Cedars-Sinai CIO Craig Kwiatkowski Talks Overhauling ERP, AI Governance & Immersive Learning

Craig Kwiatkowski, SVP & CIO, Cedars-Sinai

With so many new technology solutions hitting the market, it’s becoming increasingly common for health systems to add accelerator or incubator programs to separate the wheat from the chaff. Because although there are many brilliant ideas, there’s often a lack of understanding when it comes to workflows and other challenges, according to Craig Kwiatkowski.

The Cedars-Sinai Accelerator, established in 2022, aims to address this “knowledge gap” by granting access to end users, stakeholders, and thought leaders throughout the organization so they witness firsthand how care is delivered. It’s one reason why the reason the program has already counted several success stories, he said. Another is the fact that the Accelerator is extremely selective, focusing only on products that “scratch an itch or solve a problem.”

During a recent interview, Kwiatkowski spoke with Kate Gamble, Managing Editor and Director of Social Media, about the many initiatives his team has in place to improve efficiency and quality for providers and patients at Cedar-Sinai, an academic organization serving more than 1 million individuals across the Los Angeles community. He shared insights on the “major overhaul” of ERP systems that will help centralize services; the three-pillar AI governance structure his team has created; and the unique experience he gained during his time as a pharmacist.

LISTEN HERE USING THE PLAYER BELOW OR SUBSCRIBE THROUGH YOUR FAVORITE PODCASTING SERVICE.

Podcast: Play in new window | Download (Duration: 30:45 — 18.6MB)

Bold Statements

Our focus is on reducing friction, improving efficiency, and simplifying things where we can, and there is certainly no shortage of opportunities and possibilities to do just that.

We’ve spent the better part of the last 10 to 15 years implementing EMR and EHR technology and really focusing on the clinical solutions and tools that are so fundamental to the work that’s being done. I think that has come at the expense, in many cases, of focusing on ERP and other administrative or back-office solutions.

We try not to create bespoke technology solutions that the companies are going to have a challenging time deploying and scaling outside of Cedars. We’re helping think about what will work broadly and be forward thinking as they’re deploying their products moving forward.

Innovation is often easier than adoption. I think that’s particularly true in healthcare.

Safety in particular is something I’ve always been very passionate about. I think it connects very well to the way in which we try to serve the organization from a technology standpoint and the importance of the work we do.

Q&A with Craig Kwiatkowski, SVP & CIO, Cedars-Sinai

Gamble: Thanks so much for taking some time to speak. I appreciate it. I want to talk about your core objectives, particularly in terms of driving innovation. Let’s start with a high-level overview of Cedars-Sinai. Can you talk about where you’re located, what you have in terms of hospitals, things like that?

Kwiatkowski: Sure. That’s a good place to start. Cedars-Sinai Medical Center is a non-profit academic medical center with about 900 beds. We provide a wide range of services in the Los Angeles area with a number of different specialty programs, including cardiology, ortho, neuro, GI, cancer, and women’s health, many of which are highly regarded.

In addition to the main campus, Cedars-Sinai Health System includes Marina del Rey Hospital, which has about 130 beds; Huntington Health in Pasadena, with about 600 beds; and Torrance Memorial in the greater SoCal region, which has about 400 beds. So that’s a little more than 2,000 beds in all. Beyond the acute care facilities, we have a significant ambulatory footprint with dozens of primary and specialty care and urgent care clinics across LA and out into the valley areas.

All about people

Gamble: So certainly a lot going on. What do you consider to be your core objectives right now?

Kwiatkowski: Our core objectives continue to be centered around people — as patients, as caregivers, and staff. And so, we’re looking for ways to improve efficiency and quality across all steps of the patient care experience so we can reduce friction and make things easier. For example, we have a lot of work planned and in progress looking at patient access and virtual tools. We’re looking at ways to make it easier for patients to self-schedule and receive care.

We’re asking questions like, how can we make things better for our staff to alleviate workplace pressures and do that via technology and tools or enabling solutions? There are a number of things we’re working on and looking to explore in the near-term, including ambient listening and documentation tools — which is now called augmented response technology — for drafting notes and in-basket responses for physicians and summarizing patient charts. Those are some of the ways we’re using AI and generative tools.

We’re also pursuing virtual sitter and virtual nursing capabilities, and we’re embarking on a home care program which will go live in the coming months. We have a number of initiatives related to capacity and patient flow, which continue to be a challenge for us — we’re bursting at the seams most days of the week in terms of capacity. That comes on the heels of our ERP platform deployment, which went live earlier this year. We’re looking to optimize those tools to deploy additional functionality and automation so that we can take advantage of that tool set and begin to see some efficiency in those areas. Again, our focus is on reducing friction, improving efficiency, and simplifying things where we can, and there is certainly no shortage of opportunities and possibilities to do just that.

Improving efficiency ‘in real time’

Gamble: Very interesting. When you look at things like ambient listening, it’s starting to take off now, but it has taken a while. What do you think has been the key to getting that off the ground?

Kwiatkowski: I think it’s the evolution of the technology; we have better capabilities now. When you look at in terms of the virtual scribes or ambient listening solutions that existed prior to the explosion of generative AI, they weren’t as nimble and agile in real time. The physician would use the tools to help create the note, but there was often a long lag time before the note actually made its way into the patient’s chart. Sometimes because it required a human to actually do some QA and auditing before the note actually was finalized.

Now, with generative tools, that happens much more in real time. It can happen within seconds to a minute that a note is generated and lands in the patient’s chart for review and sign off by the physician. I think what has made it more appealing is the advancement of technology, and quite frankly, the reduction in the price point for some of those solutions, which are fairly expensive even now.

Centralized admissions

Gamble: You mentioned virtual sitting, which you’re doing with HomeCare. Can you talk more about that?

Kwiatkowski: Yeah, we’re looking at that for virtual sitters and virtual nursing, which uses in-room camera technology to monitor the patient. In doing so, it allows an individual to monitor multiple patients at the same time. In addition to make sure patients are not falling or at risk for fall, there’s AI technology embedded in some of those solutions as well to begin to predict or detect things as they’re happening in the room. We’re looking at the possibility of doing more centralized admissions and discharge, and being able to build upon some of the efficiencies we think we can gain via the centralization of those solutions.

From a home care standpoint, we’re exploring what that would look like and working on some technology with an external vendor to begin to build that program and be able to decant some of the capacity from within the acute care space.

We think it’s going to be a very slow uptake within our home care environment. But we’re hopeful that over the next few years, it will scale and have a material impact on capacity while also being a satisfier for our patient population, who would prefer to be at home. And so, hopefully it’s a win-win from both of those standpoints.

A “major overhaul” with ERP

Gamble: Interesting. You also mentioned the ERP deployment which recently went live. Can you talk about that?

Kwiatkowski: Yes, we went live with finance, supply chain, HR, payroll, grants, and funds, as well as scheduling and timekeeping, which are ERP-adjacent. We went live in sort of a phased way from the scheduling and timekeeping standpoints in late November, early December. We went live with the official cut-over with payroll, finance, and supply chain around the first of the year. It’s a major overhaul for the organization moving from a number of disparate legacy applications to a central platform and portfolio of solutions. We’re pretty excited about that — it went amazingly well. There are lots of things that still need to be optimized and addressed. But we’re on a great path to move things forward.

It was a really nice win for the organization; more of an operational transformational project that included a number of business process changes and other efficiency opportunities. And of course, technology is the enabling piece that sits underneath that. We’re excited about it and we look forward to determining what the roadmap looks like and actually starting to expand some of those tools to the broader healthcare system.

We went live at our main campus, in our Marina del Rey Hospital, and in our medical network, and we’re looking to expand that further.

“Dramatic change” in staff satisfaction

Gamble: That seems like something that can be a big satisfier to the staff.

