Where AI can help care for the sick and elderly – and where it can’t

Assistance robots like this one could be used more and more frequently in care. — dpa

As populations age, more and more people are asking how best to organise care for the elderly in future.

Trained staff will be important, as will technology and innovation – which may include artificial intelligence (AI).

Some people get fearful when talk turns to AI, a topic riddled with misconceptions.

At the end of the day, what it means is the attempt to map human decisions using computers, says Andreas Hein, an expert in assistance systems at Oldenburg University in Germany.

In a medicine or health care setting, that means providing doctors and nurses suggestions that a computer has created based on data. A person makes the final decision, says Hein.

AI is already in use when it comes to outpatient care, for example in route planning, which takes patients’ travel times or preferences into account.

For the last few years, AI has mainly been about ensuring that machines can learn, using data. No rules are defined in advance, Hein says, they are extracted from the data.

For that to work, the data has to be good. That means you need to know the basis for the decisions and the decisions that were made on the basis of those, he says.

That sounds complicated but Hein says this could apply in radiology, for example. You can train systems using archived computer tomography images and the diagnoses that were made using them.

The only problem when it comes to nursing is that so far, that data does not exist yet, but is just in the process of being developed, Hein says.

“Without data, there is no AI,” says Karin Wolf-Ostermann, a nursing scientist who heads a nursing care research department at a German university.

In the future, documentation will be one of the major AI issues, according to Hein. Nursing staff spend 30% to 50% of their time on documentation.

If that could be reduced somehow, it would free up time to help patients instead. “We believe that we will have to feed in even more data here to be able to support documentation with the help of AI,” Hein says.

The fields where AI is applued in care are a reall mix, according to Wolf-Ostermann, who lists monitoring care recipients’ health status or activities, alarm management and detecting falls, service and medicine planning.

However, so far, there is very little large-scale use of AI in these areas.

AI applications can also provide social support, interaction and helping people to be mentally and physically active. All that can be valuable when it comes for those with dementia, although it is also controversial from an ethical perspective, says Hein.

The fear is that creature-like robots, for example, or other technical substitutes will reduce the amount of time that caregivers spend with dementia patients, who won’t get the attention and interaction they need.

However, so far, studies show there have been positive effects in the short term, as long as the robots are well integrated into the patient’s daily care routine, says Hein.

It is hard to imagine nursing today without technology, says Wolf-Ostermann. Nurses are generally very open to new technologies, she says, based on her own studies.

Nursing should make active use of the scope for design by engaging in a broad discussion of technological innovation, and doing so in good time, she says.

That debate needs to be open, with a view to looking at both the opportunities and the risks.

People always need to be in focus, both researchers agree.

“By no means can this be about replacing caregivers, but about supporting them,” says Wolf-Ostermann.

You can’t replace social skills with AI, she says. – dpa