While there is legitimate excitement about AI’s potential benefits, it’s critical to establish strong foundational digital health infrastructure. Dr Ed Hutchison, Clinical Solutions Specialist at Altera Digital Health, explains
AI is now impossible to ignore. You cannot move without seeing some sort of AI advert promising a capability that is going to revolutionise your life. Need a new fridge? It better have inbuilt AI. Need to book a holiday? Why don’t you talk to our AI assistant? Want to create a presentation? Don’t bother, just let AI take care of it.
It has become a buzz phrase across every industry. Even Allbirds, a popular shoe company, recently ditched its entire business model to pivot into becoming an AI company.
Healthcare is clearly no different. Each catchy strapline promotes an AI offering that will transform your organisation into a digital health utopia. Yet, through my clinical work, I see firsthand how the reality is often far more complex.
So how do we manage this landscape?
Or, maybe it is better to ask: How do we truly know what we need to improve our healthcare systems? Certainly, one way is to speak to those who deliver healthcare. They can tell us their pain points. They can tell us their rate-determining steps. They can tell us what they actually need to solve to improve patient care.
Jokes aside, I do believe that a lot of the AI hype is legitimate and we are increasingly starting to see some real benefits. However, we must always pause and consider what we are aiming to achieve with AI, rather than racing ahead with no clear goal.
Likewise, we must always remember the foundations of digital health: useable hardware and deeply interoperable and integrated EPR software. No amount of AI is going to fix the broken mouse at my computer. No amount of AI is going to instantly remove paper notes. Don’t try to ice the cake you haven’t baked yet.
That said, who doesn’t still enjoy a taste of the icing while the cake is in the oven? While firming up the foundations, we are right to explore the benefits that AI can bring and begin to embrace its potential.
Ambient Voice Technology (AVT) is one such example. One of my favourite stats is that resident doctors spend around 80% of their time documenting and doing admin. This leaves minimal time to do actual ‘doctor-ing’ or spend time with the patient. Telling clinicians that their documents could be seamlessly written as a byproduct of a conversation will always be welcome. But only if the solution they are using doesn’t then slow them down in other ways. Is it accurate? Does it pick up conversational nuance? Is it running natively in my EPR? Does it need a separate login? Can health issues and requests be automatically captured and integrated into the patient record? If the answer is no, then there is still work to be done.
Other examples include using AI to provide patient summaries or prompt the user in the EPR on what actions might have been missed or need completing. A clinical shift is an overwhelming assault on the senses, and trying to remember everything that needs actioning can be a challenge. Just watch an episode of ‘The Pitt’. Any help is appreciated.
Helping where and when it counts
AI built into existing EPRs and workflows that summarises the key details, prompts the user with pre-configured request forms or automatically triggers care pathways is what I, as a user, want. Don’t try to change my life; just help me do an already difficult job a little more efficiently and reduce the time spent on admin. I am sure I am not alone. Speak to the users, watch how they work, then create the solution.
Leave a Reply