科護士，但所采用的方法可為專業(yè)人士所采用。例如Bayen et al.（2021）開發(fā)了一種基于人工智能的視頻監(jiān)控系統(tǒng)，并在患有阿爾茨海默氏癥的老年人中進行了測試，發(fā)現(xiàn)該系統(tǒng)有助于減少他們跌倒后在地上的時間。護理人員可即時得到通知，并能迅速作出反應。這類以人工智能為基礎的系統(tǒng)可支援老人科護士的日常工作，以協(xié)助減少跌倒后的繼發(fā)性并發(fā)癥，并改善老人跌倒后的預后。在另一項研究中，Al-Hameed等人（2019）測試了一個概念驗證的人工智能語音識別系統(tǒng)，測試對象是家中有癡呆風險的老年人，以確定是否可以識別出這種神經退行性綜合征早期階段的語言變化。這些預防性護理的新方法可以增強護士為老年人及其家人提供的護理和支持。

人工智能的局限性和風險

與任何技術一樣，構成AI的各種計算方法都有局限性，可能會引入一些風險。算法偏倚是一個潛在風險，因為用于訓練和測試AI算法的數(shù)字健康數(shù)據集可能會遺漏某些人群的信息。用于開發(fā)AI的低質量數(shù)據集可能會扭曲預測模型，并導致不適當?shù)呐R床決策。這可能會對老年人護理產生負面影響，并加劇現(xiàn)有的衛(wèi)生保健不平等（Chu等人，2022年）。人工智能技術開發(fā)所基于的許多健康數(shù)據集的回顧性也可能降低預測未來事件的能力，因為概率模型可能會遺漏可能影響老年人健康的關鍵變量，就像2019年冠狀病毒病大流行期間所看到的那樣（Chin等人，2020年）。因此，老年科護士需要意識到基于人工智能的系統(tǒng)的局限性，并繼續(xù)使用他們的臨床專業(yè)知識來支持老年人的護理。

此外，斯托克斯和帕爾默（2020）強調了人工智能融入機器人技術時存在的諸多倫理問題，因為有人擔心機器人技術可能會取代老年科護士或將其護理角色的某些方面自動化，由于機器人缺乏人類情感，如同理心，這可能會導致個性化護理減少，與老年人的治療關系變差，可能會損害他們的身體、精神、和社會健康。當人工智能引入機器人和家庭/遠程監(jiān)控技術時，個人隱私是另一個令人擔憂的問題，因為它可能導致監(jiān)控增加，并可能導致個人信息的不當披露和使用（Hasal等人，2021年），這可能會影響老年人的自主性和幸福感。有趣的是，Galambos等（2019）評估了老年人及其家人對使用智能傳感器的看法，發(fā)現(xiàn)他們意識到輔助生活的好處與個人隱私之間可能存在權衡。這一發(fā)現(xiàn)表明，隨著人們年齡的增長，使用基于人工智能的技術的利弊可能是老年人及其照顧者需要考慮的問題。

結語

鑒于人工智能在老年人護理領域的發(fā)展速度很快，老年科護士需要更多地了解這一技術趨勢，因為這將影響他們的實踐，因此他們可以為患者提供使用人工智能工具的指導。因此，護士需要更多的教育機會，以了解可用的人工智能技術的范圍，以及如何將其應用于老年人數(shù)據集（Ronquillo等人，2021年）。這些知識將使護士能夠進行人工智能研究，并評估這些預測算法是否有利于臨床決策和患者護理，以及是否值得將基于人工智能的技術引入老年護理實踐。護士可以與計算機科學，工程和技術行業(yè)的同事合作，同時鼓勵病人積極參與，協(xié)助共同設計以人工智能為基礎的工具，以滿足老年人的需要（Blakey等人，2020年）。隨著數(shù)字時代的加速發(fā)展，護理行業(yè)，尤其是從事老年醫(yī)學工作的護理行業(yè)，應該采用人工智能來幫助確定人工智能是否能夠支持老年人的健康和福祉。

THE

END

本文由“健康號”用戶上傳、授權發(fā)布，以上內容（含文字、圖片、視頻）不代表健康界立場。“健康號”系信息發(fā)布平臺，僅提供信息存儲服務，如有轉載、侵權等任何問題，請聯(lián)系健康界（jkh@hmkx.cn）處理。

相關知識

養(yǎng)老院工作人員如何維護老年人的心理健康
老年人健康管理工作計劃 老年人健康管理年度實施計劃(六篇)
老年心理護理（專業(yè)能力課） 老年人健康心理輔導 老年人心理保健的功能.pptx
多元呵護老年人心理健康
鑄就健康晚年：老年人健身的挑戰(zhàn)與機遇
智能穿戴設備在老年人健康監(jiān)護中的應用
人工智能在老年醫(yī)療中的應用探索.pptx
『老年人保健與護理』第四章　老年人的健康保健
智能健康手環(huán)，老年人健康狀況實時監(jiān)控的創(chuàng)新工具
老年人智能健康監(jiān)測手表

網址: 人工智能促進老年人健康：提高老年護理實踐的機遇 http://m.u1s5d6.cn/newsview196412.html