Kwiatkowski: Absolutely. It’s been a big satisfier already. We took such a leap forward in terms of technology. This is genuinely a contemporary technology solution; it has mobile and other capabilities that didn’t exist before. The ability to see your schedule and look for available shifts and to make changes to your time card — and do all of that in a mobile way, rather than on triplicate pieces of paper or spreadsheets — is a pretty dramatic change from where we were. It’s definitely been a satisfier so far.

“Critically important” systems

Gamble: It’s interesting; ERP isn’t exactly a buzzword, but it can make such a big impact, as you’ve seen.

Kwiatkowski: It definitely tends not to be one of the sexier topics. As healthcare organizations, we’ve spent the better part of the last 10 to 15 years implementing EMR and EHR technology and really focusing on the clinical solutions and tools that are so fundamental to the work that’s being done. I think that has come at the expense, in many cases, of focusing on ERP and other administrative or back-office solutions, which are critically important to running a large, complicated organization and health system. It’s something that we’ve certainly made a priority and we want to continue to make it a priority to take advantage of all those capabilities.

Cedars-Sinai Accelerator

Gamble: That’s a smart strategy. Let’s talk about the Cedars-Sinai Accelerator, which is a unique program you have in place. Can you tell us a bit about how it’s structured and what are the objectives?

Kwiatkowski: Sure. The Accelerator is focused on supporting the growth and development of early-stage companies with health and healthcare delivery solutions. Every year the Accelerator runs a 3-month program that essentially opens the doors of Cedars-Sinai to the expertise that exists within, giving companies access to mentors, end users, stakeholders, decision makers and thought leaders from around the organization. It’s multi-faceted.

Many companies spend a fair amount of time at the hospital and in the clinics learning and meeting with key staff to understand workflows and understand how healthcare delivery happens, which is oftentimes a gap of some of these early-stage companies.

Thinking ahead

Our IT teams spend a ton of time with the companies as well, helping them to understand and achieve their IT integration goals — essentially how to hook up the plumbing, so to speak, from an infrastructure and a technology standpoint. It also helps them understand the complexities: data use, privacy, cybersecurity and really learning what healthcare organization expect and need from them.

As an example, we try not to create bespoke technology solutions that the companies are going to have a challenging time deploying and scaling outside of Cedars. We’re helping think about what will work broadly and be forward thinking as they’re deploying their products moving forward.

There are also lectures and workshops led by various leaders from around the organization on a variety of topics related to healthcare. That speaks to the educational and didactic aspects of the program.

Demo Day

At the end of it there’s a demo day, as we call it, during which the Accelerator class companies present their solutions to investors, to potential customers, to mentors, and the press actually joins as well. Our next class is coming up in August of this year. There are usually several hundred applications for about 10 to 12 class spots. The applicant pool comes from a variety of areas beyond just digital health and technology and medical devices although that tends to be the bulk of the companies that are involved. The Accelerator also accepts companies with solutions in biotech, pharmaceutical products, molecular development, administrative functions, and some of the things I mentioned earlier around finance, human resources, supply chain, and so forth. It’s a really great program that has helped a lot of companies advance their projects and go on to great success.

Value of “immersive learning”

Gamble: I think there’s so much value in being able to come in and learn firsthand about workflow and get a better understanding of what clinicians and staff actually need.

Kwiatkowski: Indeed. That’s part of the immersive learning aspect of it — watching things happen in real time. It goes well beyond the abstract to the literal and sort of in-your-face aspects of the way in which healthcare delivery actually happens and the practice of medicine, depending on what the company is focused on. Some of those pieces really do help people understand how to best position their product, how to best enhance their product, and which frictions can be added or reduced based on the choices that are made.

Success stories

Gamble: That’s really interesting. When these companies come in and do presentations, I would think that can also shape the health system’s viewpoint too as far as what you want in a solution.

