We look at the constant hype surrounding this overblown topic and ask; can it deliver for you?
On the basis of what we have seen – probably not. Although according to multiple independent reports “When it comes to healthcare AI, the UK is the powerhouse of Europe” – there are conflicting statements by UK Government saying that “dealing with the NHS remains challenging”. And looking at a recent statements from others, the evidence from startups and those tech companies who are at the forefront of developing new AI based solutions, take up is small, and there is “little benefit to the value proposition”, from using AI per se. So there is a conflict between commercial take up – and perception of relevance.
What is worse, is that we are already seeing one hospital terminate their digitalisation journey, because – as one clinical Consultant told me – “it just doesn’t work”.
The reasons it “doesn’t work” are many and varied, but largely fall into two camps, first of which is because the introduction of AI or Digitalisation, is not an IT or Tech discussion. It is a business process discussion. It is an HR and best use of people, discussion. If it remains easier to flip the paper pages of a file to see someone’s latest notes – then that is what it is. But also – the announcement of a single IT decision point of the new NHSX quango, is itself a misnomer. At the upcoming Digital Health Conference, focussing on the new NHSX facility – there are no less than six Decision Makers all involved in the decision making process
So… what are the practical steps that you need to take, to get the best out ofAI based new technology?
1. Understand and create a Roadmap, of what you want to get out of this process? If it is simply to shore up your existing practices, then forget it.
2. Know which areas you wish to include – both from a data access point of view, and also groups of people. The more groups involved, the worse it will be.
3. Understand that what works for one hospital, may probably not work for you. Make clear choices about solutions that can deliver a specific benefit.
4. Do not engage in Trials. AI data management is not a clinical discussion, and the algorithms used are already proven. You are already good to go.
5. Have milestones of progress.
6. And only when you have all of the above written on a piece of paper – then involve your IT people.
You may well now find that the money you had previously allocated for something nebulous, will indeed deliver when broken down into manageable practical specifics.