Episode 222

Using Lab Tests to Diagnose and Improve the Healthcare System – Dr. Brian Caveney, Chief Medical Officer at Labcorp

10-05-2021

A fascination with data drew Dr. Brian Caveney to Labcorp, a lab testing and research company which has processed more than 50 million COVID-19 tests and runs more than half a billion medical tests per year around the world. For Caveney, all of that data provides opportunities for insights into how the healthcare industry can improve. As Chief Medical Officer at Labcorp and president of Labcorp Diagnostics, Caveney considers how labs can better analyze their findings, and how to best frame and communicate these findings to healthcare workers and the public. That’s a particularly urgent task in light of the COVID-19 crises, and an attendant crisis of public confidence in the medical profession. “The lab is often in the shadows of medicine, and may go back in that regard after COVID is over,” Caveney tells host Dr. Rishi Desai. But he hopes healthcare providers retain a deeper appreciation for how, if used intentionally, lab tests can bolster patient understanding and trust. Tune in to hear why political meddling at the CDC was so dangerous, the difference between law school and medical school, and how Labcorp managed huge demand for tests amid a fractured global supply chain.

Transcript

Dr. Rishi Desai: Hi, I'm Dr. Rishi Desai. Today, on Raise the Line, I'm happy to be joined by Dr. Brian Caveney, who is chief medical officer at Labcorp and president of Labcorp Diagnostics. Labcorp has been at the forefront of the Covid pandemic with more than 50 million Covid tests to date in the U.S., and it’s the first to commercially market the Covid-19 PCR test. Brian is a nationally recognized health policy expert. In his current role as chief medical officer, he oversees medical and scientific strategy. Before joining Labcorp in 2017, he was the chief medical officer at Blue Cross and Blue Shield of North Carolina. We're very grateful to have this opportunity to speak to him during this critical time in the midst of the Covid pandemic. Thank you so much for being with us today, Brian.

Dr. Brian Caveney: Great to be with you Dr. Desai. Looking forward to the conversation.

Dr. Rishi Desai: You are at the head of a very important organization right now, given the pandemic. I'd just like to rewind a little bit. Maybe you can just tell me, what started you down this journey and your interest in diagnostic lab testing as it is today?

Dr. Brian Caveney: Well, it probably starts when I was in diapers. I think every single person in my extended family is some type of a healthcare professional or involved in the industry. I'm sure my parents were arguing what kind of healthcare provider I would be before I was even walking. So that was a foregone conclusion. I've always been interested in medicine, healthcare, but really more broadly system issues, organizational and structural issues. I was weird enough to do both med school and law school at the same time, and that gave me a tremendous appreciation for the structural issues that it takes to be involved in the system.

I practiced medicine for about 10 years in North Carolina, mostly at Duke University. Then I spent about six and a half at Blue Cross, really thinking holistically about the whole healthcare dollar and how the entire system works. Then I just became fascinated with the use of data, as so many others are, and all of the different parts of healthcare—laboratory and medicine—

essentially create a number. All sorts of really cool science happens, but at the end of the day, the product is a number that a doctor looks at, and hopefully answers some clinical question for a patient. It's really important at the individual doctor-patient level, as you know so well, but really can be used to measure the health of populations, how things change over time, variances between different subpopulations, etc. It's always fascinated me.

Labcorp is I think just a world leader in that, and about two-thirds of what we do is lab testing, about one-third is clinical research globally in over a hundred countries. That just fascinated me, attracted me to come and see what we could learn about improving healthcare, and in particular, helping doctors, hospitals and other healthcare professionals use lab data to practice better medicine, deliver better outcomes, and really understand the patients they're taking care of.

Dr. Rishi Desai: I have a very quick question and then a more serious one. The quick one is... you mentioned you were in medical school, and then you were in law school. I'd just like you to share your thoughts on the cultures between those two organizations.

Dr. Brian Caveney: Well, this is I'm sure mostly a healthcare professional audience. So, I'll lean that way. It is unbelievably different. You know, law school is more like college—just a whole lot more reading—versus medical school, which is just all-inclusive living in the hospital, just living and breathing, a very different culture and environment. 

