We all know the common phrase that history repeats. I get where it comes from, but I prefer to say that history rhymes. It’s close to the same, but there are always nuanced differences with the next iteration that looks and feels very much the same. I think we’re going through that right now with AI. It’s something I’ve been talking a lot about in the healthcare AI keynotes I’ve been giving along with on podcasts and panels that I’m on. Here’s the classic cycle we’re going through right now with healthcare AI.
Every healthcare CIO wants to have one AI platform, but right now they have no choice but to choose best of breed AI solutions.
If you’re a CIO, you know exactly what I mean. You’ve been tasked with AI innovation by your board. Ideally, you’d love to buy one AI platform that you implement at your organization and it satisfies all your needs. Unfortunately, no platform exists (yet?). If you want to leverage AI, you’re going to have to use the best of breed AI solutions out there and bet on which ones will become the AI platforms of the future.
As I mentioned, we’ve been through this before. When we first implemented health IT in healthcare organizations, we had this exact same challenge. Organizations had no choice but to implement a wide variety of best of breed solutions. There was no all-in-one platform. They implemented an accounting system, an EMR system (renaming it to EHR came later), a lab system, a pharmacy system, etc. This worked fine, but implementing this many systems came with overhead. Plus, once you wanted those systems to communicate with each other, the management became a nightmare.
What happened next is a preview of what’s to come with AI in healthcare.
Little by little the EHR vendors starting rolling out their own solutions that solved more and more of the software needs of a healthcare organization. I can still remember the conversations that Epic Beaker wasn’t as good as the dedicated LIS (Lab Information Systems) solutions out there, but it wasn’t too awful.
You can imagine the conversations that happened next. Why do we have 2 vendors? Why don’t we just pay 1? The integration will be better if it’s the same system. We won’t have to worry about the finger pointing between vendors. etc etc etc. I’m sure this will bring back many memories for people that lived it.
Long story short, most healthcare organizations got rid of the niche solutions, which were actually better solutions, and went with the all-in-one EHR vendor so that they could have fewer vendors and a solution that was fully integrated. All of this led to many organizations’ policies of EHR only or at least EHR first as they evaluate solutions.
The problem with healthcare AI is that it moved so fast that the EHR vendors couldn’t keep up. The number of AI solutions in healthcare right now is mind boggling. In fact, it’s probably the hardest IT challenge that healthcare organizations have faced in a while. Where do they start and which are the best solutions to implement today. The AI solutions are coming out so quickly that even with every EHR vendor announcing a roadmap of hundreds of AI applications in their system, there is still a ton of opportunity for healthcare AI solutions to do something the EHR isn’t doing.
Thus, we’ve entered the part of the cycle where healthcare CIOs have to decide to sit out or implement the best of breed healthcare AI solutions out there. The problem with sitting out is that your organization will miss out on the benefits that AI could bring them today. Plus, there’s a lot of learning that happens when you start using a new technology in your organization. Those healthcare AI “reps” create a lot of value for an organization as it continues to evolve. It’s hard to see where AI is headed and how it can benefit your organization watching from the sidelines.
Thus, every healthcare CIO and the associated AI governance committee is putting together processes and procedures to evaluate and implement AI solutions in their organization. That’s a good thing because it’s going to drive a lot of value. However, history teaches us that a few years from now, we’ll be sunsetting a number of these AI solutions and opting for the all-in-one AI platform.
Will there be one AI platform to rule them all? Will the EHR be the one AI platform?
I can’t imagine anyone thinking the EHR won’t be one of the major AI platforms that healthcare organizations use. However, it’s hard for me to imagine a scenario where the EHR is the only AI platform for healthcare organizations. It’s probably in their best non-monopolistic interest to not be the only AI platform too.
I personally think that hospitals and health systems will have a half dozen different AI platforms that are based on very specific areas of their organization. It’s not hard to imagine having an RCM AI platform that handles all of your revenue cycle management needs. It seems obvious to me that there could be a radiology specific AI platform that does all your radiology AI. I could imagine a whole back office AI platform for hospitals and health systems. My point is that this time I think we’ll see consolidation of AI onto platforms, but I don’t think we’ll see one monolithic AI platform that covers every AI need of a hospital or health system. On the independent ambulatory side, the EHR vendor may be the AI platform for that space, but we’ll see how that plays out as well.
What do you think? How do you think the healthcare AI market will play out? Do you see the same classic tech cycle playing out like it has before or will there be some unique nuances? Let us know on social media.
No comments:
Post a Comment