fbpx
Connect with us

Insight

Can AI help keep the human experience in social care?

By Dec Norton, Director of Development, CareLineLive

Published

on

Social care, by its nature, is an inherently human activity. At its heart, it will always be that way. But AI is making inroads into the sector, driving workflow efficiencies, overcoming resource shortages and alleviating staff pressures. 

As services continue to battle a workforce crisis, just last month it was revealed that AI is being used to train staff in a “UK first for the social care sector”.

And with the government aiming for 80 per cent of CQC registered providers to have digital social care records by March 2024, AI could be crucial in meeting the administrative demands of social care. 

Now four years old, the NHS AI Lab is carrying forward its mission to safely and effectively adopt and scale AI in health and care by bringing together government, health and care providers and tech companies.

Yet despite AI development making strides in recent years, it is still in its infancy.

For many applications there is a huge cost associated with implementing AI tools with the appropriate protections for privacy and security. 

These issues are particularly apparent with AI in adult social care.

Looking to boost awareness of the technology’s use in the sector, the AI Lab “held conversations with technology companies and care providers across the country, which showed that – in general – social care is behind health on the development and adoption of AI”. 

Why is it lagging behind? And how can AI be used to improve services? 

What AI is good for health and social care? 

The recent success of OpenAI and ChatGPT have reduced the barrier to entry for many organisations, but they are not suitable for use in health and social care problem areas.

Large Language Models (LLMs) are trained in a way that the output is simply a best guess at answering any questions posed to them.

There has been plenty of coverage around hallucinations and outdated information being reported, which doesn’t suit a sector in which there is no margin for error. 

Whilst LLMs or Generative AI may not be ready for primetime in health and social care yet, there are many other AI technologies that can help with data analysis, personalised care plans, companionship and, most importantly, giving time back to caregivers by helping them to accomplish common tasks.

A successful implementation of AI technologies is always going to depend on harmony between traditional software, the providers and caregivers, and emerging technologies. 

AI’s own care role

With regards to the social care sector in particular, service providers have a plethora of operational challenges to overcome.

The AI Lab offers possible reasons for social care’s AI adoption lag, stating that “the provider landscape is generally more fragmented and many organisations still collect data manually”.

Therefore, its adoption needs to come with structural organisational changes and evolved working processes. 

Moreover, while AI can help with a lot of these challenges, it’s not a silver bullet.

Traditional software will still dominate the sector.

But it’s exciting to see what can be achieved by applying AI to routing problems to make travel more efficient, data analysis with the goal of finding new correlations between observations, and how it might be used to improve other day-to-day operations that ultimately detract from the important parts of service delivery.

And in an example of a direct practical use with patients, AI is currently being employed in south west London care homes to reduce falls and hospital stays.

Sensors – installed with the patient’s consent – are able to learn normal patterns of noise and movement and therefore send alerts to staff if unusual patterns are detected.

As the article highlights, falls are estimated to cost the NHS £2 billion a year, and “a similar pilot saw the number of night time falls more than halved.”

There is clearly great potential in such use cases, and similar systems could be set up in homecare to alert local services and staff. 

Abiding by privacy and regulation 

Data privacy is a pertinent aspect of any software program, but is especially crucial in such a vulnerable sector which is already under attack.

When care agencies implement a software solution, organisations such as Digital Social Care provide a resource for advice for being compliant and adhering to security standards. 

Customer data should never be under threat, and organisations must ensure that they have a range of protocols in place to mitigate any unforeseen attacks.

This includes pre-empting more extreme cases with backups replicated in multiple locations, failover mechanisms to restore and recover data, and a comprehensive disaster recovery plan. 

But regulating AI is a different ballgame. It carries its own costs, privacy concerns and public fears.

As conversations around AI regulation pick up pace, NICE has unveiled its AI and Digital Regulations Service to offer guidance to the NHS, social care providers, digital health innovators and the wider care ecosystem for adopting and using new AI technologies.

If AI is going to reach the potential it can, such support systems and regulations are vital for its ethical and safe implementation. 

More time for the human experience  

Social care is facing a workforce crisis that has wider consequences for the entire NHS system.

As we see record waiting lists and resources being stretched, initiatives such as the NHS AI Lab are fighting to buck this trend.

For a sector particularly feeling the strain, AI represents a viable option for reducing pressures on social care staff in particular. 

While large use cases and advanced AI technology may still remain impractical and unsuitable, there are some immediate benefits to be delivered by AI to improve day-to-day tasks and care.

With the appropriate funding, expertise, support and regulation, the technology can have a positive impact on the sector.

By using it, we can free up more time for the human experience in social care. 

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending stories