Medicine and law also have completely divergent ways of thinking. Meaning, you see a patient and you're always getting more and more information to ever refine to the narrowest possible diagnosis. Law is exactly the opposite. You have a specific fact pattern. Something happened, and everything in the law is broadening it out to determine what principles, what laws, what precedent cases could apply to that situation. So completely different ways of thinking, which I think have been very helpful to me, particularly in business life to think about, again, healthcare as a system, versus individual patients. It's critical to be able to synthesize ideas both ways.


Dr. Rishi Desai: That's a phenomenal answer to what I think was a very trivial-sounding question that I asked. So, thank you for that. You know, a point you made that I want to reflect on as well is that, at the end of the day, the deliverable is a number. A lot of times clinicians will look at that number in the context of a range. Labcorp might present a normal range. I've clinically experienced situations where that range is under debate. Maybe one country, England, Japan, or America—maybe there's some difference in terms of what they regard as normal. I'm just curious. How do you establish a normal range? How do you determine what is right for that population, knowing full well that so many clinicians are just going to essentially knee-jerk and say, “well that's normal, that's abnormal,” without taking that full context like you pointed to?


Dr. Brian Caveney: That is such a great question. We have an entire team in our science and technology group who just does that, because we have over 5,000 tests on our menu, and frankly we obsess about getting the reference interval right and measuring it over time. The official accreditation standards through the College of American Pathology, and JCAHO, and some of the others, sometimes only require a very small number. It differs: Sometimes 20, sometimes 120 measurements in a normal population or within a particular diagnosis to determine a reference interval.

One of the great advantages of being at one of the largest laboratories is we have millions of tests per year. We do over 550 million tests per year. So we have huge amounts of data. We tend to massively oversample beyond what is required for accreditation. We want huge databases, to really refine what our reference intervals are, so that we can feel very confident that we're capturing the population means in an effective way. 

What that also helps us do is not just say, “this is a population mean for America, or Canada”— where we have a very strong presence—or some other countries, but really subpopulations. Just as one simple example: We're the first lab that has developed some different reference intervals specifically for transgender populations as they're going through the transition process. It takes an enormous amount of data and a lot of clinical insight to be able to look at hormonal differences and be able to follow that transition so, as you point out, doctors can make better decisions.

Another thing that we're starting to do with certain conditions—I think thyroid disease, and some others: Even if patients have two, three, or more laboratory tests within the normal range, it can go from just above the lower limit of normal and migrate to just below the upper limit of normal. A doctor who's taking two seconds to look at that would say, "Oh, you're normal, Mrs. Jones. Don't worry about it." But we're now trying to apply all sorts of algorithms to say, "No, no, no. It's still technically normal on the population mean, but Mrs. Jones' thyroid number has changed by 19% over the last year. Doctor, you may want to actually investigate a little bit further or do something around that." So, I think there's a lot that we can do. Again, it takes a large database, and it takes the clinical acumen to look at the clinical implications of the number and really help a doctor who's ordering that test derive the right insights from it.


Dr. Rishi Desai: That's so interesting. As a pediatrician, we often talk about children falling percentiles in weight. They still may be within, you know, like you said, 5th to 95th percentiles. But if you're falling among those percentiles, you're not on your own growth trajectory. Something is amiss and you have to catch that. So, I appreciate the fact that you guys have those kind of early warning systems in process, or maybe they're released already. That's really helpful. 

You know, there's a huge surge around Covid and the Delta variant. I'm curious how a company like Labcorp deals with that massive demand for testing, especially as different organizations roll out new protocols. Like testing to get on an airplane is new for the FAA, or testing to be employed on a weekly basis—that's new, and many people are doing that. How do you manage that amazing amount of demand that is there right now?

Dr. Brian Caveney: I can laugh thinking back now to March and April of 2020, but believe me, none of us were laughing during that time when it was just an incredible amount of work to look at the whole process. I mean, just as one example, the PCR test—sort of the gold standard diagnostic test for Covid—it takes about 27 or 28 different things, pieces of supplies, reagents, the machine, etcetera, in order to do one test. We literally had people assigned to every one of them, calling every manufacturer in the world, ordering anything we could. We normally like to hyper-standardize everything so that we can have everything perfectly calibrated. We decided that we would bring up every available platform that existed, knowing that there would be supply chain disruptions because of the border travel restrictions, the border closures, etc. So, we brought up our testing on eight different platforms in 21 different laboratories, just in the U.S.—separate from what we did elsewhere—just knowing that that was going to be incredibly challenging.