Kwiatkowski: It can. We’ve had about 80 companies come through the program over the last several years. There’s been quite a number of solutions that we’ve actually deployed here at Cedars, and many that have gone on to success well beyond our organization. One example is Deep 6, a company that used AI to comb through medical records to identify patients for clinical trials.

There’s also Artera, formerly known as Well Health, which created an SMS texting solution that’s used widely in the industry and is integrated with our patient portal. Avia is a healthcare derivative of Amazon Alexa which we have in patient rooms throughout the hospital to help patients with hands-free access to the care team. Health Note uses an AI chatbot to answer patient questions prior to a visit and pre-populate notes for physicians. And then there’s Moxi, a robot developed by Diligent Robotics that assists with tasks like pickup and delivery of meds, blood, supplies, and specimens so that staff can spent more time on higher-valued tasks as well as patient care.

Scratching an itch

There are a few more advancements in the AI space. Syntho has a product that uses AI to generate synthetic data sets, allowing for research and data sharing without some of the usual privacy concerns. HealthLeap is a newer company that uses AI and machine learning to identify patients who are malnourished or at risk, and cue those up for interventions and adjustments to their care plan. Finally there’s Acolyte, which uses generative AI to create physician Avatar messages to help improve compliance with health maintenance and product follow-ups.

All of these are examples of things that were sourced from an organizational standpoint to solve a problem we had. They weren’t solutions in search of a problem. They were sourced and identified as candidates for the Accelerator because they scratch an itch or solve something that needed to be addressed within the organization. They’ve done that, and many of them have gone on to great success.

Cedars-Sinai’s AI council

Gamble: I’m definitely familiar with some of those names. There’s so much going on in the AI space. I saw that Cedars-Sinai has established an AI council to manage these needs. Can you talk about how that’s structured and how it helps with prioritization and other challenges, especially with the explosion of generative AI?

Kwiatkowski: That could be a long conversation in and of itself, but it certainly has evolved quite a bit over the last year, especially with generative AI. We’ve been using AI for many years across Cedars to support decision making with rules-based AI and machine learning and predictive analytics. From that standpoint, AI is not new to us. It’s the emphasis on the generative tools, and the imagination that invites, that has spurred a lot of conversation across the world — not just in healthcare.

In terms of our approach, we’ve got a three-pillar strategy that came out of our AI Council, which was founded in 2022. It’s a multi-disciplinary group comprised of leaders from around the organization, many of whom have worked in AI and advanced analytics, including folks who are in our computational biomedicine group. The council has a number of functions and responsibilities: maintaining priorities, identifying opportunities, policy and regulatory oversight, education and communication, and assuring we’re properly vetting and monitoring AI use. It’s a lot to stay on top off, and so, we’ve expanded the council to form some focused workgroups in a number of areas.

First pillar of AI: Investment and planning

Importantly, the first task for the council was establishing our AI strategy, which is centered around three pillars. The first pillar is around investing and planning. Like most organizations, we believe we can do almost anything, but we know we can’t do everything. This pillar is intended to reflect that and to emphasize our ongoing needs to plan and invest resources and time wisely, essentially making sure that we’re spending our time, attention, and focus on the right things to support our patient, our caregivers, and staff.

Generative AI Idea-thons

One of the ways we’ve done that was around priority gathering. In the early days of the generative AI rage, we began conducting idea-thons where we gathered frontline physicians, caregivers, and staff from around the organization so we could learn from those closest to the work how these tools might help them. At the end of the idea-thon, we had a Shark Tank-like selection process — which was much friendlier than the TV version — in which members of the AI council judged and selected the best ideas from that group. That led to prioritization, planning, and other next steps based on the viability, impact, and how those ideas aligned with organizational goals.

Second pillar: Innovation to adoption

The second pillar of our strategy is around transitioning innovation into adoption. It’s really focused on those pieces and recognizing that innovation is often easier than adoption. I think that’s particularly true in healthcare, given the complexity of the environment with all the data privacy and security complexities and the challenge of obtaining buy-in from caregivers.

