Doug Flora, MD, is the executive medical director of oncology services at St. Elizabeth Healthcare in northern Kentucky. He is also the founding editor-in-chief of AI in Precision Oncology, the new peer review journal published by Mary Ann Liebert Inc., dedicated to advancing the application of artificial intelligence (AI) in precision oncology to a global community of oncologists.
In this exclusive interview with Damian Doherty (editor-in-chief of Inside Precision Medicine), Dr. Flora discusses his clinical and leadership role at St. Elizabeth and his conviction that AI will revolutionize the practice of oncology.
This interview has been lightly edited for length and clarity.
Damian Doherty: Doug, tell me about the mission at St. Elizabeth and the work you’re doing for your community in northern Kentucky.
Doug Flora: I’m a practicing medical oncologist, so I see patients in our clinic. I have an administrative role for our healthcare system as the chief clinical officer over our cancer programs and help lead a large team over six different facilities taking care of cancer patients.
This region is really ground zero for cancer in America. Kentucky is a hotbed—it’s number one in the country for colon cancer and lung cancer in both deaths and late-stage presentations. We have our hands full, but we’re doing a lot. of good work here at St. Elizabeth.
I’ve been in clinical practice for almost 20 years, but in this role for about six years. It’s been a real treat to work at St. Elizabeth—it’s a very mission-driven place, a very kind place.
Doherty: How has this cancer center enabled you to scale this new precision oncology approach and the quality of care you’re providing to a region that has such a high incidence of cancer?
Flora: We have thrown ourselves into screening and prevention. We have a prevention cancer clinic. We have a heavy emphasis on hands-on demonstration, teaching people that food is medicine. We have a very avid center for precision medicine and genomic health with about a dozen genetic counselors, pharmacogenomic pharmacists, and other support staff.
Doherty: As a cancer survivor yourself, Doug, and an oncologist, you are acutely aware of the importance of getting an early diagnosis so we can treat cancers more precisely, which of course is where the power of AI joins this narrative. When did you have your ‘Road to Damascus’ moment with AI’s potential and when did you realize this was something that you had to bring to St. Elizabeth’s?
Flora: Probably 5–6 years ago. Many of us recognized the problem—healthcare had become purely transactional, and I felt like damage was done over the past 5–10 years to the doctor-patient relationship, where it became more of a conveyor belt. That’s not how I or the doctors in my system are wired. But in a fee-for-service system, where there are more patients than doctors or providers, it creates a stressor. We don’t have the staff that we need and that’s not likely to get better as the number of cancer patients rises exponentially.
With over 18 million cancer survivors in the United States and 13,000 cancer doctors, I realized we had to find a new approach to this. The math doesn’t work. So I started looking into technology solutions that would help scale our nurses, social workers, dietitians, and others. I was impressed with how fast AI was accelerating this field and the number of tools that were becoming available. Tech is a very reasonable solution to take away some of the more mundane tasks from them. AI is an exceptional example, so I started to lean into that, and it’s only accelerated lately.
Doherty: Did you have like-minded colleagues who felt the same way? Who inspired you?
Flora: I’m a tech enthusiast, so I read everything, including Eric Topol’s 2019 book, Deep Medicine, which was illuminating. Some of the themes in his book resonated with me in terms of repairing that doctor-patient relationship and using these tools to make care more human. As I learned more about AI and started to speak around the country on this, I realized this was a real opportunity that we needed to grab quickly—to rebuild the patient-doctor relationship by removing those other stressors that weren’t productive for doctors or patients.
Doherty: I remember reading Deep Medicine, it was a seminal book. Generative AI has come crashing into our world recently, but it’s going to be incredible to see what these models can do in pulling all that disparate data together. That’s going to be the key, isn’t it, to create that more comprehensive picture?
Flora: We’re in this unimodal phase where it’s really good at pattern recognition and reading digital pixelated slides for pathologists or image recognition for CT scans or chest x-rays and so on. But as we get into this next generation—well beyond chatbot and into clinical decision making—I think we’re going to see more cures. We’re going to find earlier identification and more patients saved. That’s just music to my ears.
Doherty: It is very hard for the clinical policy and regulatory frameworks to keep pace with this exponential change. We’ve had a recent call for a moratorium on AI and there’s talk of these large language models being prone to hallucinations. How are we going to keep all these stakeholders aligned and ensure that we’re creating robust, unbiased, and safe AI systems?
Flora: I think that’s the sweet spot. That was the impetus for starting this journal, AI in Precision Oncology. As a clinician, I love this stuff, but there are 59 other doctors in our building, most of whom don’t know the difference between deep learning, machine learning, and natural language processing (NLP). They don’t have time, they’re doing a fantastic job taking care of patients every day, but they’re overwhelmed at the same time. Industry is hurdling forward; our industry partners are 100% all in on rapid development, sometimes for commercial gain, sometimes for the greater good, but I don’t feel like the two groups are talking.
My hope for AI in Precision Oncology is to combine those two forces. I’ve recruited about 25 people on our editorial board right now—about 40% academic heavyweights, about 40% industry heavyweights, and 20% that are both. I’m really enjoying the dialogue between the two groups as they discover unique skill sets and perspectives and start to blend them. The articles that are starting to come in reflect those two teams communicating, which is exciting.
Doherty: How does the procurement for new technology work at St. Elizabeth? There must be tech companies wanting to sell you transformative AI solutions. How do you decide what’s relevant, what’s necessary, or that this is a tool that we need to look at?
