When an individual is diagnosed with cancer, he or she goes from having a primary care physician to having a team of care providers, including nutritionists, surgeons, and radiologists. It can become overwhelming, to put it mildly. It’s precisely why Sarah Cannon, the Cancer Institute of HCA Healthcare, implemented a system of cancer navigators to help guide patients and their families by coordinating appointments, answering questions, and providing education. These cancer navigators, says CIO Andy Corts, “are our most precious resource.”
The challenge came in harnessing the data — which can be complicated in any area, but is infinitely more difficult in the “incredibly fragmented” oncology environment. Corts and his team have made it their key priority to combine data sets into a common warehouse and leverage analytics to be able to “view the entire patient journey.” In this interview, he talks about how they’ve been able to define a cancer data model, how they’re partnering with Digital Reasoning to automate manual processes and enable more personalized care, and the journey that brought him to HCA, and eventually, Sarah Cannon.
- Care navigators — “They’re our most precious resource.”
- From one PCP to a team of providers
- Working with Digital Reasoning to analyze pathology reports and automate processes
- AI models – “They’re build on statistics.”
- Past experience in IT management & consulting
- Leading a small team – “It was fun. I like to get my hands dirty.”
- Sarah Cannon’s mission: “All we do is focus on care delivery.”
LISTEN NOW USING THE PLAYER BELOW OR CLICK HERE TO SUBSCRIBE TO OUR iTUNES PODCAST FEED
We don’t always employ the surgeons. We don’t always employ the radiation oncologists and the medical oncologists. But ultimately, the patients see us as a home. And so our philosophy was every patient deserves a concierge who can be there for them as they’re going through this journey.
We’re constantly updating the model as more and more data comes in. I think we’re incredibly lucky in that from a pathology perspective, HCA has over 2 million reports in our Meditech record on an annual basis. That sample size ultimately can drive a better model.
I’ve been happiest when the learning curve is the steepest. And in cancer, there’s never a dull moment. There’s always a new technology, a new genetic or genomic test, or a new implication for a treatment pathway.
‘We’re going to get you an answer as quickly as possible as to whether you need to continue investing in this drug.’ That’s what our platform is designed to do. It’s what all of our information systems are designed to do — to get the trial started as quickly as possible and get the drug to that patient.
Gamble: A big part of that, I’m sure, is helping physicians to be more efficient with their time, but then also being able to dedicate more time to patient care — which is critical in any space, but particularly in oncology.
Corts: In looking at our history of becoming a research company to manage the cancer service line for HCA, our biggest initiative that we launched was around cancer navigation. Because we’re a community hospital, we don’t always employ the surgeons. We don’t always employ the radiation oncologists and the medical oncologists. But ultimately, the patients see us as a home. And so as we launched, our philosophy was every patient deserves a concierge of sorts who can be there for them as they’re going through this journey.
When a patient is diagnosed, they can, overnight, have eight to nine different cancer care providers, whether it’s a nutritionist, a genetic counselor, surgeon, or pathological radiologist. All of a sudden a care team is reaching out to them, and so we decided to coordinate that through a cancer navigator. It made such a fundamental difference in the way patients manage their journey, because if you think about it, when you see a primary care physician on an annual basis, you probably get about 15 minutes with him. Think about what it would be like if you went to get a colonoscopy, went back to your primary care physician and got a cancer diagnosis, and had 15 minutes to spend with him. It’s just not feasible to get the kind of care that you deserve in that setting.
And so, what we offered to primary care specialists was this notion of a cancer navigator, where not only does a physician come in and deliver the diagnosis, but he can say, ‘Call this number. It’s the Sarah Cannon hotline, and we have a cancer navigator to help coordinate your appointments with the surgeon, the radiation oncologist, and the medical oncologist, and ultimately to help educate you about this diagnosis.
It was having an amazing impact, but we found that our navigators were only able to touch around 50 patients a year. They’re spending a lot of time not just with the patient, but on reading those pesky pathology reports and sifting from among 2 million reports. And so we reached out to Digital Reasoning and built out an AI model that would be able to scan all 2 million reports and immediately identify, for example, breast cancer. We could then funnel that to the breast cancer navigator so they could reach out to the patient through the primary care physician and coordinate that care for them. Or the AI model might say, ‘This is a complex GI patient’ or ‘This is a lung patient,’ and we can coordinate. As a result, we saw the navigators go from being able to do outreach to 50 patients a year to more than 250 patients, just by automating the manual process of reading pathology reports.
Gamble: Right. So a big opportunity to increase satisfaction among both patients and clinicians.
Corts: It’s a huge differentiator. The next level we want to take it is in determining the cancer pathway. A big part of our navigation program is to say, ‘This is the appropriate pathway for that patient’ and ‘Here’s when the patient needs genetic or genomic testing, here’s when they need surgery, and here’s where they need chemotherapy first.’ We’ve written all these rules, but it depends on the pathology report. And if a lot of the data is driving what the care pathway will be, the navigator then has to do data entry on the report to ultimately kick off the pathway.
