Editor’s Note: This is the fifth in our series on Chief AI Officers in Healthcare. Other recent CAIO profiles include Dennis Chornenky at UC Davis Health, Dr. Karandeep Singh at UC San Diego Health, Alda Mizaku at Children’s National Hospital and Dr. Zafar Chaudry at Seattle Children’s.
Mouneer Odeh is only three months into his job as chief data and AI officer at the California’s Cedars-Sinai health system and he already has much to talk about. Odeh is part of a growing number of health IT executives graduating to the hot new healthcare C-suite position of Chief AI Officer.
Odeh sees a small snowball at the top of a mountain rolling closer and closer. That snowball is the steadily growing artificial intelligence work being done by health IT leaders and workers over time. And in five or 10 years, when the then giant snowball reaches the bottom of the hill and hits? AI will have transformed the entire healthcare industry, he says.
For all of that to happen, chief AI officers will need to take the lead.
Healthcare IT News sat down with Odeh to discuss his new role, and what kinds of AI projects Cedars-Sinai is working on today – particularly in nursing.
Q. How did Cedars-Sinai approach you to become its Chief AI officer? What were they looking for? And who would you report to?
A. Cedars approached me through a recruiter. It was the summer of 2024. I report to the CIO. I’ve always felt it’s important to have an embedded relationship with operational leaders. So, I am part of the office of the chief medical officer, as well. I sit as an embedded member of that team. But my direct reporting relationship is with the CIO.
In terms of what Cedars was looking for, it’s a very mission-driven organization, and it’s exceptional in that it embraces progress. Cedars recognizes our industry is at an inflection point as it relates to AI not just being an important driver of innovation but being increasingly an essential part of how we sustain our mission to provide the best care to the communities we serve. AI really is an essential enabler of our mission, and that’s new and exciting.
What they we’re really looking for is for me to come in and help orchestrate our strategy across our clinical, operational and research areas. It’s a pretty broad mandate. It’s a very exciting time to be a part of this journey.
Q. What in your background makes you a good fit to be a Chief AI Officer? And what skills should anyone looking to become a Chief AI Officer have?
A. It’s an interesting question; there are so many different flavors. This is really very personal to me, how I see myself fitting into the role. The first and most important is really transformational leadership. Our role is not just about delivering exciting, cool technology. It really is how AI can be a catalyst for this transformation.
I’ve been in the healthcare provider space for the last 10 years at Jefferson Health and then Inova Health System. Transformation, what it meant, I call it Transformation 1.0. What it meant at that time was changing the operating model.
How do you pull together different hospitals and different practices and reshape that into an integrated healthcare delivery system? And then how do you build new leadership structures and new processes, and drive priorities and cascade accountability across the new health system?
When I look at the next 10 years, I think the real transformation ahead is all about workflow transformation. It’s how do we rethink, reimagine, re-engineer. Whether it’s how we care for our patients, how we run our back-end operations, how we accelerate research and discovery – all of that really gets into some very nitty-gritty, like understanding workflows and how technology can be a catalyst for that transformation.
So that’s one of the things that’s most exciting to me – being a part of that transformation. So, I would say transformational leadership is really critical to success in this space.
The second thing I’d say is it’s all about execution, right?
You can have a transformational vision, but you’ve got to deliver value. And so, really focus in on the impact side. And again, success is not measured by the projects we deliver. It’s measured by the impact we’re having.
So it’s important to be able to deliver those projects, but it’s much more important to make sure we keep sight of what are the goals, the objectives we’re trying to achieve, and then demonstrating we have achieved or surpassed those goals. So that we can really demonstrate the progress and build momentum toward more and more of this change.
I think of our job as not one big bang but a snowball effect where things accumulate. And if you think about a snowball, it starts really small and stays pretty small for some time. But after a while, it gets a critical mass and each new rotation is just so much bigger and it’s an exponential function.
That’s very much what we need to do with AI over the next five to 10 years – continue to build and build until the scale of the impact is what we need to have. And to do that, we’ve got to build a track record of success. We’ve got to be able to demonstrate that, sell the value story to the organization.
The third thing is you do need to understand the technology. And you’ve got to understand it at a pretty basic level – what are the strengths and what are the limitations, so when you’re leveraging the technology you’re not trying to solve the wrong problem with it.
Q. Please describe the AI part of your job at Cedars-Sinai in broad terms, what is expected of you? And in more specific terms, what is one of your days like?
A. There are a few different flavors. Sometimes it’s about evaluating vendors versus internal capabilities. But with AI, there’s kind of a special aspect of this, which is that the statistical performance of those models really should drive how you integrate them into the workflow.
The analogy I would use is a screening test versus a diagnostic test. They have very different statistical performance characteristics and they have different clinical meaning.
What doctors will do with that information is very different. So, for example, a screening test is really designed to make sure you can cast that net as wide as possible and try not to miss anybody. So, you’re trying to avoid or minimize the false negatives. Make sure you get everybody in that you think might have a risk.
A diagnostic test is very different. It’s really about trying to get a very clear diagnosis with the intent to treat. And therefore, you need to be really clear if somebody has a condition or not, because you’re going to make treatment decisions based on that. In that case, you’re trying to minimize the false positives.
So those are two very different statistical performance characteristics.
We give names to those in the diagnostic space. In the AI world, we have to develop that type of a language so we can think very carefully and appropriately about how we integrate it into the workflows. And then we can start to answer the questions of who gets what information, when, what should they do with that information. And explain a little bit about the why, and how they are performing those actions.
