Ginni Rometty, CEO of IBM, has famously said:
“The reality is this technology will enhance us. So instead of artificial intelligence, I think we’ll augment our intelligence.”
But when it comes to healthcare, what are we really talking about?
What can AI do in healthcare?
Healthcare AI is all about data, and the practicalities of what you can do with data. While it is correct to say that the NHS has a lot of data, much of it is not in a format that makes it usefully accessible or interoperable. The result is that the wealth of information potentially available within health records is often inaccessible and underused.
Data-driven technology can be used for:
- Individual care and patient safety
- Improving diagnosis, treatment and prevention
- Planning services and evaluating impact
Most in the sector have heard of Babylon but some lesser known innovations are tackling some of these day to day issues. For example, CogStack is an information retrieval extraction platform for natural language processing. It takes all kinds of data formats and puts them into a search engine. Within a three week pilot, teams tripled the output of data recorded within a hospital and found activity that the hospital could not find, resulting in increased revenue.
Historically, technology in healthcare settings has been in the hands of the clinicians. However, patients are becoming increasingly used to using technology in their everyday lives. With that in mind, the A&E department at Queen Mary’s Sidcup has been trialling eTriage: a digital first triage service loaded onto tablets handed to patients, which can obtain information and risk stratify those patients within minutes of their arrival. As well as improvements in patient safety by identifying critically ill patients more quickly, the pilot found that time to treatment was also reduced by an average of seven minutes per patient.
Patients are also benefitting from practical tools such as DrDoctor which uses massive data sets and algorithms to predict appointment attendance and schedule appointments at the best time for individual patients. The result is that the patient gets an appointment they are more likely to attend, improving the patient experience and saving money in missed appointments.
Concerns about AI
Concerns have been raised about the privacy of health data. We are at the beginning of a technology revolution and, if we have learnt anything from history, we must get the foundations right and have an eye on the impact for the future if we are to secure the best outcomes.
Stricter laws around the use of individuals’ data, like the GDPR, demonstrate the increasing importance of data privacy, but this is not enough on its own. As well as legal requirements we must address ethical questions around the use of data. It is therefore important to engage openly with the public, patients and staff to ensure that everyone feels they have the opportunity to be heard.
Recent research by Understanding Patient Data has shown that if people understand what their data is being used for and the benefits it could bring they are generally more willing to share it. A failure to engage risks a public backlash that could halt the adoption of an AI application in its tracks.
NHSX and the AI lab
NHSX recognises the importance of getting this right and building the right foundations. Together with Academic Health Science Networks they have been asking questions about how developers are using data-driven technology. They want to use this information to help progress their ambition to make the UK the best place in the world to develop safe, ethical and responsible AI for health and care.
In addition, NHSX has recently launched the AI lab. Set up in partnership with the Accelerated Access Collaborative, the lab aims to “fully harness the benefits of AI technology that can augment the ability of our clinical teams across the NHS” within safe and ethical boundaries, speeding up the work to get benefits to more people - patients and staff alike - more quickly.
The lab will prioritise supporting those solutions that can help busy staff in the NHS in very practical ways, what Dr Eric Topol describes as “the gift of time”, so that they have more time to spend with patients in direct care. The intention is that this, of itself, will transform patient care and experience.
Digital technology is having a huge and positive impact in all sectors, including health care. The NHS has a tainted history when it comes to IT but Matthew Gould, NHSX’s CEO, is optimistic that NHSX has the powers to get the job done. As for AI, time will tell if the outputs from the NHSX AI lab are adopted widely but addressing the legal and ethical issues at an early stage makes sense.