Lance Berberian - The Impact of Technology on New Healthcare Models (1/31/2019)

Thank you everyone. I really appreciate you taking your evening to come and hear me speak about LabCorp and healthcare and some of the things we're doing to really transform our industry. To begin, I know we touched on a couple of the places that I've worked. I'm a serial health IT guy; that's where I've spent my career. I started out in electronic data systems, which is a huge consulting company, but I was in their healthcare and life sciences practice, which allowed me to see a lot of different companies, whether they were pharmaceutical... I was their chief information officer for 23 years. Can you imagine? It's a great job. So then I went to a company called Interim HealthCare, a home health care provider. We paid about 80,000 nurses a year to go into the homes of people who were sick or elderly to take care of them. I had a brief stint outside of healthcare but missed it, so I joined Idexx Laboratories. Has anyone heard of Idexx Laboratories? Idexx is like LabCorp for veterinarians. If you take your dog, cat, bird, or snake to your veterinarian and they draw blood, they're likely to... As chief information officer, I have all the standard responsibilities you would expect: all the infrastructure like networks and data centers, all the technology within the company, and all the computer systems that our employees use, whether laboratory systems, billing systems, finance, or HR systems. But along with that, I have a few exceptions. I'm responsible for all of our product development. We have a tremendous amount of custom software development delivered to our customers for their use, whether a physician, a consumer who had lab testing done, a hospital, or a payer like UnitedHealthcare, Aetna, or Cigna. We charge them for the software depending on what the system does. I'm also responsible for our robotics. Does anybody here have exposure to robotics? Okay, so it sounds like there are some people. You'll get an interesting perspective on some of the things we do with robots within LabCorp. I also have a team called bioinformatics. The guy who leads that team describes his people as like flying fish because they're a weird mix: half biologists and half IT people. They code in Java but know what DNA means. They look at biomarkers but are analytics, big data people. They're likely to have data science backgrounds. The guy who runs it has his PhD from Yale in artificial intelligence. It's important for you to get a sense for our company because it's a little misleading. You hear the name LabCorp and think you know what that is, but we suffer from something. Remember when Google owned Google and a bunch of other companies, and they renamed the parent company to Alphabet? Think of Coach, the handbag company. They bought a whole bunch of companies, but the parent was Coach, and they owned Coach and a whole bunch of companies, then they changed the parent name to Tapestry. So we are LabCorp, and I am the CIO of the parent company, and we own... Star for us is different than it was before because we believe the combination of the companies is going to transform healthcare as we see it today. So what Covance does is a drug development company. They are focused on working with pharmaceutical companies to bring drugs to market. I'll get more specific about what they do and why it's important from a technology standpoint. But what's important for you to understand is the mission of the company. We have an important mission that affects many people: we will improve health and improve lives, and we will do it in three ways. We are going to focus on world-class diagnostics, we are going to bring innovative medicines to patients... So like I said, we are LabCorp Diagnostics. We offer about 5,000 different tests. Some are basic routine tests, like a cholesterol test. We also offer very hard, complex tests. We are the leader in the industry in something called the NIPT test. If you are a pregnant mom at risk of having a child with a developmental disorder like Down syndrome, the standard of care is you take a really long needle and stick that... It's called the NIPT test, the non-invasive prenatal test. You simply draw the mother's blood out of her arm, look at the DNA circulating in the blood. You'll find a lot of one kind of DNA and a little bit of another kind. That little bit is the baby's DNA, and then you test it. That's non-invasive prenatal testing. So it gets as simple as a cholesterol test and as interesting and complex as genetic testing and advanced testing. We do that with 700 PhDs and MDs because a lot of it is the talent. We also do it with courier drivers who drive around and pick up at all of our customers... The person who draws your blood or collects your specimen. We have 11,500 phlebotomists. Some are in those 1,850 locations, and some are in doctors' offices. Like I said, we're also now that second company, the drug development company. So when a pharmaceutical company invents a new drug, the first thing they do is patent it because they don't want anyone else to take it away. But here's the problem: you invent a new drug and patent it, but you can't sell it until the FDA allows you to sell it in the United States, and there is an equivalent of the FDA in every other country. Its nickname is first in human. So you put people, generally men, in hospital beds and put the pharmaceutical in their