And the last thing that I would say is really critical is the ability to partner. You know, this is a complex function, and we’ve got to be able to bring people with various different perspectives, multidisciplinary talents, to come together to try to solve these problems.
And that requires the ability to influence across reporting relationships, to build thought partnerships, and to embed yourself into the minds of the customers we’re trying to support with our solutions.
As for my day, each day is so different. Also, I’m only three months into this at Cedars, so I think the answer could change two years into it. But I think it’s an interesting sort of glimpse at this moment in time.
The broad responsibilities are to build and acquire then ultimately align our AI capabilities to the priorities of the organization.
So, for me, understanding what our priorities are is absolutely essential. And because I’m new, that’s a big part of what I’m really trying to do, is make sure I understand the waters I swim in so the solutions we provide are as relevant as possible and as impactful as possible.
Further, one thing that’s really awesome. Cedars-Sinai has an amazing research program. And a big part of our research program is using real-world data to drive innovations and discovery, much of that through computational biomedicine.
So, we have a leading computational biomedicine program. A big part of my job is to understand how we can accelerate that research. And then how do we take those research breakthroughs and apply them into clinical practice – from the research bench to the bedside.
And that translational function is a gap that a lot of academic medical centers will struggle with. And we’re really committed to how we can close that gap and take those ideas and try to bring them into clinical practice, learn from that, take it back to research and continue to iterate.
Sometimes people describe that as a rapid learning health system. I think with the power we have with our research community we can do some pretty amazing things here.
Additionally, AI is a very powerful tool, but it can be scary a little bit, too. So, we’ve got to have governance and accountability mechanisms that give us the confidence we are deploying AI in a responsible way. A big part of my job is trying to understand, assess and mature our governance and stewardship capabilities.
This is a new space more generally in our industry and across all industries and we’re excited to be a part of that. And I give a lot of focus to try to make sure if you don’t have that trust, then that snowball will never take off. Right? Because the trust is like the hidden fuel behind gaining momentum.
And then, a lot of days, just real nitty-gritty stuff. What is the problem we’re trying to solve? Getting very clear and specific about that. And the clearer we are about that in the beginning, the more likely we are to deliver something that hits the target. So just clarifying what it is we’re trying to solve is a big part of what I do.
Also, and this is one of the hardest things, and one of the most interesting and fun things, evaluating various different technologies. We have our own internal capabilities, we have our gaps, we’re building and closing those gaps with new skills and acquiring new technologies. But we have a lot of vendors introducing great technologies, new features, and we have tons of startups.
That is one of the most active areas of investment right now – healthcare technology startups. And almost everybody is doing AI at this point.
So, evaluating those technologies is one of the most interesting, exciting, but also challenging things because we know that most of those startups will fail. So we’ve got to be very discerning with how we evaluate capabilities and evaluate the fit with alignment to our priorities and likelihood of success.
An increasing part over the next six months for me will be how we democratize access to our AI tools. We believe that not all of this is going to come out of a centralized sort of competency. But the more we spread this capability, the more we empower people to solve their own problems using AI with the appropriate guardrails and safety mechanisms in place. Then we can turbocharge that snowball effect.
We have examples of genAI systems. We have a prompt-a-thon coming up in May where we’re bringing teams from across various different operational departments to come in for like a hackathon where they’re going to develop a prototype, hopefully an actual system.
We’ve got more than 600 people who have been trained on some of these prompt engineering capabilities. So we’re looking to see how we can continue to encourage and democratize access to these tools.
Q. Please discuss one particular AI project you are proud of that you’re working on. I know it’s only been three months, but something you’re working on that’s moving along well, and what you’re seeing.
A. We have so many different use cases going on – AI is going to accelerate our research and discovery. It’s going to improve our patient outcomes. It’s going to drive operational efficiencies.
But one of the areas I think is maybe not as well appreciated just yet is how it will transform the experience our patients have and our team members have. And in many ways, I think it’s huge. Humanizing.
Technology has almost been a tyranny. We’ve all become slaves to our ERP systems, our EHR systems, and things like that. It creates burden. AI is transforming that relationship. It is through those large language models. And things like ambient technologies doctors are using.
What I’ll hone in and address is our nursing burden. Nurses are really overwhelmed by the amount of documentation they need to do. And the documentation is important because it’s communicating to somebody else who needs to see that information so we can provide the best care for that patient.
But it is tremendously burdensome because you have to enter that information into an electronic health record system that is trying to force you to think in discrete data elements, whereas people don’t think that way in natural language.
So with these large language models being applied to nursing, we are starting to see very substantive savings in the amount of time it takes them to document. Burnout is coming down.
One thing that is somewhat surprising but makes sense when you think about it, is our patient experience scores are dramatically improving as a result of that, because our nurses aren’t spending as much time capturing the information on a keyboard and they’re spending more time in interaction face to face. And it’s being documented in a much more natural and intuitive way. And the feedback we hear from nurses is truly inspiring.
This is saving them an hour-and-a-half to two hours a day. They love it. It brings the joy back into their work. And I think that’s a really great example of just how this technology is transforming the way we interact with technology more generally. It’s humanizing it at a completely different level, and it’s alleviating some of the burdens.
One of the big challenges we have is how people are getting burnt out. And technology is a contributor to that. And I think now with AI, technology can be a great solution to that problem.
For a 5-minute video of bonus content not found in this story, click here. In the video, Mouneer Odeh shares tips for IT executives looking to become a Chief AI Officer for a hospital or health system, and offers his views on the best ways a Chief AI Officer can work together with his or her peers in the healthcare C-suite to ensure AI gets done right.
Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.
WATCH NOW: Grabbing the Chief AI Officer brass ring (and working with top brass)