Most people agree that technology has a key role to play in taking pressure off the NHS.
Amid all the recent noise around AI and its applications across practically every sector imaginable, we’ve seen the green shoots of several projects promising to equip trusts with “AI nurses” or “AI receptionists”.
While the potential here is great, more “basic” tools can have just as big an impact on NHS operations without the same cost or complexity.
The NHS’s no-show problem
Missed appointments, from GPS to hospital to non-urgent blood tests, are draining the NHS.
The number of missed appointments and the projected cost of this wasted time is almost mythical now – rising every time new data comes out.
Most recent numbers from the NHS estimate there are eight million missed hospital appointments each year with an estimated annual cost of £1.2 billion to the NHS.
The exact “no-show” rate varies from trust to trust but recent figures predict the average to be around 6.4 per cent, but in recent history, we’ve seen cases as high as 13 per cent.
Even ignoring the one billion pounds shaped hole in the NHS budget, these no-shows create many secondary problems.
Having Drs or nurses on downtime is a waste of resources and reduces the capacity of hospitals or GPs to see patients, resulting in longer wait times for patients.
NHS waiting lists have become infamous with an estimated 7.2 million people waiting for hospital treatment in England alone.
This is partly a hangover from the pandemic but equally down to a capacity crunch – which these missed appointments extend.
NHS trusts can attempt to manage and control these missed appointments, but without supporting technology it can be a nightmare.
Hospitals or GPs often overbook appointments to try and account for no-shows, but if you overbook by 6 per cent you will still see a major variation in attendance day-to-day, leaving things up to chance.
A way around this can be greater emphasis on communication from receptionists, calling to confirm appointments, but this can quickly destroy productivity and pile pressure on staff.
Is AI the answer?
Clearly, this problem needs solving and supportive technology will have a big part to play.
NHS trusts all recognise this and we’re seeing a number of different solutions and trials being rolled out.
One technology that has been highly touted for its potential to transform healthcare (and many other sectors) is artificial intelligence.
For example, trials of “AI receptionists” which “predict likely missed appointments using insights including the weather, traffic and jobs” have recently begun.
While AI will undoubtedly transform many areas of the NHS, it’s possible that, at this moment in time at least, such solutions are overcomplicating the problem.
Using AI on this level also comes with a number of additional challenges.
First among these is their need for patient data to function, which presents security, privacy and operational questions that we may not have the answers to right now.
In addition to this, there can be unintended consequences to relying on predictive algorithms.
Marking people as “high-risk” for did-not-attends (DTAs) for example is already something that should be approached with caution, but when we start asking algorithms to make these decisions based on data alone it has the potential to cause problems.
AI can discriminate against certain groups if the data sets they’re trained on feature examples of discrimination in one way or another.
These challenges can and will be overcome and AI will have a huge part to play in healthcare over the coming years.
It’s good then that the NHS is investing and trailing different solutions, but the immediate answer to the NHS no-show problems lies with humbler tools.
Getting the basics right first
The immediate answer to missed NHS appointments doesn’t lie with a convoluted “Minority Report” style solution that predicts missed appointments before they happen, but by simply enabling patients to self-serve.
Having an automated system allows patients to book and cancel appointments without having to wait in a queue to speak to a receptionist, and can reduce no-shows by up to 62 per cent.
Online scheduling tools are used by some (but certainly not all) NHS trusts, and have been shown to reduce DNAs to 3 per cent – with missed appointments costing the NHS £120 each, this would result in a colossal saving of £48,360,000 a month.
Compared to “AI receptionists” it’s perhaps an unglamorous, boring operational change but it delivers huge benefits.
First of all, it makes it easier for patients to cancel appointments.
While this might feel counter-intuitive, a cancellation is a far better result than a no-show, as it makes it possible to fill the space rather than have it wasted.
Automating this digitally also means filling these gaps is incredibly easy, as patients have visibility over available slots – from a trust’s point of view, the gaps fill themselves.
Arguably, this is where so much of the value lies – not only do patients not have to wait to speak to a receptionist, but staff get so much time back that productivity and staff wellbeing see huge lifts as a result.
The right digital scheduling tool can also improve patient accessibility.
Healthcare systems have to work for everyone, so having multiple options for communication, like online forms and automated phone lines can help with this.
Studies have shown that over 73 per cent of all patients want to be able to book medical appointments online.
The final major benefit of a self-serve scheduling solution is that they are fairly lean, meaning they are easy and inexpensive to implement and don’t need extensive integration or access to patient records.
While AI is not to be ignored, and will undoubtedly change the way we live and work forever, the answer to the NHS no-show problem is (thankfully) much simpler than this.
A fairly basic scheduling tool can have just as big an impact on missed appointments, productivity, and patient and staff satisfaction.
This is a more cost-effective way to increase the NHS’s productivity and save millions, if not billions, of pounds.