body for the very first time in very small quantities. You increase those quantities over time, monitoring that person, checking their vitals, checking their diagnostic test results to make sure there are no adverse effects. These people don't even have the disease the drug treats; you're just looking to see if there are any adverse effects. We do that. We have a 300-bed hospital in Dallas, and we have them in England. We have these facilities to... You're really applying it in this manner. If it passes that, if the FDA will let you go to Phase 3, that's a larger population. At the end of Phase 3, if the FDA approves your submission, you get to sell that drug. Phase 4 is when you learn later that it might actually be used for other things. I'm going to talk today about a drug called Keytruda. Keytruda is an important drug for non-small-cell lung cancer. But guess what? After it was in the market, they realized it also treats melanoma. So that's another indication for that drug, which would be considered a Phase 4 indication. We do the laboratory testing across those phases, and we also use something called real-world evidence to try to expedite the trials. I'm going to talk to you a little bit later about that. LabCorp is a growing company. The blue is LabCorp Diagnostics, the orange is Covance. You can see that we'll finish last year at about $11.3 billion in sales. Our earnings per share increases on a consistent basis. That means profitability, right? Earnings per share. So it's revenue and profitability growth. One of the things I like the most is we invest in technology. If you look at our capital... I'll talk to you about some of the innovation that we do that supports our mission across those two companies. We all know from hearing on the news, and maybe experiencing it yourself, that healthcare is a problem in our country. Part of it is controlling the cost, and part of it is providing better care at that controlled cost. So at LabCorp Diagnostics, we are focused on improving the quality of care and managing to lower the cost of care as well. We are also focused on enhancing the consumer experience because what we are learning is as people have to pay more of their healthcare bill, they start to ask questions. They start to care about who's... They said at my age I might be very polite and ask some questions. At your age, you really want to know what's going on, right? So the culture of interacting with the doctor is shifting over time, and it's partially because people have to pay most of it out of their pocket. So we provide a tool to physicians called LabCorp Link. It's an app, a responsive app that works on a cell phone, tablet, or computer. It gives that physician an enhanced view of laboratory data. They can place an order for laboratory work, but they can review the results. It has a reference range that goes along with it. It basically says if you take the large majority of people, there's sort of a high value and a low value you should be in between, sort of like your cholesterol test. Now imagine a physician meets a patient, runs the test, and it determines that person's cholesterol is pretty high in the range. But it's in the range, but it's pretty high. Is that good or bad? I don't know. Because if the last time they had a cholesterol test it was way high outside of the range, then that doctor should be saying to the patient, "Great job! You really have been improving the way you're eating and exercising and lowering your cholesterol." If the last... That allows the physician to have that enhanced view and interact with that patient in a more value-added way. We also provide enhanced reporting to the physicians. In the old days in the lab business, you sent the data. What we want to send is the insights that cause that patient to be treated better. It also makes the physician more efficient. Being able to share those insights through an enhanced test report is an important part of what we are doing. We also invented something... I know, I should just... There you go. So this is a key piece of technology comprised of two main components. One of them is MDSL, Medical Domain-Specific Language. This is a language that lets a clinician author rules so that you don't have to have programmers write code to create the rules. Then those MDSL rules get inserted into a high-speed rules engine. I would almost describe the rules engine, which is called the Matrix, as if you think of a Rubik's Cube except every square of the Rubik's Cube can also be a Rubik's Cube. So instead of having a billion if-then... And we do that across a variety of disease states like chronic kidney disease and diabetes. These are all ways that help the doctor understand, based on a complex rule set that involves lab values and non-lab values, what they can do to improve care for the patient. As you can imagine, in the healthcare ecosystem, everything is very large. LabCorp is an at-scale company. We're doing half a billion tests a year. Well, you can't type anything; it's just too high volume. So we do a lot of EDI connectivity. You can think of that as you go to the hospital or a doctor's office, they put that lab order... An enhanced view of the app I showed you earlier, but at least in the medical record of their core system, they have those test results. We build between 8,000 and 10,000 of those EDI connections every year. We have 67,000 of them running every single day to get data in and out