ICS prescribes Amazon’s Alexa to patients

Staffordshire and Stoke-on-Trent ICS has provided around 400 smart speakers to patients to improve their self-managed health and wellbeing.

It is thought to be among the first examples in the NHS of a health system prescribing Amazon Echo to patients. The smart speakers are known as Echo, but the product is more commonly referred to as “Alexa”, which — when spoken by their user — activates the software. 

The scheme is run by staff from North Staffordshire Clinical Commissioning Group, University Hospitals of North Midlands Trust, North Staffordshire Combined Healthcare Trust, and start-up tech-training company Wavemaker. Funding for the Alexa Echo Show devices came from North Staffordshire CCG, NHS England and community interest company Wavemaker.

Paul Beaney, one of the project leaders, told HSJ the feedback from patients was “overwhelmingly positive”.

“The health benefits from using digital assistants were remarkable, especially in terms of medication adherence and mental health,” he said.

Mr Beaney added examples of how the Alexa Echo device had helped patients included:

• Diabetes patients being able to record their insulin doses and set reminders;
• Cardiac rehabilitation patients accessing hospital exercise videos; and
• People struggling with loneliness finding comfort in talking to another voice in their property.

Scheme avoided information governance requirements

However, when the NHS signed a deal with Amazon in 2019 to enable Alexa devices to quote health advice from the NHS’ official website, it prompted some concerns from campaigners over privacy and data.

Asked about the project’s information governance, Ruth Chambers, who led the CCG’s involvement, said the devices were deployed in a way that “avoided NHS IG requirements”. This was done by getting patients or their carers to sign disclaimers, and ensuring there was no clinical interaction between the patients and the NHS. 

Ms Chambers said: “The information fed into the device by the patient or carer (eg the messages, usage data or responses) was not seen by our team. The devices were gifted in line with the disclaimer in return for them using it for a few weeks and pledging to respond to an evaluation call.

She said they did not seek IG approval because there was “no direct clinical interaction in identifying the patient or accessing any of their information”.

Instead of clinicians choosing patients, recipients were identified by a range of public and third sector groups, including social prescribing leads in primary care networks, a diabetes patient group and local cardiac rehabilitation services.

Ms Chambers said the team reported its model to the ICS’ digital workstream board and two other relevant groups in the county and “no challenge of IG approval omission was brought up”. 

Mr Beaney added: “We advised how Amazon collects and stores data, and the privacy features available to use. Thus, patients were empowered to decide their own level of risk and comfort with sharing data as anyone does when they sign up to a new website or join social media…

“Taking this pragmatic approach can enable ordinary citizens to benefit from current accessible devices for their self-care as technological advancement inevitably outpace IG legislation by the day.”

Mr Beaney said the scheme had been delivered through “successfully built-up interdisciplinary and multi-organisation collaboration” within the ICS.

He added: “The way we delivered the project exemplifies the importance of forming local partnerships across health, social care and third sector organisations — something that ICSs are expected to prioritise.”

Ms Chambers said the scheme will continue in Staffordshire, with another 14 devices set to be distributed shortly. She and Mr Beaney have written a further article on their work (see attached).