For skilled care providers who need to save time due to talent shortages, generative AI could become a more important part of caregiving.
WellSky, a software analytics company serving healthcare providers, announced a partnership with Google Cloud late last week to develop advanced data analytics tools, machine learning capabilities, and artificial intelligence in healthcare settings. Joel Dolisy, WellSky’s chief technology officer, told Skilled Nursing News that the move is aimed at addressing burnout in the industry and increasing clinicians’ time at the bedside.
Skilled Nursing News sat down with Dolisy to discuss the specific goals WellSky is seeking to achieve through its partnership with Google Cloud and integration of generative AI tools known as Vertex AI. He also discussed the potential impact of his organization’s technology on post-acute care and skilled nursing departments.
The transcript below has been edited for length and clarity.
SNN: Can you give an example of how generative AI tools are specifically applied in skilled nursing settings?
Dolisy: Last summer, we launched a number of different We visited all the experts in the field. We talked to almost every post-acute end market, including home health, hospice, skilled nursing, and long-term care. What has become clear is that there are so many similarities when you have a very sophisticated patient journey from receiving referrals to pre-admission to admission, and then from visits and encounters to discharge. I was told that there is. It became clear that each of these steps in the patient journey between different end markets had common issues that applied to all of those end markets. Some of them are, for example, when receiving an incoming referral, [Continuity Care Document] For example, a CCD has 100 pages of clinical documentation that someone has to look at right now and decipher whether it’s a list of problems, allergies, or a list of medications they’re currently taking. It won’t. This is one of the key applications that generative AI excels at, for example. Good at summarizing and extracting information from documents. And rather than imagining things, the model actually pulls out important information and allows you to provide references to documents where the facts are found. This is very important so that you can use the technology’s summarization capabilities while ensuring that your model is actually rooted in fact and not just in your imagination.
SNN: What are some other examples where generative AI can be useful in SNF?
Dolisy: We’ve talked to long-term care facilities, skilled nursing facilities, and experts in the field on discharge, and generative AI does a very meaningful but short summary discharge before the patient is discharged. may be helpful. And that could mean the difference between the two. Immediate admission of the patient to another home health or other care facility. Again, we have access to all patient information as part of the patient record maintained within our system. It actually summarizes that information based on the last few visits and the last assessment done, creating a relevant summary view of the patient’s status that helps ensure that patients receive care when they are discharged. Wouldn’t it be great if you could do it? Will they actually be accepted into another nursing care setting right away?
SNN: Can you talk about the steps taken to protect patient information?
Dolisy: We are trusted to process data on behalf of the healthcare providers and patients we serve. That’s why we use this responsible AI framework that works closely with Google Cloud. This is because Google Cloud adheres to the same philosophy regarding separation of data from technology and models. We really want to ensure that level of thoroughness, enforce privacy and security, and provide transparency in how we arrive at the results we deliver. This is key to keeping the user at the center of everything.
SNN: Will AI completely eliminate the need for clinician judgment?
Drissy: Our goal now is not to replace humans. I don’t think we’ll ever get there, at least not while I’m working, and I’m probably wrong about that. But for now, we want clinicians to be informed as well. They’re in the middle of making decisions, and they’re taking what’s suggested by the AI engine as a way to pre-fill, auto-complete to some extent, in a way that’s much smarter than what they can currently do around it. But they always have the option of actually looking at it and saying, “Oh, I believe that,” or “No, that doesn’t really match what I heard during this encounter, for example.” Edit it and then put it back. Let’s. ” For example, when it comes to ratings, we haven’t gotten to the point of fully autofilling them, we just say, “Yes, that’s fine,” without doing any reviews at all.
At WellSky, we have established a responsible AI governance committee to ensure that every AI project passes a series of reviews, and that the various features planned for the product actually pass through a series of gates and filters. Make sure you provide transparency and ensure certainty. Fair characteristics. Therefore, a series of deliberate steps are required for WellSky to address this issue in a highly responsible manner for patients and users.
SNN: You’ve said that there is “no shortage of possibilities” in the emerging field of AI. Can you provide insight into potential future developments?
Dolisy: Currently, our focus is clinical, but there is another area that is very interesting. It’s the back-office, medical coding side of the revenue cycle management side, and a lot of the solutions now are, to some extent, using what used to be the cutting edge of AI technology called RPA (robotic process automation). I am. I believe that many of these current capabilities will be enhanced in the future by generative AI adaptive capabilities, increasing some of the accuracy of automated coding and minimizing the amount of human review required. The key is to really think that this is really a clinician-centered approach, a patient-centered approach…That’s the key thing that we want to make sure our employees and our clients understand. We’re not there to replace them, and we’re trying to get them to spend more time doing what they signed up to do in medicine: helping patients.