of the company. One of the powerful things about LabCorp coming together with Covance is the idea of health systems being used for clinical trials. We know that the pharmaceutical company wants to compress the time frame of the trial because, as I said earlier... I'm going to go back to Keytruda. Peak sales for Keytruda only for the non-small-cell lung cancer indication is $6 billion a year. I like that because I'm not good at math, and $6 billion a year is easy to figure out. It's half a billion a month. Every month you delay the clinical trials, they lose half a billion dollars in sales under patent coverage. Well, LabCorp has great relationships with health systems and hospitals. Health systems and hospitals love to participate in clinical trials. It's very prestigious and a revenue source. So can't LabCorp and Covance coordinate to help the pharmaceutical companies and the health systems? In a company like ours, an at-scale company, what you'll find—and this is a common question people ask me as a CIO—is, "Lance, when do you buy software and when do you build software?" It's an interesting question. My thought is that if you can't find it in the market to buy and it's a differentiator for the company, you need to build it. In a company that does the specific thing we do, there are certain examples where for our internal employee use systems and for products we're developing to give to customers, we need to build software. Then we are layering that on top of hyper-converged computing in our data centers, combined with cloud-based services in Amazon and other cloud-based services. We've integrated into it a big data platform because a lot of the capabilities we provide to customers and use internally are based on the huge amount of data that is the natural exhaust from the laboratory testing we do. All of that data can be used for statistical analysis, for serving patients, for serving physicians. All of that is integrated into a modern platform. When I talk about that big data platform, this is very important to a laboratory. Data is pumped into our HANA platform, which is an in-memory computing platform. If you think of the orders of magnitude faster memory is accessed compared to disk drives, by pumping that data up into that HANA platform, it gives us extraordinary performance when accessing that data. We do that for simple things like reporting, but we do that for advanced things like data sciences and other capabilities we offer. We also have a system called AQuA. AQuA is a system that makes sure that when the phlebotomist is collecting your specimen, they're doing it correctly. Because the color of the top is different, right? If you get one drawn into what's called a lavender top, that has a chemical in it called EDTA. EDTA is an anti-clotting agent. So if they're going to do a blood cell count, they don't want your blood to clot, so they'll draw that into a lavender top tube. Probably the most common tube is called a speckle top tube. I call it that; it's also called a serum separator tube. The most common type of specimen used in a laboratory is serum. It's used in chemistry, serology, special chemistry—all throughout the lab. If you take your blood and put it in one of those speckle top tubes, you'll notice... When they take it out of the centrifuge, it's going to be three stripes: a dark red stripe at the bottom (that's all your red and white blood cells), the goo separator, and then this yellowish liquid on the top that is your serum. That's where all the chemicals in your blood are suspended. Your cholesterol is a chemical; it's suspended in serum. It's a very common substance used throughout the laboratory. Well, if you think of that phlebotomist drawing your blood, they need to draw the correct amount into each type of tube to accommodate all of your tests, and then they need to treat the tubes correctly. Sometimes they're refrigerated, sometimes they're... This guides that phlebotomist to collect the correct amount of specimen in the correct containers to serve your entire lab order. Again, improve the quality of care, improve the cost of care, and serve the patient, serve the consumer. So I spent a little extra time on what is serum for a reason. I mentioned before that we're involved with robotics. It's an important part of what we do. We design and build robots from the ground up, not just buy them and install them in the lab. We do some of that too, but for key robots we design... It's called Propel. By the way, we make a variety of types of robots, but Propel is 7,000 square feet for one of them. It does serum preparation in 250,000 test tubes in an eight-hour shift. Sometimes when I say this, people don't believe me, so let's take a look. So it's grabbing those tubes and now it's going to dump that serum off the top of the tube like I described to you after it was centrifuged. All of this again designed and built by... Tapping the specimen to preserve the integrity when it's going into storage. At some point, you're going to see it panning across, and you'll get the full sense of the size of the robot. That whole thing is the robot, 100% designed and built, software and hardware, managed through cloud computing by LabCorp. So I mentioned a little bit earlier the... Everything has DNA. You have DNA in every cell in your body. If I took the DNA out of one cell of your body and stretched it out, it would be that long—from one cell. It's all curled up. It has 3.2 billion data points. We all know DNA is shaped like a ladder that you twist, a double helix. Every rung is a piece of data because there's a chemical: adenine, guanine, cytosine. So it's either an A, T, C, or G at every point. That gives me everything about me. It's all coming from my DNA. I got half from my dad, half from my mom. Knowing that is important because you inherit things because of it. Those variants cause things about your health. So anybody remember when Angelina Jolie had a prophylactic double mastectomy? Why? Because she had a variant in her DNA. She had the BRCA variant. 