Hackensack Meridian Health is one of the largest integrated healthcare networks in New Jersey, with18 hospitals and more than 500 patient care locations, including ambulatory care centers, surgery centers, home health services, ambulance services, air medical transportation, rehabilitation centers, urgent care centers, physician practice locations, and a fitness and wellness center. The network has more than 35,000 team members and 7,000 physicians.
Sameer Sethi, Hackensack Meridian Health SVP and chief AI and insights officer, spoke with MobiHealthNews about how much of its 2025 budget will be allocated for AI adoption?
MobiHealthNews: Do you have plans on spending a percentage of your 2025 budget on AI?
Sameer Sethi: We have been allocating an amount of budget to AI for quite a few years, at least for the two and a half years that I have been here. AI has been a separate cost center for me, which I manage.
The evolution of that is what we call AI 2.0, because we have run through AI 1.0 in the past two and a half years that I have been here. When I came here, I helped develop an AI strategy, and we have learned and evolved from that. As a result of that, we have now developed our AI 2.0 strategy. That costs money.
Without disclosing numbers, we have a higher allocation of budget. I started here as a data and analytics officer. I am now the chief AI officer. That is an investment and a commitment from the organization to take this further. So, we will be doing more with AI.
My role here is AI but also RPA, robotics processing and automation.
When you fuse that together we call that IPA, which is intelligence process automation.
We take the concepts of AI and the concepts of automation and we put it together to best use.
The evolution in addition to additional costs and, therefore, the use cases, are dependent on a lot of the successes. The successes being that in the last few years we have built a solid foundation that we can build AI on top of.
We have hired the people, we trained them. We continue to hire and train them. And we have built a lot of use cases and have learned from that exercise and experience of what works and what does not work.
MHN: What are your main focus areas regarding AI?
Sethi: At first, we had three buckets: We will do things to bring in operational efficiency, we will build things to help decision support, and the third is we will build things that will build patient experience. From that we evolved. Now we have six focus areas, or buckets, for AI and automation.
We will build AI to create personalized and equitable experiences, we will create AI to streamline administrative and clinical efficiencies, we will develop AI for capacity management and burnout alleviation, we will build AI for disease detection, and we will build AI for precision treatment and research and innovation.
MHN: What are positive contributions of AI in healthcare, and what are some of your concerns related to AI?
Sethi: AI is controversial. AI becomes controversial when people start to think of AI as one size fits all. It's not. It is similar to any other software. All this has to be fine-tuned and built to purpose.
If you build AI to purpose then that controversy starts to go away. Part of the controversy around AI is: Is it wrong? Is this correct? That is one piece of this. There is another side of the controversy around what it does to the workforce. Does it take jobs away?
AI is not going to take jobs away. It is going to take jobs away from people who do not use AI.
When I'm asked this question, AI is not going to replace physicians, but AI will replace physicians that do not use AI. What we have seen is that AI is helping physicians.
For example, when you go into a physician's office, the physician generally does not look at you eye-to-eye and is actually staring at the screen and reading something. What they are reading is your clinical information, which to some degree they should have time to do prior to coming into the room.
There are a lot of pressures around seeing more patients every day, and these physicians don't get time to read the hundreds of lines of information that are available in a patient's EMR or clinical record.
What we do with AI is we summarize that information for the clinician.
Today, information is coming from your watch, your phone and heart rate monitors, and from five other doctors that you saw or the hundred doctors that you saw since you have been alive.
All that data is a lot of data. It takes a lot of time for humans to process that data. That is where AI comes in. AI is helping them, it is not replacing them.
AI is decision support. AI is helping a physician make a decision amongst all other noises that they have to deal with. Physicians are trained to do this; we are just making their jobs better.
To me, the controversy is coming around understanding what AI is and how it should be used.
If you educate people and train the models, and you train the end users, as you would with any other software, then the controversy for the most part goes away.
We, as an organization, take AI governance very seriously. And when I say AI governance, it's not just building what needs to be built, but it is also the safety. We look at 108 different ways of what AI should and should not be. We have 13 domains that result in our 108-point check. That determines and looks at bias safety, security and privacy. We evaluate all of these things to remove more of that controversy. With these three things AI starts to become less controversial and very safe.
MHN: Does AI have a bright future in hospitals and health systems?
Sethi: I am personally committed and very optimistic that AI is going to make things better.
I also want to point out that there is an element of cost. There are opportunities in healthcare. The cost of healthcare is tremendous.
Our elderly population is increasing. Us, as a country overall, are not getting a whole lot healthier, and we require care, between the expensive cost of healthcare, plus the aging population, plus we are being more careful when diseases come around. We need help, and AI can help with those things. It doesn't solve those problems 100%. It doesn't become a doctor, but it can definitely take away certain things that today cost us a lot.
So, I think that the avenues by which it is coming at us, unlike the internet, is really fast. We as a society need to spend more time educating ourselves, trying it out and giving it a chance. And if we do that, I think things will go a lot easier.