It also touches on the translational research component. As I mentioned, we have a pretty robust computational biomedicine or bio informatics program and an incredible team of researchers across many clinical programs who are applying these computational and statistical methods to move research from bench to the bedside. That’s something we’re pretty proud of here at Cedars; we have a long history of doing that; AI is just one of the ways that we’re pursuing it.

Third pillar: Ethical AI

The third pillar is around supporting the sound and ethical use in governance of AI. To that end, we’re establishing ethical and unbiased standards for these technologies to really validate that. All of that is reviewed upfront and throughout the life cycle of these solutions — that’s key, in our opinion. Monitoring the tools, re-running them through our AI assurance process and making sure that we’ve operationalized these in a way that aligns with appropriate use in accordance with our strategy. Those are really the three pillars in our AI council.

Gamble: Definitely. In terms of what you’ve seen at the Idea-thons and what people identified as key needs, was it what you expected or were there surprises?

Kwiatkowski: I’d say most of it was expected. It was interesting. We took a very open approach to it. The way it was structured was through 20 to 30-minute education sessions on what generative AI tools are and how they work. But we kept it open from the standpoint that we didn’t try to constrain their thinking at all. It was more, ‘imagine how you could use things,’ with a little bit of the art of the possible, and how you might use these tools to impact the way in which you work.

What was interesting is that their ideas aligned fairly consistently with many of the things we know our major vendors are working on. And by that, I mean the EHR and ERP vendors, among others. Many of the ideas they came up with were things that overlapped. That was good from a validation standpoint to know that our vendors are on the right track.

And then there’s a hodgepodge of other things across all facets in terms of how care is delivered and the way in which a hospital functions. We had a physician centric idea-thon. We had a non-physician idea-thon with clinicians, nurses, pharmacists, dietitians, etc., and we had more of an administrative or non-clinical idea-thon with folks from legal, finance, HR, facilities, and other areas. We definitely run the gamut of areas and ideas. It’s been a really fun thing to be part of, both for us and the participants.

Starting out in pharmacy

Gamble: The last area I want to talk about is your background as a pharmacist. I worked for a pharmacy publication years ago and it was so eye opening. Pharmacists are on the frontline and have so much access to patients. Can you talk about how that background has shaped you as a leader?

Kwiatkowski: Yes, that’s a good question. I appreciate it. I like to reflect on my background. Going from being a pharmacist to Chief Information Officer is somewhat unique. I know of two or three others who have had the same career trajectory. As you described, that background offers a unique perspective that allows us to approach technology as a clinician and as an operations leader. I think folks tend to think of pharmacists as people who know the most about medications; that’s certainly true or at least tends to be true. But they also tend to be good at closing care gaps, developing care plans, and navigating the complexity of healthcare on the frontline. Pharmacists tend to find themselves positioned between the physician or nurse and the patient or the caregiver, not to mention at times finance administration. All of that offers an opportunity for pharmacists to have a broad view of how healthcare delivery really happens.

Safety first

I also think there’s a synergy that exists between pharmacists and IT employees. Pharmacists — and this is a generalization — are wired in their DNA with a tendency toward precision, accuracy, logic, and safety. Those things translate well to work in IT. Safety in particular is something I’ve always been very passionate about. I think it connects very well to the way in which we try to serve the organization from a technology standpoint and the importance of the work we do. There are countless little things we do and touch that can have positive consequences and unintended consequences and there’s an expectation that systems and solutions are delivered safely and reliably.

I think pharmacists, historically speaking, have probably been a little ahead of the curve relative to some other clinicians in terms of technology adoption and innovation, particularly with some of the earlier entry systems and clinical decision support tools — things that existed 30 years ago before many folks in other clinical areas were as immersed as they are now in technology.

Gamble: Really interesting. It sounds like it was a great experience, and it has certainly benefited you.

Kwiatkowski: It’s been a great ride and a great experience. I don’t think I’d change a thing.

Related Posts:

Share Your Thoughts Cancel reply