The UK is 鈥渙n the cusp of a huge聽health tech revolution that聽could transform patient experience鈥�, said health minister Matt Hancock when he announced 拢250 million to fund a new AI Lab for the National Health Service earlier this month. The lab has been set up聽to bring together academics and technology companies to work on some of the biggest challenges in health and care.
But the AI sector has a reputation for overpromising on what it can deliver 鈥� as do politicians. I met with Indra Joshi, head of digital health and AI at the newly established NHSX聽鈥� an organisation tasked with digitally transforming the NHS and running the AI Lab聽鈥� to ask her how it is going to achieve these aims.
A former medic, Joshi says she聽still knows what it is like at聽medicine鈥檚 coalface. She recounts a time her ward ran out of pillowcases, so staff had to wrap up pillows in sheets instead. 鈥淎s much as in my day job we feel that we are going to solve the world, we can鈥檛 lose sight that these are some of the issues that staff face,鈥� she says.
Advertisement
How is AI going to make a difference? Joshi tells me about efforts to develop software to聽look at scans, such as mammograms for breast cancer screening, and measure the size of any lump and flag it for a radiologist鈥檚 attention if necessary. 鈥淭he more mundane tasks are taken away, leaving time to do more complicated ones,鈥� she says.
Also in the works is a computer-based check-in system for emergency departments. At the moment, people initially speak to a receptionist and then a triage nurse, before seeing a doctor. Instead of those first two steps, it is hoped people could record their symptoms through a touchscreen device that could also take some of their vital signs such as heart rate.
One advantage would be that every person鈥檚 data is digitised, so patterns can be spotted more quickly. 鈥淭hat allows a hospital to understand what鈥檚 happening on the ground,鈥� says Joshi.
At a more basic level, her team聽is also investigating the potential for phone apps to let people book and rearrange hospital appointments, and even find out test results.
These kinds of initiatives may聽sound promising but some doctors have responded with scepticism, pointing to the recent explosion of commercial apps making various health claims without evidence.
Joshi says her team is tackling this by laying the groundwork for ensuring all future health tech is ethical and supported by evidence. It has drawn up a code of conduct and has worked with the National Institute for Health and Care Excellence (NICE) on setting the standards for the levels of evidence required for different types of technology.
鈥淲e need to help the market understand that when you鈥檙e developing [technology] you聽claim is diagnostic or therapeutic, you need to go聽through the same peer-reviewed process as with any other product that鈥檚 therapeutic or diagnostic,鈥� she says.
Another criticism from frontline staff is that the NHS shouldn鈥檛 be working on new tech without first fixing existing IT infrastructure problems. Dominic Pimenta, a cardiologist in London, has complained that his computer can take half an hour just to get going in the morning. To run his clinic, he聽says he must log in to 10聽different software packages 鈥� to see someone鈥檚 notes, scan results, ECG and so on. 鈥淭he other day it was taking so long I just gave up and started again in a different room,鈥� he tells me.
Other problems include a lack聽of tech support out of hours and different hospitals using incompatible software. I ask Joshi if founding the AI Lab is like trying to build an extension onto a house whose walls are about to fall down. 鈥淭his isn鈥檛 an聽鈥榚ither-or鈥�,鈥� she says. NHSX, which only launched last month, is taking measures to聽address these problems.
Joshi says she understands people鈥檚 frustrations. 鈥淲hat we want to do is build frameworks to ensure we can deploy [technology] and it works where聽we say it will work.鈥�
