Episode 7: Is Data Management the Glue of Modern Clinical Trials?
(gentle music) - Good morning, good afternoon, good evening, wherever you are joining us today. I am really excited to be hosting this, the latest episode of our podcast, the Future of Digital Trials. And I'm very excited today because we're actually filming from our live face to face summit here in Boston, and I'm even more excited because I'm being joined by someone who's very important and a very important figurehead in our industry, and that's Louis Torres who heads up programming for LabCorp's FSP Business.
So with that, Louis, let me turn to you and let me first start by inviting you to perhaps share a little bit about background and a more broad spectrum introduction of yourself to our audience today. - Well, first of all, thank you, Richard, for having me, and it's a pleasure to be here. I've been with LabCorp for 24 years, it'll be 25 years now in December. Time flies when you're having fun. And in my current role, I am the head of clinical programming within the FSP Business unit, but in my role, I have the pleasure of supporting the clinical programming team and just navigating through the many challenges that we go through and along the way, coming up with innovation and solving problems and getting very creative. - I like that phrase, problem solving.
I think we'll come back to it a few times in this presentation. But is that how you see your role? Solving problems? - I think that in our roles, there's always something that comes up that is new, and typically you have to really think outside the box to resolve them. It's not a problem, it is just a challenge that you have to go through. - Okay, let's come back to that. I would love to start...
The fact that you've been in this industry 25 years as have I, and you look about 15 years younger than I do, I will hold onto that for another time. But let's go back to when you began. How has data management changed from when you began to where you are today and where do you see it perhaps changing as we continue to advance? - So like many of us, I started in the paper days and I actually started my career as a temp doing data entry. So there was a lot of paper, and I remember when I started doing data entry, all the temps, we were about 10 or 12, and we were in this room that we call Guam because it was so isolated and it would get really hot, I think from the body heat and all the computers in the room. But yeah, there was a lot of paper, a lot of data entry. And then, initially, or not initially, but a little bit after that, then I began to get involved with Oracle Clinical, where I started my career doing database builds and got into programming both around PL/SQL, SQL and SAS programming.
- Now, as part of that, where do you think perhaps data management is going in the future? - Going back to the paper days, I remember we had... If you think back, it was very inefficient because we even had a dedicated resource that would just go around from cubicle to cubicle looking for a case report form. And if a data manager needed to do something with a particular patient, this resource had to go and track the case report form because maybe somebody forgot to sign it off from the file room. We don't have that now. So we have created opportunities for people to move from those type of activities to now performing more of the data management role.
And I think that going into the future, the data management role, and we talk a lot about the data scientists, and I think the data manager's gonna become more of, I like to call it a glue, where they're gonna be using all these different skill sets, programming skill sets or technical skillset sets, working with biostats and all these different departments and groups to be able to perform their jobs. - Let's think about that for a second. You and I, just a few minutes before we started this, shared a little bit about this Rubik's Cube. This is not here accidentally, it's here quite deliberately.
The Veeva branded Rubik's Cube, which you're welcome to take away. This idea of problem solving, which is where you started this conversation and I want to spend some time now, it feels like this is the preeminent subject of data management today. I feel like we as an industry are coming together to solve clinical trials and to find ways to solve problems that just haven't been fixed to this point.
I've just published a few ruminations on solving the Rubik's Cube. It's a little bit unfair that you've had all of about eight minutes to digest that. But you have your own thoughts on solving problems and what this idea of solving a Rubik's cube could mean.
Perhaps I could invite to share a few thoughts there. - Yeah. The first thing that I think about if using the Rubik's Cube analogy is that, well, first of all, it's not going to resolve itself, so you do have to attack it and you have to pick it up and work with it. But one of the things about the Rubik's Cube is that the centerpiece of each side is what determines the color.
So as you're trying to solve the Rubik's Cube and let's say if you try to solve the color green, you have to focus on that centerpiece. I like to think of the centerpiece as people, process, technology. Those are the ones that I'm closer to. The other three, we can figure out what those are, but I think that in order for us to really solve the challenges that we're going through, you really have to focus first on that center piece, the process, people, and technology. - I like this analogy, and I like what you were describing earlier as well about having to solve the Rubik's cube.
And I think you started to stretch the analogy in ways that I haven't thought of previously, which is it's about not just having to solve it as a black and white statement, it's now how do you solve it? How quickly do you solve it? Which side do you solve first? Is there a preference? Does it matter? Does it not matter? Is it actually important whether you solve it in five seconds or six seconds? Where do you think we could perhaps focus on this analogy a little bit as we explore how data management can become a more and more impactful part of the clinical trial process? - I think you had mentioned about that when you solve one of the sides and you're not able to see the other sides when you're looking at the Rubik's Cube, is that you tend to not really know what's happening on the other side. So I think that when we look at data management and some other things that we have been talking about, it's not just about solving it as fast as you can, but it's also making sure that you are working together with all the other sites and the other groups to make it happen. - You're not the only person trying to solve the Rubik's Cube.
And one of the message I've taken away as we've tried to build this story is, wherever you stand in relation to the Rubik's Cube, you can only ever see three sides. You have to either pick the Rubik's Cube up or you have to move in order to see other sides. And as much as therefore a change you make could impact someone else, other changes could negatively impact you. And I think that's an interesting thing for us to talk about when we talk about problem solving as data management.
I'd also just wanna pick up on something you said, which is, as you start to solve the Rubik's Cube and you think about solving one or two sides, you have to be prepared to tear those sides down before you can progress. Otherwise, it's impossible to solve the rest. If we bring that back to what you do and problem solving within clinical trials today, where do you think the biggest challenge actually sits in your universe today? - I think that in our industry, there's many ways of doing things.
I would say that one of the bigger challenges that we have that I see, especially speaking from the FSP perspective, is that we have to not only work with multiple sponsors, but then we also have multiple standards. Sometimes we have to use our own SOPs, sometimes it's their SOPs or there's a hybrid. Those are some of the challenges with not having a true standard in the industry. I really like what SCDM is doing with introducing standards and guides so hopefully we'll be able to be there one day.
But even even through that challenge, sometimes it creates some opportunities where we are able to make recommendations to some of our customers and say, maybe you can try this technology or you can try this different process. Even with those challenges, sometimes there's also some opportunities that come up. - Which perhaps leans to advances.
Then let's come back to what you and LabCorp are doing as an organization. If you think about problem solving, you think about the challenge of people, process, technology. What are the big things in front of you right now and what are the things that you are doing to try and solve this Rubik's Cube? - On the people side, we were fortunate that about five years ago, we expanded into Latin America, specifically in Costa Rica. I'm actually the country manager for Costa Rica too. But we were able to expand programming and data review and also a testing group there. So it created a bigger pool of resources.
That has really helped during this time because we are having challenges with resources and we know what's happening with the world that really has sort of shrank the resources in the market. So this has really helped us. And then on the technology side, at LabCorp, we have our own homegrown systems, but we also leverage systems that are off the shelf. Recently, in 2020, we went through the enablement of Veeva CDMS, and not just with Veeva CDMS, but now we're also getting very involved with CDB. We actually had a head start where we're supporting Veeva by providing resources there and we wanna be heavily involved because we think that that is going to solve a lot of the challenges within data management and the technology.
But I think that one of the key components about these challenges is that you always have to look at your systems and always very frequently do an assessment to see where you at. And if there's any opportunities or any candidates for improvement, you have to look at new technology because as you know, technology is always evolving. - And I think that's important because when you think about that people, process, technology, I think if you go back to when we started, it was a five-year or a 10-year plan and technology survived a long time because it was a big investment.
Now I look at the newer technology ideas and it is plug and play. It's going to rotate much faster. I don't think we are going to be talking about systems we talk about today in five years time. Probably not in three years time, maybe not in two years time. That evolution is huge.
But that change in speed and cadence must bring pressure on people. It must bring pressure on your business to deliver faster and faster every day. - Yeah, there's always that pressure. I remember we were trying to build databases in 12 weeks and then that continued to come down and now even eight weeks, it's too long, we wanna build them sooner and we wanna get to the two weeks. And I think we can get there, but again, we have to solve some of the challenges that we're facing. And again, that's why I think it's so important that you're looking at your internal systems and make sure that they'll be able to support being able to get into those goals.
