Yasir Tarabichi, MD, is in chief of health SubwayThe public health system based in Cleveland. He is also CMIO in Cleveland Ovatwhich provides coordination of care for primary care, urgent care and behavioral care using a unified technological platform. Dr. Tarabichi sat with Health care innovation Editor -in -chief Mark Hagland during Vive25, which takes place this week at the Nashville Music City Center, to discuss the real state of the adoption of artificial intelligence in patient care organizations at the moment. You will find below extracts from this interview.
After a long period of media threshing and high expectations, where are the managers of patient care organizations currently in terms of improving AI?
It depends on where you are as an organization on the innovation curve. The organizations that jumped forward spent a lot of time, energy and money, the determinant, and probably helped everyone save time this way. My role at Metrohealth is to identify opportunities and to strategically guide the organization so that we do not waste resources, and for us to invest, buying, resources, which work for us. So what is the real value proposition or the king [return on investment]? Sometimes the king is who makes your clinicians better adjusted. And it’s great, but the organization could say, it’s good, can you see more patients?
And in the reimbursement environment, we must think carefully in terms of return on investment. I cochair the Metrohealth AI advisory committee, with a trading partner, like Dyade. We pollinate crossed. So I’m talking about the risk of a clinical point of view; He reminds me of operational problems, it could hurt us financially, which could hurt us strategically. The risks are therefore parallel to the clinic, but different. So we want to see what exists and understanding what we solve can be a little better informed rather than trying something novo. We have to choose solutions that pollute all these objectives.
What are the initiatives you are working on at the moment?
We have made a lot of predictive analysis in the clinical space. We have built models and assessed them. We want to do it fairly. Here is an example: a common problem is access to care in clinics, and a current problem is that systems overbooked patients, which is honestly a terrible idea. So, in a zero -sum system, those who are already late are the most ready to lose. As soon as you say, this person may not introduce himself – and he could be a person of color, disadvantaged, etc. – And then what do they get if they arise? They have a terrible patient experience: they are upset, the clinician is upset.
I would post that dual book patients for clinical meetings are a very bad solution to a problem, as it worsens disparities. We are a community security network system, and we think that if you make an appointment, this appointment belongs to you. And we have all of these phone calls, SMS, patient portal messages to patients, but some patients just don’t answer. So what can we do? Call them. It turns out that there is a segment of the population, mainly black, which has a high rate of non-presentation. So if we two meetings, it is this group of patients who will tend to be disadvantaged. But they pick up the phone if we call them.
Consequently, we have implemented a solution with a standardized route, associated with telephone calls. And in doing so, we have reduced the non-presentation rate in the African-American community by 15%.
In other words, you have twinned an analysis of the data facilitated by AI with relatively low technology action – telephone calls.
Yes, that’s right: the question is how technology works in the real world, with our patients in the field? And we can predict anything, but what does that mean? That doesn’t tell me what I have to do. The solution is not technology. At this point, we have finished being in love and excited by technology; We have to make it work. It is a high -tech approach and in Touch.
How would you characterize this moment in terms of AI AI adoption and generative development?
I am probably less excited by the place where large language models landed today; They stagnated. What I can say is that what the generator is the best for what is ambient listening, and the other, a recovery of information increased from a occupied and terrible DSE [electronic health record]. An example on the Ovatient side is the way we have managed the use of antibiotics. The classic situation is when a patient comes to see a doctor with a potential urinary tract infection, and that the doctor orders a prescription for an antibiotic, but said to the patient: “OK, I ordered a prescription for an antibiotic, but Wait for your US test is positive to take the antibiotic, ok? Well, what does the patient do? They automatically start to take the antibiotic. But with a generative AI, as a doctor, I can detect the interaction, based on predictive analysis, which can predict whether the symptoms of a patient correspond to the UTI, before the tests.
What will happen in the coming years, especially around generator AI?
The technology will become cheaper and more accessible, and the next step will be to ask why we use it. I therefore think that if you have swept all the information from the DSE and understood best practices and protocols, now, given the knowledge base this area. And the generative players of the AI will hit this door. And if you can install an agentic AI in a patient portal, the portal In a portal with an agentic AI, and he can make an appointment with you, he creates a arms race with DSE sellers trying to have a better experience.
An agent could reformat and do things faster for you; He will organize the experience for my taste, I look forward to this and that patients are more independent. And I also think a lot about access. Access to health care navigation is difficult and it fears. And unless a patient has a full -time coordinator waiting for satee, to his teams, by helping them at each stage – this coordination is another opportunity. But the agent IA will have to understand the system. However, we must also repair the broken health care provision system.