We’re working with Digital Reasoning to take the unstructured text living in the report that ultimately will determine whether or not a patient goes to chemotherapy first or to surgery first. Can we have the machine abstract it as opposed to the human, and can we pick up more time from that function so care navigators can go on to see 300 more patients annually? And so we’ve been working not just on doing the identification function, but also the notion of automated abstraction to drive more productivity. Because at the end of the day, these cancer navigators are our most precious resource. The funny thing is when we get letters from patients, they won’t call out how great the surgeon was. They won’t call how great the radiation oncologist was. They’ll say they couldn’t have made it through the journey without the navigator.
Gamble: In terms of the work you’re doing with Digital Reasoning, is that going to be a continual effort?
Corts: It’s definitely a continual effort. When you think about it, these AI models are built on a lot on statistics; sample size is critical to getting less false positives and false negatives, and so we’re constantly updating the model as more and more data comes in. I think we’re incredibly lucky in that from a pathology perspective, HCA has over 2 million reports in our Meditech record on an annual basis. That sample size ultimately can drive a better and better model. I think we’re close, at least on the identification side of the house, to 99 percent accuracy.
On the automated abstraction side, we’re in the early phases. We’re around 80 percent, but if you assume more and more data comes in, and as the model learns more and more through the user validation and the feedback loops, you’ll ultimately see that number get much higher in terms of being able to automate on behalf of the user. The nice thing about these models is they’re always getting smarter. They’re always getting more accurate, and it’s pretty amazing the dataset that we have can drive that.
Gamble: Right. Now, looking at your career, you’ve been with Sarah Cannon since 2008, correct?
Corts: Yes, at the end of 2008. Sarah Cannon become a member of the joint venture around 2006. When I joined it was around 70 people. I had a team of about five IT employees, and was a director at the time. We have undergone tremendous growth to where now we’re close to 1,000 people, and I have about 90 reporting to me. The timing was right, and I was really tight with the CEO. And so when it was decided the organization needed a CIO and a leader — even though my team was only about 5 to 10 people — they put their faith in me.
Gamble: And you had previously been with HCA, so you knew the organization pretty well.
Corts: Yes, that was definitely helpful. My career actually started in consulting. In 2000, I went to work with a company called Cap Gemini Ernst & Young, where I did IT strategy work as well as some big clinical information system installs like Allscripts. Mostly, I went to academic health systems and wrote their IT strategies on where they were going to make their investments over the next 5 to 10 years. I did that for around four years, which gave me a good background, and I was ultimately recruited to HCA to manage their IT strategy.
When I arrived, they were pretty early on in terms of developing an IT strategy team, and so I set up the process for how the hospitals would all come together under one IT strategy so we could better surface what the needs were across the enterprise. I did that for about four years; during that time, I was asked to create a strategy around clinical trials, particularly leveraging the electronic health record. At that point, HCA just made an investment in Sarah Cannon to become their clinical trial hub. So I went there and met the CEO, and helped her to create a strategy. I had been doing strategy for about four years at HCA, and had taken the process about as far as it could go. Then my entrepreneurial spirit kicked it, and went over to Sarah Cannon when it was still in its infancy stages. I’ve really enjoyed the growth in the past 10 years.
Gamble: You said you had about 5 people working underneath you, and it increased to 90. That’s some serious growth.
Corts: It was fun. I like to get my hands dirty. We definitely scale small here, but when we need to, we can scale large as well.
Gamble: As far as being oncology, it’s an area where care is certainly fragmented, but one where research is advancing so quickly. Is that part of the draw in being in this area of care?
Corts: Throughout my career, I’ve been happiest when the learning curve is the steepest. And in cancer, there’s never a dull moment. There’s always a new technology, a new genetic or genomic test, or a new implication for a treatment pathway. Being on the research side of the house, it’s easy to be mission-motivated simply because of what we’re doing. We enroll more than 3,000 patients a year to clinical trials. The FDA is approving new immunotherapy and molecularly-targeted therapies at a rate unseen, and Sarah Cannon has had its hand in the majority of those being an early stage developer. We’ve been exposed to so many success stories; it’s truly been a great journey these past 10 years.
Gamble: Sure. And when you think about speeding up and improving the clinical trials process, I know that’s been a focus for a long time.
Corts: Definitely. We’re not an academic medical institution; we like to think that what we do, particularly on the clinical trial side of the house, is focused on care delivery. There isn’t as much politics about which trial we decide to do. Our whole goal is efficiency — the faster we can get a study going and patients enrolled, the quicker pharma can get to an answer. That’s a big piece of our valued proposition; we tell pharma, ‘we’re going to get you an answer as quickly as possible as to whether you need to continue investing in this drug.’ That’s what our platform is designed to do. It’s what all of our information systems are designed to do — to get the trial started as quickly as possible and get the drug to that patient as quickly as possible.
Gamble: Really interesting. Every one of us knows someone who has been impacted by cancer. Your team is doing incredible work.
Corts: And the more we can scale it, the more that we can get it into every market around HCA, the more that we can get the best cancer treatments to patients close to their home. That’s our mission.
Gamble: Right. Well, that answers my questions. This has been great. Thank you so much for your time.
Corts: It was a wonderful opportunity. Thanks for including me.
healthsystemCIO’s Interviews and Podcasts are sponsored by: