There is no doubt that artificial intelligence (AI) in healthcare is a high-energy market. We can’t tell you how many times healthcare organizations have asked us what we know about it.
In our last general AI report in 2019, we wanted to separate the truth from the noise. We wanted to know not only about the client experience but also about outcomes. So we created a market framework and evaluated the performance of early vendors.
The fact is that this year, AI is still an emerging market. It didn’t make sense for us to reevaluate vendor performance. But a lot of decisions are being made in this market. And so we into who is making those decisions, what decisions are being made, and why in our Healthcare AI 2020 report.
A Slowly Maturing AI Market
It is no surprise to us that the interest in incorporating AI in healthcare continues to exceed the available technology. However, the market is slowly but surely maturing.
We found that some organizations have just initiated pilot programs to see what they can accomplish. They have high hopes for their vendor partners, but they may not necessarily understand what they are getting themselves into.
Other organizations are a little further along. They know what they want, and they know what they can bring to the table. They are also realizing the time and energy they need to invest to realize outcomes. We wouldn’t say that the honeymoon period is quite over, but some pain points are starting to show.
Many healthcare AI companies are still largely dependent on their customers’ IT departments to embed AI models into their workflows. Then, once those workflows are embedded, organizations are also responsible for driving clinician adoption. Doctors and nurses have to trust that the data is accurate and will help them provide better patient and financial outcomes. That requires good internal education about the product as well as a good understanding of the why and how around predictive and prescriptive output. Even when adoption has been realized and the technology is working, the struggle is not over. These AI models are, of course, designed to bring to light performance gaps and other problems in the facility. Fixing those problems requires good change management and willing clinicians. All of these things tend to fall on clients’ shoulders. And many customers are not seeing an ROI come easily.
A few vendors are including a service wraparound to help clients with the arduous tasks of implementing the technology, driving clinician adoption, and even helping with change management. But these vendors are still in the early stages of developing a service arm, and many customers still feel that it is up to them to realize outcomes.
The changes in the market aren’t all negative. Some new vendors are starting to show up, such as ClosedLoop.ai. Excitement about using AI in healthcare is building. There are organizations that are starting to see outcomes. And although it isn’t surfacing as much due to the recent pandemic, AI is moving in the right direction.
Four Notable Vendors
As you peruse our latest report or look at the chart below, you may notice four vendors popping out at you: Epic, Jvion, ClosedLoop.ai, and KenSci.
You might say that the Epic AI platform is a catalyst for the market. In almost every call we have had about AI with Epic customers, its name comes up. Even though Epic’s AI customers are almost exclusively EMR customers as well, many other non-Epic organizations like to compare other AI vendors to Epic. The vendor’s platform is still a little immature, and they don’t have a robust service wraparound, but Epic’s customers can turn on prebuilt models or build their own and really delve into the world of machine learning. Of course, the Epic models also integrate well with the rest of the platform, and that is really attractive.
Jvion has been in the AI market for a long time — at least compared with many other AI vendors in healthcare. It had its start with some of the most well-respected health systems and has a large database of prebuilt models designed to solve real healthcare issues. With those kinds of references and technologies in an emerging market, the vendor is understandably a market leader. Jvion even has a service wraparound. Unfortunately, it seems like all of that is not quite enough to help clients operationalize and adopt their technology. Some customers struggle with the adoption stage. Other clients are failing to see the outcomes they desire and are losing their patience.
ClosedLoop.ai is a promising up-and-coming vendor that is starting to gain traction through its strategy of population health management, risk adjustment, and quality improvement. They offer a platform for customers to build their own models as well as a library of prebuilt models, and are healthcare specific. That is the best of both worlds for many customers. No wonder their performance scores and consideration rates are high for a new vendor.
Customers select KenSci for their expertise in both healthcare and AI, their partnership, and the flexibility they provide for organizations to build their own models. KenSci is still growing and lacks specific functionality that many potential customers need, but their current customers are optimistic. They are looking forward to getting new dashboards and expanding clinician adoption in their healthcare system.
AI still has a ways to go before it will grant us our wildest healthcare dreams. But we believe that with time and a little patience and persistence, dreams can become reality.
If you would like more information about what healthcare AI technology decisions organizations are making today, please check out our latest report.
This piece was written by Lois Krotz, Director of Research Strategy, Financial & Services, and Ryan Pretnik, Director of Strategy/Research & Segment Owner – I, BI, Payer Analytics and Quality Management, with KLAS.