The other thing we did was making sure we spread it geographically to all the different labs, because we knew so little in the beginning. If we had a terrible outbreak and had to close down one of the labs or every worker would be quarantined, we didn't know what to do. So, we have geographic differences. All of our labs, even pre-Covid, run 24/7 basically, so three shifts. We would actually have shift A leave the building, do some cleaning and then have shift B come in before we really knew that it was fully airborne, because we didn't want any cross-contamination between shifts in case, same thing, we had an outbreak. We had all sorts of redundancies to be able to transmit supplies all across the country. We hired a couple thousand extra people to help us do all of the manual components that are required to do the testing. Literally, to every manufacturer, we said, "We'll buy any machine, all the reagents, pipette tips, all of those supplies. We'll buy anything that you create, because we just need to build our capacity to be ready for what may happen." That was just an incredible period of time we went through. Thank goodness we made that bet. Sadly, that capacity was needed last summer.

Dr. Rishi Desai: What is your sense on the level of understanding among clinicians, but also among patients, of how to interpret these various tests? I feel like our lexicon has changed dramatically in the last year, and now we talked about positive predictive value, negative predictive value, in a way that like, literally nobody talked about that stuff a year or two ago.

Dr. Brian Caveney: Exactly.

Dr. Rishi Desai: Have you noticed a shift in terms of not just clinicians but patients' savviness around these tests, and how to interpret a positive or negative, and a control and what that means, and things like that?

Dr. Brian Caveney: Much more in terms of inbound inquiries we've been getting from the patient or consumer, the public population. We tried to do an enormous amount of education to frankly, political figures, members of various public health agencies that were non-scientific. to help them understand the data that we were transmitting to them on a daily basis of all of the positives. I guess some of the positives and negatives, you know: Most healthcare professionals have had some amount of training during their training around ordering tests, etc. But it was amazing, the number of conversations I had last year with ordering physicians who really didn't understand what you're saying—test performance, characteristics, sensitivity, specificity, limits of detection. And this was a big one: The timing of when somebody was potentially exposed to a known positive and then the time at which they would test. 

We had lots and lots of people who would be exposed to somebody. They find out during the exposure, and then they want to go get tested right away. Well, it's obviously going to be negative because of the biological process required for the viral replication. So, even that type of education I think could have been handled better probably, not by the labs but by public health communications, the CDC, other groups. That was important. 
 

If I can put one plug again—your audience being so many learners across the curriculum of healthcare professions -- lab is often in the shadows of medicine, and may go back in that regard after Covid is over. But I hope all healthcare providers have a little better appreciation of the fact that when you order a test, or before you order a test, you really need to think through: What is the pretest probability of the outcome, and of all of the range of potential answers that I may get? Whether it's yes or no for a Covid test, or a number for a regular blood test, what am I going to do with this information? What decision am I going to make with my patient around either making the diagnosis, predicting the risk of a future diagnosis, the treatment that I may undertake? It's really, really, really important. 

It's taken for granted -- and I think this last year has really highlighted a lot for my science staff -- how important it is that we're crystal clear on explaining what these tests mean and sadly, the fact that no test is 100% perfect. There is no test that's 100% sensitive, 100% specific in all situations that you might consider. That's frustrating to the public. But I think it’s critical for healthcare providers who are doing the interpretation with their patient to understand that—to constantly refresh themselves on how good the tests are that they're using, and the information that they're getting from them, and then when to retest, or what the next step may be as a result of that information.


Dr. Rishi Desai: It's a good point and a reminder of the fact that, when people get a result, if they're told it's not perfect, the test is not perfect, sometimes people jump to the conclusion of "Well, then what's the point? Then why even do it if it's not telling me what I need to know." I'd be curious to get your thoughts on how you bridge that gap in understanding. Also, tied to this is concordance. "I did these two tests and they told me different things. Which one do I believe? Or are they both useless?" And discordance as well. How do you reconcile when one test says one thing, and the other test says another? As well as a time-based differences. "I was positive today. A week later I'm negative, etc." How do you explain that to patients? Do you get these questions from clinicians as well?

