IQVIA Holdings Inc. (IQV) Earnings Call Transcript & Summary

Hello, everyone. Welcome to today's live broadcast, Connecting Real-world Data and Domain Expertise to Enhance Trial Design and Planning. I am Lisa Henderson, the Editorial Director of Applied Clinical Trials, and I'll be your moderator for today's event. We are pleased to bring you this webcast presented by Applied Clinical Trials and sponsored by IQVIA. And now I'd like to share a statement from our sponsor. IQVIA is a leading global provider of advanced analytics technology solutions and clinical research services to the life sciences industry. IQVIA creates intelligent connections to deliver powerful insights with speed and agility, enabling customers to accelerate the clinical development and commercialization of innovative medical treatments that improve health care outcomes for patients. With approximately 82,000 employees, IQVIA conducts operations in more than 100 countries, and you can learn more at www.iqvia.com. We have a few important announcements before we begin. This webcast is designed to be interactive, and we do encourage you to ask questions during the event. [Operator Instructions]. And now I would like to introduce today's speakers. We are pleased to be joined today by Tina Hearon; Sheetal Telang; and Nathan Sommerford. Tina Hearon is Vice President of Clinical Trial Analytics and Strategic Insights at IQVIA, and she's been -- has over 18 years of experience in the pharma industry. In her current role, Tina leads a global team of 5 therapeutically aligned groups supported by a global analytics team. With a strong background in data analytics, Tina has been instrumental in driving data-driven decision-making and insights across clinical trial programs. Sheetal Telang, Vice President and America's Head of Therapeutics Strategy at IQVIA is a senior clinical research professional with over 17 years of experience in various leadership and operational roles in the pharma, CRO and data science industry; and Nathan Sommerford is the global lead for IQVIA's AI for R&D solutions. He has more than 20 years of pharmaceutical industry experience, both in consulting and industry. In addition to his experience with artificial intelligence and machine learning and R&D, Nathan's areas of expertise also include real-world evidence study design, statistical analysis and execution and electronic medical record analysis or EMR. So thank you all for joining us today. And Sheetal, could you get us started?

Thank you, Lisa. Good morning, everyone. My name is Sheetal Telang, and I am Vice President of Therapeutic Strategy leading the Americas Therapeutic Strategy Group at IQVIA. Today, with my colleagues, Tina Hearon and Nathan Sommerford, we will be presenting Connecting Real-world Data and Domain Expertise to Trial Design and Planning. Next slide, please. So I want to start today's webcast by briefly talking about some of the pressing needs in the research and development industry. So for several decades, our industry has predominantly relied on previous experience to help drive operational decisions. Over the last few years, we have increased our reliance wherever possible on any available data sets that we have. But these data sets have historically always been separate and siloed and have been also interrogated separately. So as a result of that, the relationship and the interdependencies between data sets was not recognized and couldn't be incorporated into any types of operational planning. So the end users who also relied on the data were siloed. So we often ask questions on data based on the risks of a protocol as it occurred. And we receive the information at very different time intervals. So there wasn't a consistent flow of when we ask the question, when we receive the information and the types of questions we ask in a proactive manner. In today's discussion, we're going to focus on solutions for some of the challenges that our industry faces within R&D. So starting with one of our largest challenges today, developing the right protocol and how we do that by applying design analytics and how design analytics proactively helps identify concerns with clinical trial design, including trial entry criteria, so what makes the patient eligible to participate in the trial, assessments and end points which are ultimately going to drive your clinical study report and take it to the next step of your development and reviewing the protocol for consistency, which is really important to avoid cost associated with amendments. The next piece is getting to the right country and site strategy and this is very important. And the use of data and analytics with some practical examples that we will share today and case studies will demonstrate how we are using it to help drive early planning decisions on a clinical trial, enabling diversification not just of countries but also trial populations. We will also discuss examples today of integrated evaluations where large data sets could be very quickly collated across multiple data sources, so we can have actionable insights using AI and ML capabilities. Combined, of course, with expertise. And the third piece is evaluating enrollment rates using data and analytics, so we can predict a study time line, which from my operational background will tell me that it is the single most important driver of a clinical trial cost. Next slide, please. So let's talk about some of the key business issues and how real-world data and analytics can be used end-to-end in a clinical trial setting to provide meaningful as well as consistent and predictable insights to solve for some of the issues that we face. One of our biggest challenges that our operational teams face today on a clinical trial is the protocol amendment process. Last month, Tufts released the results of their most recent protocol amendment study in their impact report, which examines the protocol amendment experience as well as its consequences in the clinical trial lifestyle. The key takeaways that are -- the key takeaways here are the prevalence, and the mean number of protocol amendments continues to rise across all phases of a study. Also, more than 70% of amended protocols require modification of study procedures and trial design. So it makes them important protocol amendments. And when we rush through protocol development, hoping to adhere to some milestone at the very start of the development cycle, but we disregard robust upfront optimization of the design. We ultimately result in causing longer delays to time lines in the future when we realize that changes should have been made at the start. So we want to start with avoiding changes to the future in protocols. An upfront and more deliberate investment that you make at the start to apply design analytics, to identify issues proactively helps in the long run to reduce the likelihood that you're going to have a prolonged time line of development and also increase burden both for sites as well as the cost of the development of the molecule. So let's move to our next business issue with study time lines. Statistics from Tufts show that 48% of clinical trials missed their enrollment time line. If we deep dive into what causes these time lines to extend, often, we start with the very first operational step of country and site identification. This results in reduced months of enrollment, which ultimately results in an extension of the protocol time line. So one solution to this key risk is the use of predictive tools to identify country and sites using real-world data and analytics that are more interested and more likely to succeed given the clinical trial to reduce the number of nonenrolling sites and have sites that are motivated to participate on the study. So when you select the right countries and approach the right size, what you really enable is a faster site identification process more motivated sites with the right populations that ultimately leads to starting them faster and have studies with enough time to enroll patients with no delays to your ultimate recruitment time line. So throughout this presentation, we're going to share specific case studies where the optimal use of integrated real-world data and analytics, coupled with domain expertise, has resulted in achieving trial time lines. And last but probably the most impactful to the overall development time line is a key issue that 80% of the trials in our industry face today. The delays to recruitment. There is usually multiple drivers that result in recruitment delays, and it's often a combination of factors, some of which we've discussed above. But in addition to the data using tech and insights to further drive faster and more cost-sensitive patient enrollment is going to be an important factor for us as an industry to consider moving forward. Next slide. So now that we've talked about the risks and some of the key issues that we face in clinical trials, let's do a deep dive into the types of design analytics. So at IQVIA, we have multiple design analytics to pressure test protocol design consistency. Today, I'm going to focus on only one of them due to time constraints. So we'll talk about the design consistency analytics. Firstly, I like to think of this analytic as an internal audit of the protocol so to make sure everything is consistent. As an example, we're conforming objectives to end points. We're also confirming eligibility criteria to the study procedures and vice versa. So without this clear line of evidence, the probability of success of your trial may be compromised. You may experience higher risk because you're missing some key data points or you may be spending a lot of resources collecting data points that are unnecessary and you don't need for your study. Next slide, please. So let's do a slightly deeper dive into how we can apply design analytics to protocol decisions. So design analytics can really be applied anywhere along the continuum of protocol development. So during the development of study concepts, you can apply it to a draft worksheet, you can apply it to a synopsis, you can apply it to your final protocol or just your protocol before you start your submissions and even after your initial amendment if you have to have one. But specific analytics need to be applied at different time points, but the level of impact of when you apply them will vary at each stage. So for example, you may interrogate real-world data to understand and define your target patient population and eligibility criteria. You're in your very early phases of study design concepts. But you would reserve analytics to optimize your schedule of assessments for a later draft of protocol or synopsis. But what we recommend is applying design analytics to pressure test your design decisions at least once before you have a final protocol. So with that, I'm now going to hand over to my colleague, Tina Hearon, to take us through some of the global data assets and their applicability to the clinical trial strategies. Tina, over to you.

Thank you, Sheetal, and hello, everybody. Again, thank you for joining us today. Appreciate everybody making the time. So if we can advance on to the next slide. In today's data-driven world, the strategies for clinical trials, as Sheetal was mentioning, benefit greatly from leveraging global data. And within IQVIA, we have a wide variety of data. So the depth and the breadth is there as well as the unique variety of sources. And why is that so important, the variation? And it's because it provides us that insight into the full continuum of care for a patient. So across the multitude of countries that you see here and the sources listed here, we have prescription data, the claims data, sales data. We can truly understand the life of the patient, which then allows us to gain those better insights into the who, what and the why diseases are treated and managed certain ways. And of course, when we look at data at a global scale, the access varies, right? So a lot of things affect that, the countries, the laws, the health care systems as well as the coverage just as a geographic -- when we are at the national level, pretty robust data assets when we get into greater detail, when we get to things such as like the subnational levels. So we'll get into more specifics as we move through the presentation. However, as you can see, all of these assets, they play a really important part in our decision-making and execution. So the key takeaway here really being, we have a lot of data, right? But we also have the expertise in working with these global assets in order to drive that successful application across our various use cases. So if we go on to the next slide, we talk about the data, the importance of access, the variety, being able to work with the data. But we also need to ensure we're using the right data, right? We're asking the right questions in order to determine the right insights, again, another really important piece in our decision-making and our execution. And depending on the goals or the objectives of a given study, we may use all of the assets. We may use some. It's really going to depend on the questions that we're trying to answer and what are the key insights that we're finding within that data. As well as I had mentioned before, the country, understanding the data at the country level is important. For example, we know in certain countries such as U.S., the U.K., EU 4, they have very rich data assets available for us to use. The granularity and the specificness of the data does vary at that country level, therefore, the type of analysis possible also varies for us to use by country. So having the right data and the right insights then allows us to get to those actions. What are the actions that we need to take in order to optimize the trial strategy? We all know that meeting enrollment targets is often a challenging task. So to Sheetal's points earlier, in the planning stage, if we're selecting the right mix of countries and sites in the planning phase, that plays a crucial role in overcoming some of those challenges. Let's move on to the next slide, and then we can discuss some of those key actions that we take, we consider when we're developing a strategy. And I think this is a build slide. So you can build it up all the way if you would like there. But before we build a strategy, we really ask ourselves some very basic questions about the study. We want to understand about the protocol, right? Does the protocol have operational risks and dependencies? This is going to help us derisk the study as we move through the strategy. Who is the patient and where are they? Where can we find them? This is going to give us that geographical view. Where's the patient? Who's treating them and where are they being treated? This, again, is leading us into the site, that data-driven site identification process. And then, how do we convert the patients that where we're identifying them into the clinical trials? How do we get them engaged and into the clinical trial and convert it into a trial subject? So we aim to answer these questions using what we refer to as our connected intelligence, right? We're bringing, as we talked about the right data, the right analytics, bringing the technology into play. We all know in today's world technology is a huge part of how we do our jobs. It changes the way that we think about how we do our jobs and helps us enhance the way we do our job, giving us not only more efficient ways to do planning and execution, but giving us uses of a holistical end-to-end process as well. So bringing all the data, the analytics, the tech, ultimately, we bring all that together and then empower it with our deep domain expertise in health. So our strategy then become largely data-driven and backed by the evidence. We also have that element of expertise, which makes our strategies development that combination of an art and a science. So if we drill down a little bit deeper into this approach that you see on the slide, where do we look right, to get some of these answers to the questions that we're asking? And we talked a little bit on the past slides about the importance of not only having that data, the expertise, but the variety, right? So firstly, it's important to consider many factors when we're looking to answer these questions. So if we look at when we're selecting countries and sites, we may look at different things such as the patient demographics, of course, very important, right? Patient is the key piece of a successful clinical trial. So we look at patient demographics. We look at the prevalence of the disease, what's the competition, historical recruitment rates, regulatory and infrastructure, very important factors. But we also will look at -- we'll also consider language, culture, accessibility, also play a crucial role in the patient enrollment. Secondly, as we move into site analysis, how do we choose the best sites? We use several techniques for this selecting the right mix of countries and sites. And again, data and analytics playing critical parts in that decision making. So how do we determine what is the right countries to go to, what's the appropriate number of sites? What countries and sites have the highest potential for the patient enrollment? All these questions are constantly through our heads as we're planning and looking within the data. In these areas, we'll again look at things such as the trial experience, the past performance. What is the study-specific eligibility criteria? What are the treatment paths and the patient's journey that they're on so we can really understand what's happening in the real world, and again, who was treating those patients. So in conclusion, optimizing the trial strategy, using the right data and insights gets us to that point of selecting the right country mix and the right site in order to achieve those crucial milestones and meeting our enrollment target and ensuring the success of the clinical trial. In these steps, we talked a lot about these steps being important in the planning phase, but they're also very important as we're conducting the trial. We're continuously holding ourselves accountable to monitor and evaluate their strategies to ensure that we're having the best possible outcomes. So if we move on to the next slide, this is really just a great example of that patient-first approach. Looking at the patient, we understand with the help and the support of those data and analytics, what is that journey that the patient's on? We want to understand it from the initial symptoms all the way to when are they having their first visits, when are they being diagnosed? What are the treatments, what are the events that are happening along those appointments? Who are they seeing at those appointments, and so forth. So you can see very straightforward and basic example here on understanding all of the different pieces that are really important in understanding how a patient is treated within a certain indication. If we go to the next example, right, we move on to -- we start with understanding -- if we go on to the next slide, please. If we go -- we understand the patient, understand the treatment, the pathway, looking at all the information we have, and then we want to understand who's treating the patient. And how do we identify those physicians that have -- within their specialties that have the highest -- the patient potential, excuse me, for that protocol-specific population? So in this example, and let me just lead off to with starting the number of physicians that we have here is not reflective of all physicians specifically doing -- conducting research, it's looking holistically at treating patients. So we have our team for this exercise, take a look focusing in on our U.S. medical claims, primarily looking at ICD-10 codes to determine patient density amongst physicians. Very straightforward, there's other available methodologies that we can use, looking at like drug [ assets ]. However, for this specific protocol, it involves combination therapies. So the data is not always -- as we know, the data is not always reflective of the treatment paradigm. And so our decision internally was let's focus on ICD-10 codes to drive the strategy. So as we were looking at the data we used. In this example, we use our claims data, which is approximately like 75% coverage, we wanted to gain insights into a few things. First, the prevalence of Afib, ASCVD, ACS and AIS. We just wanted to understand it, the market. We then looked at the data to help us identify who are the physicians with the highest volume of patients meeting that criteria and who could potentially be best positioned to recruit effectively for the study. So we identify -- we identified between 5 million to 7 million patients with a medical claim for the ICD-10 codes of interest for us. And there are -- the insights coming from the data, there were a variety of physicians that are caring for the patients with those indications that have large volumes of patients. Mainly focusing being on the cardiologist, right? So our focus for Afib and ASCVD, we've identified cardiologists as the primary investigators, and we would still look at the alternate specialists for referring opportunities or partners. For the ACS trial, we saw that both cardiologists and neurologists would be a great focus for our primary investigators. So as you can see, the insights that the data is telling us, naturally, in -- on a cardiovascular study, the obvious or our traditional way of doing work would have been to just focus on that one specialty. However, doing this exercise, it provided us information that there could be some valid points in exploring some other specialties to help moderate patient volumes and minimize the competition at certain sites. If we move on to the next example that we have, so you'll see the theme using that data, the insights, how to apply it and the variety, right? So you can see we're carrying that through from the patient to the physician, to the location. So in this example, you'll see a lot of different things on this slide of how we carry that through, right? The data is now informing us where the appropriate facilities that we should be identifying, that have those physicians within those specialties treating those protocols eligible patients required for the site identification. So again, as we were talking about through this presentation, bring the variety of the right data, the right insights to drive those actions and our planning process. And you can see on this example, we showcase U.S. and ex U.S. So we did the U.S. example that we talked about on the previous slides. But we also conducted that similar exercise in other countries, France, England and Brazil using some of those databases that we talked about earlier in the presentation, and again, where the coverage and the methodology varies. So in the countries where we don't have access to the claims data or ICD-10 level information to arrive at that patient density or those insights at a site, we use prescription data to identify those facilities. So we have an algorithm and a methodology that we would use that helps us to determine who are the high ranked physicians and sites using the product to be able to point us in directionally to the appropriate centers that are treating the Afib, ASCVD, AIS patients and so forth. So with that, I'm going to hand it over to Nathan to walk us through some other ways that we've had successfully used the data and the application for planning trials.

Thank you, Tina. And just to echo my thanks for everyone joining the webinar today. We really do appreciate your time. I'm going to walk you through some of the case studies that we've utilized in a real-world setting. If we can move to the next slide. But before I get into the details of the case studies themselves, I just want to kind of wrap up what Sheetal and Tina were saying around utilizing real-world data to drive your dynamic trial strategy and site insights. So what we've learned over the many years, applying these methodologies and [ policies ] is the ability to really understand the granular level of the information and how to apply that to optimizing the trial design and also giving operational directions for when managing trials for responses as well. So we can -- moving left to right on the boxes on the slide, we obviously can apply that granular intelligence so we can look globally to help clients understand which countries to invest in from a trial design and operations perspective. But we can also take that granular real-world data to the site level and help those clients understand the patient potential at site level. But also one of the big drivers in this is competition. So it's very easy to go after the big sites that may have the biggest patient potential. But an understanding of those sites, they might have 3 or 4 similarly placed protocols in front of a sponsor's protocol that's going to make it difficult for the recruitment of those patients at that specific site. So understanding the competitive landscape at the site level through the application of real-world data allows us to potentially pivot the trial design or the countries in the country basket to optimize that sponsor's trial design moving forward. So along the bottom row, global patient density obviously gives us the ability to look at patient potential by disease and prevalence. We can do that at a global scale, utilizing real-world evidence or we can take it all the way down to the site level and help our sponsors understand the patient potential and the treating physicians at site. One of the biggest drivers at the moment is the diversity requirements, and I'll come on to a case study that tackles just that momentarily. And one of the things that we started to apply here over the last few years is the proximity site identification, it's one thing to understand your referral sites, it's another thing to understand which sites have patient potential to refer into those referral sites to try and optimize that trial's enrollment rate of patient potential within specific geographies. So moving on to -- if we can skip the next slide, please, and then moving directly on to the case study, that would be great. Thank you. So our first case study looks at the application of artificial intelligence and machine learning to diversity and inclusion within clinical trial enrollment. So our expert data scientists collaborated with a large biopharmaceutical company to address not only their recurring problem, but the recurring problem in the enrollment for many clinical trials is of underrepresented populations and their lack of engagement and enrollment in clinical trials. So the data science team utilized what we call FRMM, which stands for Fair Ranking with Missing Modalities. It's designed to optimize and maximize racial diversity in enrolled populations. And through reinforcement learning, it optimizes the trade-off between performance or enrollment in this case, versus patient diversity. But it gives the flexibility to this specific biopharmaceutical organization to adjust those parameters at the site and country level as well. We tested this on 4 enrollment models for this specific client across a 10-week period to address these challenges of site selection in enrollment for underrepresented populations, and suffice to say, the model performed well across all of those 4 enrollment models, across different indications as well. So we were looking at oncological indications, prostate cancer being one of them. And the engagement went very well, and the team are now looking at utilizing FRMM on a number of their clinical trial programs moving forward. So this is a nice case study that gives us an understanding and the intersection of the application of artificial intelligence and machine learning in the clinical trial design space. And that's one of the things that we are really pushing on at IQVIA at the moment. to help our sponsors, to educate our sponsors in where to apply our artificial intelligence and machine learning to get the most value from. And this is one of those case studies that prove that out. So moving on to the next case study, please? This one relates to many of the aspects that Sheetal and Tina were mentioning, the clinical trial analytics. So utilizing real-world evidence in the Alzheimer's disease space. We worked with a large biopharmaceutical organization who are rebuilding their neuroscience capability, and they wanted to explore how the application of real-world data could give them better insights into feasibility, patient enrollment, patient density across a number of geographies. So we created the analysis and the approaches of 5 countries within 3 regions. We utilized things like biomarker data and lab data to give them a full picture of the moderate cognitive impairment subindication of Alzheimer's disease. We managed to achieve that strategic vision for the client. We built the dashboarding and gave them the competitive landscape from a country ranking perspective. It's very well received and is forming a cornerstone of the evidence package that's driving the rebuilding that neuroscience space for that organization. But again, this is a nice real world example of where the utilization of real-world data can really provide those insights, those granular insights not only into trial competition, patient populations, patient pathways, an understanding of the competitive landscape for assets that are competing with the specific organization as well. And we try where possible to round out that view, utilizing real-world data to give that 360 approach. And then last but not least, we have a final case study but I'll spend a few minutes on it. If we can move to the next slide, please. And this one relates to a Phase II rheumatoid arthritis study with a mid- to large pharmaceutical organization. The study was conducted in Europe and was a proof-of-concept study for the use of a study drug in a new indication. On this study, we were able to use IQVIA's connected intelligence capabilities to assess current treatment patterns, revealing the highest potential markers for the study and pinpointing sites in these regions. In addition to utilizing the connected intelligence to identify the best countries and sites, the project team, echoing what Sheetal was saying, leverage its domain expertise and was very involved in study startup. The combination of IQVIA's connected intelligence, domain expertise and study delivery led to, as you can see on the right, a higher than overall recruitment rate than past Phase II RA studies and 22% less nonenrolling sites. This effectively led to recruitment completion 2 months ahead of schedule and a 26% reduction in recruitment time. Real-world data, again, echoing what's been said throughout this presentation, is a core component of helping to optimize the design of a trial and the operationalization of the trial. And this ultimately led to significant time line impacts and recruitment completion for this particular study. So moving to our final slide and just closing out this webinar today. Bear with me. So just to summarize the utilization of real-world data to drive optimal trials for patients and sponsors, and we've pulled up 5 key things. If we can move forward, please. So informing clinical development choices to mitigate operational risks. I think we can all agree the presentation has shown that utilizing real-world data in informing clinical development helps to not only mitigate operational risk from when you're designing a trial but also when you're looking at which countries to go into, which patient populations to position the trial in. And that helps from an integration of complex and sometimes competing data points as well. So the application of AI and ML in this space helps us to bring in huge amounts of data and really draw the value from that information for our sponsor partners to help them understand which areas to go into geographically, which patients to look at. And again, that leads into the optimal country site mix, which should, by definition, lead into better enrollment rates, more successful trial completions and effectively getting drugs to patients much faster and more successful. We can also involve patients at the site level, with an understanding of that patient pathway. So not only can we look at this stuff from a global scale, we can go right the way down to patient pathway, understanding and patient pathway optimization to help our sponsors understand how patients are being treated at the moment, how that can be folded into the protocol design to ensure that patient burden is lessened when recruiting patients onto their specific trials. Ultimately, this increases productivity with a focus on critical business tasks and improved time lines, as I mentioned, to get drugs to patients faster. And with that, that brings us to the end of this webinar. Thank you very much for your time, and we'll turn it over to Q&A.

Excellent. Thank you so much, Nathan, and Tina and Sheetal also for very informative presentations. [Operator Instructions] So let's go to our first question. Our first question for Sheetal, I think. Is this process used for every study you work on or -- I'm sorry. Sheetal, are patient journey maps a useful way to capture and present real-world patient data used in planning and designing trials?

Did you want me or Sheetal?

Sorry, Sheetal.

It is one of the -- it is one of the data and analytics points that we use is the patient journey maps, and I'll give you a really quick example. If you're working on a trial in stroke, where your enrollment really for the patient starts in an emergency room setting, we know that the emergency room setting is chaotic, and we want to limit the amount of data that we're collecting in that setting. So what we would do is we would help enhance the protocol to only collect what you need. The other example is investigators, right? Who is the right investigator in an urgent care setting? Is it the ER doc who won't meet the patient again? Or is it the specialty doctor who will treat that patient longer term? So these are just a few examples of how we use the patient's maps and the patient journey to be able to use that as we strategize for a trial in the very beginning.

Excellent, thank you. And our next question, is this process used for every -- that's the one I asked before, sorry. Is this process used for every study you work on? Or does it only apply at certain phase studies? Sheetal or Tina?

Yes, I can take that. So it does. So we use it in all trials with patients. So we don't use it in all these volunteer studies, but for every trial that recruits patients from Phase I all the way through to our real-world trials, we use this data. Yes.

If I can just add to that as well, Lisa, we utilize it on our full-service trials, but we also utilize it for potential sponsors that are running their own trials. So we can apply this intelligence, a lot of this intelligence, for sponsors running their own trials who even might be working with another CRO as well. The decision was made a few years ago now, the importance of our intelligence and our connected intelligence needs to be not democratized but open to potential clients to work with us regardless whether it is full service or stand-alone.

Excellent, thank you, Nathan. So our next question, can you explain what you mean by patient density?

I can take that one if you would like, Sheetal. So how we define patient density is the patients that are meeting our indications that we're looking to research or the protocol-specific criteria. So the number of patients that we're seeing in that real-world data that meet that criteria is how we define that.

Excellent. Our next question, how do you handle data privacy concerns? Has that impact your data quality and subsequent results? I'm not sure who wants to take that one.

I can take a stab at that, if you want. So IQVIA being the size that it is, we have data privacy teams that work across all of our data providers. So by the time that the data gets to our teams to work upon, the data privacy concerns and regulations have been adhered to. So a lot of the GDPR questions and potential issues, they've all been rounded up before we're able to utilize the data. Having said that, when we do use the data, especially from a real-world setting, we are very diligent in the understanding that the outputs we provide our potential client partners cannot be back engineered with regards to the identification of patients. So they're given at such a level that is informative, very informative and directional, but there's nothing in there that's potentially identifiable from a patient perspective, even to the point where in specific areas, we might have to blind patient counts and utilize ranges. So we're very diligent in understanding of the data privacy issues. But thankfully, from an IQVIA perspective, a lot of that data privacy work is done before we get to work with the data. Happy for Tina or Sheetal to add anything to that if I'm missing something.

I was just wanting to also add into that -- I think you handled it -- you answered it very well from like the IQVIA and how when the data comes to us, but also to Nathan's point, being having the very large [ volume of ] data that we have, that is always top of priority for us at that global enterprise level. But each of us as employees, we're held accountable to that in our day-to-day jobs, how we're utilizing the data and the efforts and stuff. So privacy is always top of mind, top priority for all of us IQVIA using data.

Yes. I was recently at a conference and yes, the data privacy issue is top of mind across all of pharma also. So that's excellent that you guys have addressed that. So our next question, do you have an automated methodology to match similar studies for enrollment rate mining?

I can take that, and yes, we do. We use it quite routinely. It's called our protocol similarity tool. And what it does is it matches similar studies and then there's obviously a component over that of domain expertise that we apply. Nathan, I don't know if you want to add anything. I see you nodding.