A recent debate article, published in BMC Geriatrics, has explored whether artificial intelligence (AI) health monitoring may be a suitable enhancement or replacement for elder care.
An ageing population
The life expectancy of individuals worldwide is increasing. In turn, this challenges the sustainability of traditional care models that have depended on direct monitoring. Estimates indicate that in 2050 the global population age 65 and over will double to 17%. Longer life expectancy is met with longer periods of living with impairments and chronic conditions that impact quality of life. With an ageing population comes needs of higher personal assistance and care. Unfortunately, in many countries, there is a severe shortage of direct care workers. Simultaneously, the health workforce is also ageing and replacement remains an ongoing challenge. The ongoing COVID-19 pandemic has displayed a disproportionate impact on older adults, particularly those within care facilities. Therefore, it is essential to develop further strategies to help individuals receive health monitoring with minimal in-person contact.
Remote monitoring technologies, such as video cameras, provide some assistance in supporting older adults to live independently. However, these technologies still rely on human operators and can be prone to human distractions and errors. As a result, attention has turned to the use of automated technologies, such as AI health monitoring, to enhance older adults’ ability to live safely in their desired settings.
What can AI health monitoring offer?
AI-powered health monitoring technologies go beyond commercial and medical devices as they are endowed with processes that mimic human intelligence. They play a unique role in caring for older adults. They complement current care provision, reduce burden on family caregivers and improve quality of care. Machine learning algorithms can analyse a vast amount of data from longitudinal observations. For example, AI-enabled blood pressure or electrocardiogram monitors can help predict various health concerns, e.g. hypertension.
Other AI home-health monitoring systems, such as computer vision analytics, can classify activities, e.g. standing or walking. These systems can then iteratively learn expected movements or activities. Importantly, some AI monitoring programs can analyse input data and detect whether an older person is taking gradually longer to gain balance while trying to stand. Once the system has predicted health decline, it can decide to intervene based on a pre-set risk threshold. These automated alerts can facilitate timely and safe care and prevent potentially serious injuries.
Clinically meaningful and actionable AI monitoring data can also be used to inform medical decision making and transform health delivery. These technologies can optimise human resources as healthcare providers can ensure that the right patient has timely access to appropriate care. This information could also be integrated with electronic health records to facilitate therapeutic interactions and may be particularly helpful to support elderly patients who find it difficult to articulate their symptoms or needs. Furthermore, AI monitoring will generate continuous data which will provide a better idea of health assessment over time rather than at one time point.
What are the ethical considerations of using AI health monitoring?
A recent systematic review studying sensor monitoring reported that there is only limited evidence to suggest its effectiveness, as most of these technologies are still in early stages of development. Another concern is the fact that most AI health technologies are designed without explicit health considerations regarding the impact of these technologies on end-users. In turn, this may limit the responsive and ethical translational potential of these technologies in people’s homes.
The authors suggest that exploring the goals and priorities of older adults using health monitoring platforms may help to develop and implement technologies that can truly promote autonomy. For example, researchers have shown that older adults are interested in co-designing these technologies and controlling their data. However, many existing systems are relatively established and are difficult to customise according to user preference.
It is important to consider how older adults and their families may weigh personal and data privacy considerations against gains in independent living, physical safety and convenience. People have also raised questions about how AI technologies will impact family dynamics.
Conclusion
Automated monitoring systems can provide an overview of older adults’ activity patterns and can be valuable tools for healthy aging. If developed ethically, these technologies can be assistants for older adults, family caregivers and healthcare professionals. Nevertheless, it is important to design these technologies with intersecting clinical and ethical factors in mind. In addition, the authors emphasise that we must ensure that these monitoring technologies provide older adults with greater access to the healthcare system rather than exacerbate the digital divide or social isolation.
Image credit: https://www.freepik.com/vectors/people People vector created by stories – www.freepik.com