3% of women in the United States get breast cancer. If you have the BRCA variant, 54% of women get breast cancer. The second type of testing we do is oncology testing. If you think about it, cancer is mutated DNA. If you can know the specific mutation, you can prescribe the drug better. Knowing that particular variant means you have to sequence the DNA of the cancer. We also sometimes are looking for the DNA of something that's in your body. You're not looking for the patient's DNA or cancer DNA; you might be looking at HIV virus DNA because the specific type... Trying to figure out who somebody is, that's identity. A lot of times we're doing identity because the mom is saying he's the dad, the dad is saying he's not the dad, and you have to test and figure out based on the child's and the father's if it is right. That's another reason you do identity. Little-known fact: LabCorp is one of the last large users of Polaroid film. Remember the film that goes like that? The reason is that in many cases, by law, you have to have a Polaroid of the dad with a needle in his arm to make sure it ties back through the chain of custody. Interesting piece of trivia. So I talked a little bit before about patients as consumers. In today's day and age of healthcare, they get to decide stuff... After receiving their healthcare services from LabCorp, about how much they're going to get charged. That's common in healthcare in the United States today. You don't know; your insurance company does something, and eventually you get a bill for some amount. We would like to inform patients upfront. We think it's the right thing to do. We're creating patient-friendly test reports. We talked earlier about physician-friendly test reports, but if you're a patient, you might likely want to see what your test results are in a way you can understand them and consume the information. We're starting to put LabCorp blood draw stations in Walgreens locations. Consumers love this, especially since we have a feature to send you a text when it's time to get your blood draw. So... Rate this as a fantastic thing. We have them in about 20 stores as a pilot, and we're now identifying 600 more. We're releasing Pixel by LabCorp. Pixel is a test kit that you can buy. You can lance your finger, put your blood in it, snap it shut, send it to LabCorp, and we'll run the test off that device. Again, very consumer-oriented. We value your time and want to provide the most efficient experience before, during, and after your visit to our... Location to schedule your visit, then answer a few questions, and you're set. Our system will then remember your information to save you time when scheduling future visits. With Pre-Check, say goodbye to completing forms on clipboards when you arrive for your visit. The LabCorp Express station provides a quick self-guided check-in process. All you need is the QR code sent to your device, your driver's license, or a state-issued ID. The LabCorp Express station also provides an easy way to make a payment after your visit. Sign into or activate your LabCorp Patient Account. It's an easy way to securely view your lab results, pay your bill, schedule your next visit, and more from any device. We look forward to providing you with a prompt and reliable... At that iPad tablet that had that little tray you drop your driver's license onto and your insurance card onto, we're stopping the patient from having to hammer all that information in. When they drop their driver's license on, it's going to snap a picture of that driver's license, suck it up into the AWS cloud, apply machine learning algorithms to identify the specific card type, which will allow us to rip the data off using OCR and pump it into the system and present it back to the person so they don't have to type anything. Then when they do the insurance card, we do the same and follow it with a real-time call to the payer—whether it's Aetna, Cigna, Humana, Medicare, or Medicaid—to say, "Does this person have coverage and what kind?" And we present... That other piece, right? This is the piece where we're trying to focus on getting the drugs to market faster. The trick to this is we have to improve the trials using real-world evidence. I want to talk to you about something called companion diagnostics, as well as some targeted software we provide to pharmaceutical companies. The trick to expediting the trial and shrinking the time frame so the drug comes to market quicker is three things have to come together. There is a protocol. The protocol says the inclusion/exclusion criteria for a patient to participate in this trial: they need to have these lab values, there needs to be a certain number of them that are men... Can be part of the trial. Now you can imagine if you design that protocol too restrictive, you'll never be able to find the patients fast enough. So knowing how to design that protocol becomes very important to actually recruiting the patients into the trial. Then there's a matter of identifying where are these patients, and hopefully you have clusters of patients around doctors that will operate the trial. They're called investigators. A doctor that participates in a trial is called an investigator in clinical trials. So you really want to find the investigators and sites that are near patients that fit the protocol. Now you can imagine... Stuff. So let me give you an example. Take an ulcerative colitis trial. This is an actual trial. We look in our LabCorp database and find about 52,000 patients that have an ulcerative colitis diagnosis code. You then start to apply the inclusion/exclusion criteria, and you end up—this is a true trial—with a screen fail rate of about 75%. So now we no longer have that massive number of patients we can pursue; we have a smaller number that meet all the criteria that make up the inclusion/exclusion criteria. Okay, we know about sites because we do clinical trials, so we know which doctors are good at which disease states for clinical trials. We also know how many they're running right now because you don't want to pick a doctor that's really busy doing a ton of different trials because they can't give yours the attention. So in the end, we're able to take the location of the patients and the location of the doctors that are good at those clinical trials and find the clusters where we need to recruit patients into the trial. Again, all for the purpose of getting that drug out to market and into the patients that really need it—in many cases, a life-saving drug. LabCorp knows all about laboratory testing. Well, it turns out that some drugs are very good in some cases but not in others, depending upon a laboratory test. So I'm back to Keytruda for non-small-cell lung cancer. Keytruda works because of a very specific interaction. In every one of your bodies, you are born with a type of cell that fights cancer. Did you know that? It's called a T cell. T cells fight cancer. Here's the problem: also in your body, you have a mechanism to tell the T cell... Object. There has to be a way for that kind of cell to say to the T cell, "Don't kill me, I'm good." That's called the PD-1/PD-L1 interaction. It stands for Programmed Death Ligand 1. It's the way to say "don't kill me" to the T cell. What happens is that some cancers learned that trick and are able to say to the T cells, "Don't kill me." What Keytruda does is it reinvigorates the T cells so they don't fall for the trick, basically. Now, when you do the trial in a case like this, you create what's called a... Work because it's not using that PD-L1 interaction. So it improves the effectiveness during the trial. But of course, what will the FDA say? Of course, if you did the test to limit the people coming into the trial, you have to do the test when you prescribe the drug. So it becomes on-label for the drug. So let's think about that. LabCorp Diagnostics can help the pharmaceutical company come up with the companion diagnostic that is going to make that trial successful. If then we can commercialize the test to make it available for people to use, it can make sure that drug that might... We also provide very specific solutions to pharmaceutical companies that are differentiating. We're not trying to be the primary software vendor for pharmaceutical companies, but when we look at the process that pharmaceutical companies go through, we see areas that are focused that can be addressed with software that we deliver to customers, and we sell that to the pharmaceutical companies. So you might think of something like PharmAcuity. PharmAcuity is a product that helps the customer design the trial for success. What country should I run it in? In the European Union, there are a lot of countries. Which one is most likely to have a short and efficient... We have a product called Endpoint. Now, when you do a clinical trial, generally it's a double-blind trial. There are exceptions for oncology and others, but generally it's a double-blind trial. That means some people are on the drug and some people are on a placebo because in the end, the FDA is going to want to compare those two populations and see that the people taking the drug do better than the people taking the placebo. Well, something has to determine: did this patient meet the inclusion/exclusion criteria? Then which ones are going to get the placebo, which ones are going to get the drug? And then manage the shipment of the correct material to the correct physician so that you have a person consistently throughout the trial... Monitoring: are the sites handling the trial correctly? Are the patients at any sort of risk? By doing that, it helps to ensure the success of the trial by using technology. The fourth one, Global Specimen Solutions, is basically a way for the pharmaceutical company, the patient, the physician, and the lab to share data in a way that optimizes the flow of laboratory testing in the midst of the trial. So I hope as you think about the things I shared with you, the topics I talked about all facilitate the three primary goals that... Okay, so if you have a question, we actually have two mics. I'm going to give our ambassador here one. We're going to give this ambassador one. Raise your hand and please wait until you get a microphone so we can capture the question on video.

