A Glimpse at Vendor Performance in AI During COVID-19
During the COVID-19 pandemic, the progress of various AI initiatives has been hit or miss. The difference has frequently come down to how healthcare providers have wanted to implement AI tools. If organizations could see a way to use AI for problems related to COVID-19, they continued to be interested in and wanted to learn more about AI. However, other organizations that might have been less comfortable with AI and organizations doing custom work outside of needs related to COVID-19 put many projects on hold due to a lack of resources and cost constraints.
Currently, we are gearing up to publish a report on vendor performance in AI early next year. However, as we dig into the research we are collecting, we would like to share a few updates ahead of the report.
The Current Report Card
There have definitely been some shifts in vendor performance over the last year. As you can see from the current scores, some vendors have really improved, but we’re hearing from customers that other vendors are struggling.
Epic and Cerner
Many provider organizations are using Epic and Cerner for their AI projects because they are their EMR vendors, and the organizations tend to go to those vendors first with new projects. Many Epic customers are currently using the included or free models. Alternatively, some customers are also using Epic’s developer platform, which is a more advanced AI offering.
Cerner customers are a little more hesitant than Epic’s, but some are coming on board via Cerner’s managed model library. This library is one of Cerner’s two vastly different approaches with AI, and the other is HealtheDataLab. The main difference between the two is that if you are a healthcare provider using managed models, then you get more help from Cerner, and if you are using HealtheDataLab, you get little to no support.
From an overall performance perspective, both Epic and Cerner have some areas that they stand out in for doing a good job and areas where they can work to improve to increase their customers’ satisfaction and outcomes (which will be covered in the report) and provide case studies of how customers are using the vendors.
Consistent and Improving Performers
ClosedLoop.ai is a consistent performer in this space, and they are a dedicated healthcare player. Their customers can point to outcomes that they’re seeing. Health Catalyst, on the other hand, has a completely revamped approach on applying AI to data, and the purpose is to drive non-transactional outcomes. Some of the customers that used to rate Health Catalyst in the 60s now rate them in the 90s. So, if you look at the vendor’s performance simply by comparing the numbers, their new approach seems to be working.
Additional Vendors
The last vendor on the chart, Jvion, seems to be scoring lower at this point, but we’ll be digging more into their performance in the report that’s coming out early next year. There are also a few other vendors, Medial EarlySign and KenSci, that we will not include performance data on because their solutions are emerging, and it is too soon to tell how they are doing. However, those vendors will be included in the upcoming report with case studies of what their customers have been able to do with them.
The Ups and Downs
AI has seen its ups and downs during COVID-19, but KLAS is identifying use cases where providers are using AI to help solve clinical, operational, and financial challenges while increasing overall care quality. For a complete, deeper look at this space, keep an eye out for the report on the KLAS website.
Photo credit: Black Jack, Adobe Stock