Dr. Brian Caveney: We sure do. We try to make—particularly for the more esoteric testing, a very specialized test—we try to get the lab scientists who run those assays, who understand the machines, who are publishing papers on the different methodologies, to answer those questions for inquiring clinicians. If something looks like it may be a spurious result, if something looks like it may be discordant in some way, it often does come down, as you mention, to the communication between the doctor and the patient—how it was described to the patient. What information was going to be available before the test was done? It comes down to the rest of the clinical scenario. Does it fit or does it not fit the clinician's gestalt of what the probable diagnosis was? And if it does, then they're more likely to follow that pathway. If it's the opposite of what they expected, then is there some other thing that might be a tiebreaker, whether it's another test, another physical exam maneuver, something else that might help determine what the next step needs to be?

I can't remember the exact old saw, but I knew back in med school. What is it? You know, "Treat the patient, not the X-ray. Treat the patient, not the number." Those kinds of things really are important because you can't just look at the number that we deliver back and say, “this explains all things and everything about the patient.” Again, it really comes down to the pretest probability. If you get discordant results, which one was more likely to be consistent with the fact pattern that led to that number coming back? And then what should the next step be for that clinician to allay the anxieties of the patient? 
 

That doesn't mean medicine is wrong and doctors don't know what they're doing and the lab's screwed up. It doesn't mean that. It means, "Gee, your biomarker pattern isn't necessarily exactly what I would have predicted. Maybe there's another layer deeper that we need—in the history, the physical, other follow-ups that we may need to do. But rest assured, I'm going to try to get to the bottom of your situation and make sure we implement the right treatment patterns for whatever your case is.”


Dr. Rishi Desai: One of the things that I was really, in some ways I suppose, surprised by, was how much home testing became a part of our arsenal against Covid-19. Home testing for travel purposes, for work purposes, for school purposes. You've released a home testing kit for Covid. What has that experience been like for Labcorp? Maybe you've done this with many other diagnostics and if so, how does it compare to the other ones you've released that are for home testing purposes?

Dr. Brian Caveney: Sure. Yes, the great news is, we already did have a whole platform, technology, website, and the ability to perform home testing prior to Covid. So that made it very easy. Not easy. That's the wrong word, but it made a smoother pathway for our scientists to be able to take the exact same Covid test that we're doing in our labs, that you would get from your doctor's office or hospital, and be able to do all of the validation with the FDA for the other pieces of it. So, the testing itself has exactly the same materials: supplies, specimen, specimen transport tube, and equipment and reagents in our labs. The only difference is what we validated very deliberately with the FDA on lots of patient samples: The ability for, I'll say the average person, to read very simple infographic instructions and follow it. 

That required us first to be able to validate an anterior nares specimen collection. Instead of a nasopharyngeal swab, which never could be done by the public, just anterior nares. So it looks basically like a Q-tip—three swirls in each nostril. Then making sure that they can get it into the inactivation buffer, close the specimen so it didn't spill, and then get it into FedEx. 

So we FedEx it to the person's house. They do their sample and then they drop it in a FedEx box, or FedEx picks it up, or they can drop it off at one of our patient service centers, and it gets sent directly to our lab. As part of the validation, we checked each of those steps—in particular, people being able to test in themselves. It was very highly concordant with a nurse or doctor-originated swab. 

But then, we also, with the FDA, did an enormous amount of what I'll call specimen integrity validation. Meaning, we flew it across the country a couple of times. We would leave it out on a really hot tarmac for a couple of days. We would put it in a really cold environment, in direct sunlight, just for all of the different things that happen out in the world, in getting it from a person's home, or wherever they would test themselves, back to our lab. That was very highly concordant with again our healthcare professional-originated specimens.

So finally, the FDA validated it, and gave us our FDA authorization in order to market it. We were the first ones to make that available. Now, if you think about access being such a big deal last year…before the federal government started doing those drive-thru sites, access was tremendously challenging. In some states, 30-35% of doctors' offices were closed and telemedicine only—meaning there was no actual way to go get the swab done. So, this was actually a critical access point for people to be able to get tested in the early days. It continues to be very important, as you say, for pure clinical purposes mostly, but also for some other administrative purposes: return to work, or for travel. Some are using it for going to Broadway events, sporting events, or other places where you might need to demonstrate a negative test in order to participate in something. So, it's been very helpful in that regard.

Dr. Rishi Desai: That makes a lot of sense. You know, it brings me to my next question. Maybe you can help us with just taking a step back. What has the Covid crisis revealed not just for the healthcare industry but specifically for diagnostics? What lessons have you learned as a company? Or maybe even insights you've pulled from seeing how things have unfolded, that have come as a result of this last year and a half, two years.