That's specific to the CIO role. It's very much a half business, half IT role. That's a part that I think is an important differentiator for someone trying to go into that particular type of job. You really have to understand the business enough because the CIO is primarily responsible for making sure that the company strategy links to the IT strategy. If they diverge, it ends in a bad place. Thank you for the question.

Yes, it's a very interesting point. If we can forge the electronic connection to that customer, then it's unlikely that they would shift to another competitive lab if they had to switch to paper. So we do evaluate each request to build an EDI connection to see if it's cost-justifiable and if it makes sense based on the customer volumes and the longevity of the contract we would expect to have with the customer. It causes us to have a very sticky relationship with that customer when we have that kind of integration.

This is an area that keeps me up at night: cybersecurity. It is a very difficult environment that requires us to have a level of attention to it that is... He most recently was the chief information security officer at the Mayo Clinic, so he has a healthcare background and an understanding. He reports to Chuck McAllister, who reports to me, who's sitting in the back. We spend a lot of time trying to delve into how we ensure the privacy and security of that information. There are laws about it, but even more importantly than the laws, it's just the right thing to do. You're talking about my mom, sister, brother, or family member. I would hate to think that their personal healthcare information would be shared by any healthcare institution without them knowing it. It means that... The attacks are mutating so quickly that you need to have other, more effective methods. We're applying a lot of our artificial intelligence to the cybersecurity problem because you can't know if you're going to detect a pattern anymore; it could be something brand new, a zero-day attack. So now we're starting to say, can we look for patterns in the communication and in the systems that would allow us to identify it even if it's unknown malware, just by the pattern of its functioning or how it's transporting through our different systems? It is a huge challenge for us, but one we are certainly addressing.

Yes, we do. We feel that's the right thing to do for the patient; it improves their quality of care. As a separate note, how many people here use the Apple Health app? If you do, we were the first laboratory to be connected to the Apple Health app. So if you go to it, you'll see your LabCorp lab results in the Apple Health app. We do feel like integrating to share data to support that consumer is critically important.

We have a number of different threat types. I think the privacy threat is much more on the diagnostic side because, as I mentioned, when you go through a clinical trial, it's typically a double-blinded study. Most of the data you have, you don't know who the patient is; you just have a subject number. So there's less value in the data on the Covance drug development side. On the Covance drug development side, the risks tend to be... Areas for us in that business. I think we have time for one more question.

It's interesting because at an executive level, we have these discussions. In our world, the influence of our customer segments is shifting dramatically. I talked a little bit about the consumer now starting to take more control. Physicians that might be accumulating within a health system, so the health systems are becoming very important. The payers are pretty much opening up their networks to both large labs, so they are not the differentiator of where the lab volume is going to go. The big question mark is this question you just asked me about the employer. I think it will be a good thing for us because if you think about laboratory testing, 3% of the healthcare spend is laboratory testing—only 3%—but it influences 85% of the decisions made about your care. So when a doctor is deciding what to do about your care, 85% of the time they're... Lab values drive a lot of key decisions for care. I think employers are going to get more involved. Most people think that if you take out your insurance card and it says Aetna, Cigna, Humana, or whoever, that's the insurance company paying the bill. It's not usually, because if you're a large company like General Motors, you're not going to pay the uplift for the insurance fee. You're going to go to that provider and say, "I want to sign what's called an ASO contract." If you were a small company, you'd be at risk of someone getting a very expensive disease that could put you out of business, but not General Motors, not LabCorp. I have a UnitedHealthcare card, but UnitedHealthcare doesn't pay my bill; LabCorp pays my bill. So the employers are starting to get very interested in how to improve the quality of care for their employees and lower the cost of care for their employees. It is a really important point with a lot of unknowns surrounding it right now. Can I make one quick comment before we wrap up? First of all, thank you very much for taking the time. Second of all, we hire a lot of people from NC State. Some of our best talent. This is a great... You learn different things. It's fun, it's rewarding. You can move the needle on achieving the mission of LabCorp or whatever company you end up working for. But I think it's important for you to know that as the CIO of an $11 billion company, when I think NC State, I think of a top of the top of the heap when it comes to technology graduates. So thank you very much for letting me speak with you. I really appreciate it.