- I think historically there was a scientific vision for what clinical trial excellence looked like, and adaptive trials, platform trials, bucket trials is perhaps the upper limit of that today. But technology and perhaps data management has failed to deliver that vision because science always wanted more than technology could deliver. I genuinely feel we're close to perhaps even reversing that. Where do you think data managers come into that? And do you think it's important that data managers pick the technology or is it technology picks the data managers? - We need data management and the end users to drive technology, they have to be involved from day one. Because otherwise you can end up implementing a technology that one, may not be ready, or two, the users are not comfortable using it, and then it fails. I'll share a quick story if I may.
So my mother, she's 86 years old and one time, a few years ago we had to come over to and live with us, I'm a technology guy so I thought, well, let's put Alexa in her room so that she can control all the lights and the fans and be able to call us from her room into our room. And we quickly found out that Alexa doesn't speak Spanglish. So it didn't really work. But a few years later, now Alexa has Spanish capabilities, and it's great.
Now she's able to communicate with it, she's able to use it, able to call us. It's an example of if we exclude the end user, you're going to have potentially a great technology, but it's not going to work because the users are not able to use it. And we see that a lot with even visual analytics and creating dashboards. If the end users are not part of it, it's not gonna work.
- If there's any consolation, my wife's from the northern England and her and Alexa do not have a close relationship. There we go. If you think about that sensible adoption of technology, if I can put that phrase into what you just said, it is something about total experience, and that's why I like this idea of science and technology have to be in lockstep, not kind of pulling each other in different directions. So I think it's very critical what you said. I guess though, as a CRO, there's another layer or level of responsibility in that as well because you are delivering on behalf of a customer or a sponsor with an expectation that your decisions are always correct.
Is that true, do you think, even today? Or do you think there's more of a risk sharing approach or do you feel the burden of that responsibility? - We definitely share the burden. I think that if we think about data quality, we can easily say that data quality is the sponsor's responsibility and they have the ownership of data quality. But no one is really going to say, I know LabCorp and I don't think any CRO is going to tell a customer, well that was not my responsibility. You're the owner.
You have to make it personal. Not just data quality, but everything that you do, it almost becomes personal and you have to want to really be able to do the best that you can and really take that ownership. I think that's the key, is making it personal. - When I think about data management, it's relatively easy for me to think about two use cases. You've gotta get a study up and running.
So in the case of data management, build a database. And two, I've gotta be able to use your data. So reach decisions generally through reporting. Now, you happen to be responsible for both of those things. So let me ask you first, what do you think data managers would like from the experience of building a database? Do they want to just do it themselves or do you need a broad range of people in order to deliver a database? - I would say that it depends on the organization. I think that data managers want to and they definitely have to be involved in the database build process because again, I see them as the glue.
If they're not the ones building the databases, then they have to be very close with the programmers and everybody that's involved. And then on the reporting side, I think that data managers need to have the ability to configure their own reports. I don't think they wanna be doing any programming because again, their roles are not going to be specific on any areas. I think they have to have this overall ownership of the entire process. I think there's a part of the data manager's responsibility around data exploration.
So they need to be able to configure their own reports, ad hoc reports, and be able to do that so that if there's any particular data that they have to drill down into or somebody's requesting, "Hey, what about this data?" They're not necessarily asking other people and looking around to create that report. - If I'm thinking about reporting, it's a natural next step to think about data cleaning and delivering clean data, which is our primary role in the end of the day. What role do you think CROs play in that chain of delivery in terms of bringing clean data to sponsors? - So again, going back to the taking ownership, I think as a CRO, we always take that ownership to provide data quality and clean data so that when we turn it over from data management to SDTM and and submission, the data has to be super clean. So I think it's definitely something that we need to own. - Super clean is an interesting one for me.
I guess when we grew up, where we started, perfection was the only acceptable standard. I mean, I remember just the concept of a database unlock, you would need like 15 layers of approval to do such a thing. It was like a career limiting step, let's be honest. But now we use different phrases. We talk about acceptable error limits, we talk about veracity in data.