Flora: We are a forward-leaning system that cares about this technology, and we are looking into it heavily with very bright IT teams, very accommodating C-suite, and board members who believe in what we’re trying to build at St. Elizabeth from a cancer perspective. We are in the process generation phase: how do you handle governance? Who is on the committee that will adjudicate this?
Everybody in the industry is approaching St. Elizabeth as an early adopter, and we can’t buy 200 programs at once. We don’t have the IT muscle. So we have to determine that carefully. I think there’s a user case right now for the things that have been validated.
We’re looking at things like NLP to remove documentation burdens for doctors. I think there are some things going on in that world with Microsoft, Nuance, and Epic that were announced recently. Our radiologists have pushed through mammography—AI augmentation of their reads that are highly accurate and reduce workloads. We’re dipping our toes in, trying to gain experience.
I have been almost heretical here and I hope people aren’t getting turned off. The velocity of change is not what we’ve seen before. When we had to adapt to COVID, suddenly the world was upended, and you had to build everything new. We went to virtual health, to quarantine, and people working from home and the whole system had to adapt immediately.
I think 18 months from now, we’re going to find ourselves immersed in AI and we should be vetting these tools carefully. But we should view them as a tool like a stethoscope or a CT scan, to help the patients.
Doherty: I’m glad you mentioned NLPs because there seems to be a paucity of institutions that are still using them to alleviate that administrative burden. I assume you’re also implementing AI on the digital pathology side as well?
Flora: Yes. Our pathology colleagues are digitizing all our slides this year. We’re not quite there with NLP for voice recognition. We use a product—Nuance—that we’ve used for many years for dictation and that’s been NLP from the beginning. But it’s refining now with the advent of generative AI coming to the masses. The next round of products is going to be hands-free. It will be able to take large unstructured data sets and make sense of them. We’re not there yet, but I’m going to be asking for pilots and trials and, please, let oncology have it first. I’d love to do that with my own division. My doctors are ready.
Doherty: Another important part of AI and precision medicine is education. Many physicians weren’t trained in genomics. And there’s still a huge education gap, even in residency. It can be confounding for physicians to know what test to order, and which panels to use? Do I do the whole genome or exome? How do I interpret variants of unknown significance? How have you approached that educational side at St. Elizabeth?
Flora: I hope they’re teaching it to our new medical students because it is half of my life as a cancer doctor. The other half might be immunology, which we also didn’t have heavy training on either.
We’re blessed in our system. We have a dedicated center for precision medicine and genomic health, led by a very bright, dedicated medical oncologist. Brooke Phillips, MD. She trained in cancer genetics at City of Hope and at Cleveland Clinic and has been leading our charge. The director of our center, Jaime Grund, is also extremely involved in providing education. It’s really important for me, as an early adopter, to listen to their counsel and say, maybe what’s being promised isn’t quite ready.
Doherty: Let’s talk about the new journal. How are you getting on with submissions and preparations for launch?
Flora: There will be some interesting front matter in terms of education. One of our goals is to help oncologists get caught up. One of our editors, Scott Penberthy, is managing director of applied AI for Google, and he will present instructional front matter each month—Definitions. What is NLP? What is multimodal, unimodal? We’ll be doing multimedia and webinars as well.
We have a good piece from Lee Hood, MD, PhD (co-founder of the Institute for Systems Biology (ISB), based in Seattle, WA). Dr. Hood has a 50-year perspective on this field and has been talking about personalized medicine for half of my life. He’s seen everything from the original automated DNA sequencer he helped design to where we are now. He launched a company that’s digging into gigantic sets of data to introduce these concepts.
As a working clinician, I want digestible, easy-to-use tools that make my life easier. Every month, there will be a section called ‘‘Prompt Assistance’’—how to construct a generative AI prompt that helps you do your job to fight an insurance company’s denial or to write a letter of recommendation for a fellow or a resident. How can we use these tools to use our voice, but accelerate the efficiencies with which we work? So there’ll be a lot of user-friendly content in the journal.
Doherty: What would you perceive to be a measure of success? And what would you like your legacy to be when you hang up your white coat?
Flora: I have accomplished most of the goals in my career I always dreamed of doing, and now I’m trying to build something that is more impactful. Each day I think, what levers can I pull that will reduce suffering the most or save the greatest number of patients with screening and detection or education? I might save more cancer patients as an administrator, building screening programs, or as an editor, educating clinicians on tools that they could use to take better care of patients or make better clinical decisions. 5–10 years from now, I hope I will have added to the Global Happiness Index because we have reduced the burden of cancer on society.
These [AI] tools have incredibly high potential. I’m an eternal optimist, and I think these things are here to help, but we need guardrails. We need strong legislation. I think some of the leaders who understand this technology the most are the ones asking for help to rein it in before it becomes unwieldy. I think we’re in a critical phase where we’re going to figure this out in the next year or two.
Damian Doherty has been in media and publishing for nearly 30 years, beginning in the early nineties at News Corporation. Damian has managed, edited, and launched life science titles in drug discovery and precision medicine. He was features editor of Drug Discovery World for fourteen years and founded, established, and edited the Journal of Precision Medicine in 2014. In parallel, Damian founded and organized the Precision Medicine Leaders’ Summit, a global, immersive 3-day senior leadership conference that still runs today. He edited AIMed magazine in 2019 before launching Photo51Media, a platform for illuminating untold, compelling stories in precision healthcare. Damian joined Mary Ann Liebert in 2021 to help steer the new rebrand and relaunch of Clinical OMICS to Inside Precision Medicine.
