From listening devices that detect falls to “patient sitter” systems in hospitals and robots helping with exercise in care homes, Singapore is looking to artificial intelligence to help manage the health of its elderly population.
By 2030, a quarter of Singaporeans will be 65 or older — in 2010, the figure was one in 10 — and it’s estimated that around 6,000 nurses and care staff will need to be hired annually to meet Singapore’s health workforce targets.
Technology is much needed to help fill the care gap in Singapore and elsewhere, according to Chuan De Foo, a research fellow at Singapore’s Saw Swee Hock School of Public Health. Societies around the world are “dismally unprepared” for an aging population, Foo wrote in the science journal Frontiers last month, and with his co-authors described AI and other technologies as “pivotal forces with the potential to drive a paradigm shift in healthcare.”
For Foo, artificial intelligence is set to play a “huge” role in elder care in Singapore, both in terms of helping clinicians manage non-acute conditions and in overseeing administrative tasks such as monitoring the availability of hospital beds, he said in an email to CNBC. “As the elderly in Singapore get more IT savvy, we see them turning to teleconsultations and digital tools that utilize AI technology,” he said.
AI is also being used to detect diseases earlier, an area of personal interest for Dr Han Ei Chew, a research fellow at the Lee Kuan Yew School of Public Policy in Singapore. He said his late mother’s diabetic eye disease could have been diagnosed — and treated — earlier had AI testing methods been available when she was alive, as they are are now. “That would have been so useful when the family was going through that journey,” Chew told CNBC by phone.
A big focus for Singapore is “aging in place,” according to Chew. “We can deploy the AI, but it isn’t about fully replacing human care … it is really about assisting the caregivers and helping seniors to stay independent and age in place,” he told CNBC via video call.
Chew said Singapore’s Housing and Development Board is even offering built-in home technology to detect when someone falls down, with an alert sent to a resident’s next of kin or connected to a call center for help.
These types of monitoring technology need to be used carefully, Chew said, in whatever jurisdiction they are deployed. “The AI should empower the seniors and not strip them of control. They still need to have the choice to opt in, set boundaries, and, more importantly, to turn it off when they want,” he told CNBC.
A care ‘co-pilot’
It’s not only Singapore that is looking at using AI for elderly care. In the United States, Sensi.AI is a fast-growing “care co-pilot” that monitors elderly people using audio devices that are usually plugged into three areas of their homes.
Company co-founder and CEO Romi Gubes said the technology can provide caregivers with more than 100 different insights, alerting them to early signs of urinary tract or respiratory infections, or to falls or cognitive decline. “We’re combining multiple indicators that are coming from audio,” Gubes told CNBC by video call. “Think about, for example, respiratory infection. This will [take into account] the cadence of the coughing, the frequency, the type of coughing, together with … complaints around fever, dizziness,” she said.
When Sensi.AI is installed in a home, it creates a “baseline” over two weeks, noting a range of “acoustic indicators,” Gubes said, including non-verbal sounds like objects being moved, footsteps or snores, which it combines with its team’s clinical knowledge. Once the AI knows the baseline sounds in a home, it can alert caregivers to any audio anomalies that might suggest a health issue.
Gubes said Sensi is being used by “tens of thousands” of seniors in the U.S. and a spokesperson said the company is in discussions about a potential expansion in Asia.
Ageism in AI
The experts CNBC spoke to warned that AI must be used carefully when it comes to senior health care.
Foo warned that the over-use of AI in consultations might lead to “poorer health outcomes” as not all elderly people can use technology, and he warned that it must be correctly designed to avoid “perpetuating digital ageism.” Indeed, the World Health Organization cautioned, “The implicit and explicit biases of society, including around age, are often replicated in AI technologies,” and its 2022 policy brief urged developers to have older people participate in the design of new technology.
In Singapore, the government’s “Action Plan for Successful Ageing” details its aims, such as to reach 550,000 seniors with a health and wellness program and reduce hospital deaths from 61% to 51% between 2023 and 2028.
But Foo said seniors’ opinions needed to be taken into account when determining how AI can address their health needs. “Like all new initiatives, failure will be inevitable if the target audience, i.e. the elderly, are not on board. We [need] to hear their voices and tailor the national health-AI strategy to suit their needs while not removing the human element of healthcare. That is the challenge,” he told CNBC by email.
For Chew, the approach to elder care will need to blend human and machine, describing it as “high tech, but high touch.” “The AI is probably best used as an extra set of eyes, ears and the robots [are an] extra set of hands, but not as a replacement for the high touch human care giving,” he said.