HIMSS Analytics looks at the adoption and use of healthcare IT in Europe. We assessed its recent findings to understand rate of e-health adoption. Here are key takeaways for change agents within healthcare:
Use of AI in healthcare is very limited at present
A massive 84% of health professionals across Europe were either not using AI tools, or not aware of doing so. This is unsurprising, given that almost 60% of healthcare facilities were not already using AI tools, and had no plans to start doing so. A quarter had specific plans for AI tools, but only 16% were already using AI tools, and those tools were generally fairly basic.
AI use in healthcare varies considerably—and possibly in unexpected ways—across Europe
AI use and planned use were both highest in the Nordics, followed by the Netherlands, at over 20% use, and 30% planned use. In the DACH region, however, actual use was 10% or less, and planned use only slightly higher in Germany or Austria. Planned use was highest in Spain, at 42%.
Software vendors were more likely to be planning a launch than healthcare providers a purchase
Over 50% of software vendors were planning to launch new AI products for the healthcare market, and the vast majority (over 75%) within the next two years. However, only 25% of healthcare providers planned to purchase AI tools within that period. This suggests that the market may be at risk becoming saturated long before providers are ready to buy, which may cause development to stagnate later.
Planned AI purchases were not necessarily supported by funding
Of those planning to use AI tools, 32% planned to do so within a year, and a further 37% within 2 years. However, when asked whether their organisation provided enough budget for AI tools, two thirds said ‘no’. There is an argument that nobody ever feels that they have enough budget, but this nevertheless indicates that ‘plans’ may not have budgetary support. This, in turn, suggests that the market may not take off in quite the way that vendors hope.
There is generally a reasonable balance between supply and demand—but which is the driver?
Overall, there is similarity between the AI tools that healthcare providers are using or planning to buy, and what is available or in the pipeline. There is, however, a question about whether customer needs are driving software development, or reflecting what is available. After all, healthcare professionals have little time to spare thinking about what might be possible with AI, and are reliant on vendors to suggest ideas. Might vendors’ innovation processes benefit from a little co-creation and design thinking?
The expected benefits from AI very widely between vendors and healthcare providers
Healthcare professionals were most likely to expect the benefits from AI to focus on improving the quality of care, and of medical decision-making. Vendors, however, were more likely to suggest that the benefits would be from reducing both costs and medical errors. This is a fairly fundamental mismatch in expectations, and again, suggests that perhaps a little more user involvement in design processes might be beneficial.
Perceptions of challenge areas show a wide gap between customers and vendors
Healthcare professionals say the biggest challenge is lack of maturity of AI products, but software vendors believe the problem lies in lack of trust among healthcare professionals, and lack of good quality data on which to train models. In other words, both sides are blaming each other for the slow uptake of AI tools in healthcare, which does not bode well.
Nobody is worrying about patients’ perspectives
Interestingly, nobody—vendors, healthcare providers or government agencies—thinks that patient acceptance of AI tools will be a problem. Perhaps we had all better hope that this optimism is founded on evidence, and not just a lack of awareness and/or research, because patients’ refusal to engage could result in serious wastage.
Healthcare professionals have high hopes of AI
The most-used words in response to a question about the likely benefits of AI were ‘better’, ‘quality’, ‘support’ and ‘data’, suggesting that there is huge optimism that AI will help to deliver improvements in quality. These are expected to include diagnostics, individualised and personalised care, and better decision-making.
Despite being optimistic, healthcare providers recognise that the benefits of AI could take years to deliver
Healthcare providers were optimistic about the potential for AI in healthcare. They were also however, cautious and realistic about the fact that the potential might never be delivered, or that it might take far longer than expected for the benefits to be realised.