These are terms I didn't grow up with, but they're now more commonplace. And I guess that's partly due to the explosion of data, the explosion of data types. Is that kind of true of your experience as well or do you still think there is an element of perfection as the the only acceptable standard? - I guess when I say super clean, the data that's critical, then that has to be super clean. I know that there's risk base, you're not going to be able to look at everything. Not like the way that we used to do back in the days. And you really shouldn't because again, we're looking into speeding clinical trials and you're not going to be able to look at every data point unless you have to.
If you have to, then you look at it. But if you don't then... But yeah, I don't think we need to. We're certainly not at the point where we used to be years ago where everything needed to be reviewed. - So the last couple of questions for you.
One, you talked about this relationship between LabCorp and Veeva that we've been very fortunate to partner on, and partner's the word I'd just like to take your guidance on. It feels like there's a special relationship. We've absolutely leaned into your organization to help guide in places.
Not specific to Veeva, but just CRO to technology. What are you looking for as a partner and what do you hope to get from the multitude of technology companies you work with? - What has worked with working with Veeva is that... Veeva as a technology partner, 'cause I also like using that word... As a CRO, we have a lot of lessons learned and we can share that experience with Veeva and technology partners. And I think it's beneficial not just to the vendor but to the industry, when the vendors do listen and when they are able to implement some of the enhancements that we recommend and some of the recommendations that we make. It's not that we're making recommendations just because we want to, it's because we have gone through it.
We have years of experience, years of working with different sponsors, and that's what has been so successful working with Veeva is just the way that Veeva has listened to the recommendations that we have made. - You know where we go next? - Yes. - It's the final question. I've put this to everyone, so I'll put this to you.
In the scenario that we could give you a magic wand and you could wave that wand and you could fix some problems, maybe there's some things you can't do today that suddenly you could, and maybe there's two things that you wish you could consign to room 101 to never be done ever again, what would you wish for. - Does the wand only work in our industry? Can it be something else? (laughing) - I don't wanna cause any marital issues, so you can decide to answer however you want. - Probably the first thing that I wanna think about is...
We talk a lot about automation, AI and machine learning, and that has done a lot of the automation and I think that we're gonna be doing a lot more, but one area that I wish that we could have spent a little bit more is natural language. I think I would have systems and platforms to be able to have the capability to allow the users and data managers to just be able to query the databases in a way that is just like the way that you speak to Alexa or that you just type a requests. That's why, by the way, I like what CQL it's going into and I think there's going to be some good things that are gonna happen there. But I would say that, and then the second thing, again, SCDM is doing a great job with implementing a lot of standards and guidance.
I don't know, maybe one thing I would like to see is maybe more standard pricing, if that makes sense. Maybe some type of MSRP on data management services. What was the other side? What I can get rid of? - [Richard] Yep.
- I've been involved with both the EDC and reporting and we do see a lot of, still, some manual listings and there's some sponsors that they still like their SAS listings and Excel, and I would like to get rid of that once and for all. - I have to say, I have this reaction to Excel in the same way. My personal room 101 would be just paper in all forms and that includes Excel.
Excel for me is just another form of paper, same with Word. In a clinical trial process, they just have to go. So I'm absolutely on board with you on that one. - It was funny when we would do all these demos on visual analytics and you have these beautiful dashboards with graphs and sometimes probably the question that we will get the most is, can you export it into Excel? (Luis laughing) - Well, can you print it off? - Yeah. - Perfect.
Well, I know we're up on time. I need to say thank you very much, Luis, for joining us today, for sharing your thoughts. I know the audience will be very, very receptive and grateful for what you've done, but can I give you the final word, anything you would like to share with our audience today? - And thank you for that, it's been a pleasure being able to do this. Probably the one thing I would say is, always find joy in what you do. Always put a smile in your face when you're working. I think if you're able to find joy in what you do, you can accomplish a lot of things.
It doesn't matter the technology, the process, but if you're not enjoying yourself and having fun, it gets very difficult. - What a great way to end. Thank you, Louis, and thank you, everyone, for taking time to be with us today. - Thank you. (gentle music)