Dr. Brian Caveney: Yeah, that's a great question. The National Academy of Medicine is putting together a very long white paper on what we learned—supply chain, diagnostics, therapeutics, and vaccines—to be better prepared for future pandemic-type scenarios. That will be very important. The diagnostics industry, frankly because it's small as a percentage of the total healthcare dollar, has been squeezed quite a bit by public and private payers. It really wasn't that ready, to be honest, to handle the challenges and to be able to scale capacity.

Most of the labs are very low margins and run very lean, efficient operations. So they weren't really able to scale up quickly at all because there was such a supply chain disruption. Just as one example, the vast, vast, vast majority of all of the samples that we use for many of our different laboratory operations, and certainly for the Covid testing, are not sourced in the U.S. Most of them are global sources, all across Europe, Asia-Pac, South America. We get stuff from everywhere.

Our supply chain is truly universal, which is usually fantastic, and very lean and efficient when there aren't border closures and travel restrictions. We couldn't get pallets of pipette tips through customs that we desperately needed in our lab. So we learned those types of things…building redundancy in the supply chain, making sure that you don't just have just-in-time inventories but you actually prepare and in some cases possibly stockpile for things that are likely to be used, or even in the absence of a pandemic could still be consumed in the normal course of business: easy things like pipette tips, tubes, gloves, masks, you name it. We certainly can learn a lot about the overall supply chain going forward, and make sure we plan for it.

Dr. Rishi Desai: Yes, that supply chain of course has affected the entire economy and has reminded us that this is a very small world in a way. Listen, your insights are amazing. Is there any parting advice you have for our audience of mostly, as you correctly pointed out, health professionals, young medical students, nursing students, and the like.

Dr. Brian Caveney: This one may be a little nerdy but I'm going to do it anyway, given the great audience that you have here. One of the lessons that I've learned over the last year is, while we're driven by evidence-based medicine—and it's critical that we do the right studies and we learn the right things—I think the public health communication around this into the public was a little too focused on p-values and vaccine effectiveness numbers early on and not enough on just educating the public on what's happening, what this means, how you can interpret it. 

We lost our sense of risk management—risk and probability analysis, if you will. What makes me almost more sad than anything else -- and for you, having been an EIS officer, it probably hit home -- I never once in my entire career have ever questioned the fact that the CDC has the best epidemiologists on the planet. They are the source of truth of everything. It was very sad to me that they seem to have lost some of their credibility for other political reasons last year. 

The medical community never before really questioned the integrity of information coming from the CDC. I hope and pray it does not take a generation to restore that, because they are so important to the healthcare system here. They need to be well-funded. They need to be at the center of preparing us for all infectious disease monitoring going forward, not even in just the next pandemic, but flu, STDs are on the rise, all sorts of things like that. It's so critical that we all as part of the healthcare system, understand and support the information sharing that can come from a trusted source like that. Just a sad outcome of the last year is the public anxiety, rightly so, around behavior of being in public has significantly decreased appropriate care-seeking behavior. The number of cancer screenings is way down. The number of people monitoring their known chronic diseases is significantly down. We're seeing that in the testing that we're doing. We're doing some public health projects right now, seeing Hemoglobin A1Cs rise in underserved areas. We're seeing other non-Covid conditions get worse because all of the focus is on Covid.

You've got to think holistically about your patient, all of their conditions, all of the things going on. We need good quality, credible, trusted, evidence-based information that we all can believe in, and use to educate our patients appropriately going forward. Hopefully, all of your audience will do that.

Dr. Rishi Desai: Yes. That's a really important point, Brian. I appreciate you making it. That trust between the different parts of the healthcare system needs to be restored. Like you said, it's been corroded, and I hope that we can work together to help restore that. Listen. I want to thank you very much for your insights. Absolutely phenomenal role you're playing, and your company is playing, in this fight against the virus, and I just want to applaud you for that. Thank you.

Dr. Brian Caveney: Sounds great. I enjoyed it very much. Thank you for what you're doing to educate the public and learners about how they can improve the system.

Dr. Rishi Desai: I'm Rishi Desai. Thanks for checking out today's show. Remember to do your part to flatten the curve and raise the line. We're all in this together.