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

Hello, everyone. Welcome to today's live broadcast applied design analytics to pressure test protocols, minimize avoidable amendments and mitigate study risks. My name is Andy Studna, Editor of Applied Clinical Trials and I will be a moderator for today's event. We are pleased to bring you this webcast presented by applied clinical trials and sponsored by IQVIA. I would 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. Learn more at www.iqvia.com. We have a few important announcements before we begin. This webcast is designed to be interactive, and we encourage you to ask questions during the event. [Operator Instructions] You can enlarge the slide window by clicking on the small icon in the bottom right corner of the media player. The slides will advance automatically during the event. And if you have any technical problems viewing or hearing this presentation, please click on the question mark help widget in the top right of your presentation window. I would now like to introduce today's speaker. We are pleased to be joined today by Tammie Nguyen. Tammie Nguyen is a Director of Clinical Planning and Design Analytics at IQVIA. She is a pharmacist by training with a background in biochemistry and cell biology. Nguyen has more than 10 years of clinical drug development experience and has pressure tested more than 150 protocol assessments using historical and real-world data to highlight areas for protocol optimization prior to execution. Thank you for joining us today, and Tammie, if you would, please get us started.

Thank you, Andy, for the introduction, and thank you all for joining today's webinar. I'm going to start off with a personal story. About a year ago, I had a baby and my husband and I decided, we both needed to update our cars to something safer for our son. After tons of research, I decided on something I thought was pretty practical, my Toyota SUV hybrid. I'm able to fit a car seat and our 2 large dogs in the trunk. My husband, on the other hand, decided on a sporty BMW luxury car because he's always wanted one. We can barely fit the car seat and the only thing that fits in a small trunk happens to be, of course, his golf clubs. Now when it comes to the maintenance of our cars, we also differ there. I follow a well-planned schedule because it helps me plan for related expenses. I never let it get to the point where the service light appears, and if it does, I usually get it fixed within the week. And when I do bring my car in for routine maintenance, I always request a 12-point inspection versus the courtesy inspection they always offer, where it's the tire changes or rotation and the oil change. My husband, on the other hand, is of a different mindset. If his car is running fine, there's no need for maintenance. Basically, he said, if it ain't broke, don't fix it. Let's just say, he's had a maintenance light lit in his car for the past 6 months and has yet to investigate what the issue is. So sure. My method has upfront cost in both time and money to complete these checks and routine maintenance appointments. But from my experience, it helps my car last much, much longer. The last car I drove was for nearly 15 years, and it's still in operation today. And it should also reduce my risk of massive mechanical issues that likely will cost more money out of pocket just later on. Much like my car example, so is protocol design. It may end up costing you a lot more money and time downstream, if you have an amendment or experienced recruitment or retention issues versus taking the time to optimize the protocol upfront. In today's discussion, I'm going to focus on some of the challenges our industry currently faces with protocol execution as well as the solution to this challenge and highlight the value of this solution. So essentially, what do we know to be true of our industry today? What are we currently doing about the issue? And where do we want to keep moving towards? So what are some of the challenges that our industry faces today? For the past year, IQVIA has been fortunate enough to participate in TAS recent protocol amendment study. In fact, this month, TAS released the results of their study in their impact report, which examines protocol amendment experience and its consequences. The key takeaways are that the prevalence and mean number of protocol amendments continues to rise across all phases of studies with the most recent study showing nearly 77% of clinical protocols require at least one substantial amendment. These protocols are generally larger in size, so larger amount of patients, sites, countries and are more complex in their trial design. For example, protocols with an amendment have nearly 25% more endpoints and 16% more eligibility criteria on average versus protocols without an amendment. Also, more than 70% of these amended protocols require modification of study procedures and trial designs. And when we look at the TAS impact report in 2015 on cost and time, these amendments are ranging a cost between 140,000 to 535,000 on average, but 3 or more months delay in timelines. So when we rush through protocol development, hoping to adhere to milestones set forth, but disregard robust upfront optimization of the protocol design, the consequences later may end up causing the longer timelines we were trying to avoid to begin with. Applying design analytics to proactively identify issues with the trial design upfront can be a strong solution to these industry challenges. At IQVIA, we will plan multiple analytics to pressure test protocol design decisions. Today, I'm going to highlight just 3 of them due to the time constraints. So those are the 3 big bubbles that you see on the slide. If there is interest, we may add another webinar later on to cover more. So the other smaller bubbles that you see on the slide, including our design analytics that leverage our real-world data to assess eligibility criteria and study procedure frequency. Design analytics really can be applied anywhere along the protocol development continuum during the development of study concepts to a draft version of the Synopsys of protocol and even after an initial amendment. Specific analytics can be applied at different time points, but the level of impact will vary at each stage. For instance, you might interrogate real-world data to understand and define your target, patient population and eligibility criteria during the early specifications of the study design concepts. Whereas you would reserve analytics to assess and optimize the scheduled activities for a later draft of the Synopsys protocol. We really recommend applying design analytics to pressure test your design decision, at least once along this continuum. Let's go ahead and dive into the 3 analytics that we at IQVIA apply most often. Since the TAS impact report is hot off the press, I'm going to keep running with the stats from that publication for the purposes of our webinar. 72% of implemented amendments require modification to study assessments. The design consistency analytic is one of our analytics that reviews this element. I'd like to think of this analytic as an internal audit of the protocol, ensuring everything is consistent, meaning we are confirming a line of evidence from objectives to endpoints, endpoints and eligibility criteria to procedures and vice versa. Without this clear line of evidence that the probability of success of your trial may be compromised. You may experience higher risk due to missing data or spend resources collecting wrong or unnecessary data. Let's explore a couple of examples on the next slide. The first example to the left is the Phase III COVID study. We apply design consistency analytics and identified 10 different inconsistencies in the draft protocol. Just a reminder, not every section in the protocol is reviewed in this analytic. Our focus is on key components of the protocol that have the most impact. So that would be your objectives, end points, eligibility criteria and schedule of assessments. Some of the key findings revealed by applying this analytic in this example include extraneous COVID tests at the end of the study that did not support any end point, another endpoint that was unclear, and there was no clear direction on what procedures needed to be done in the case the patient became hospitalized during the study. Overall, the sponsor updated the protocol to remove 50% of the inconsistencies that were identified. And the second example to the right is the Phase Ib neurology study. Our consistency analytic revealed 16 different inconsistencies. Areas of increased risk within the sponsor's protocol included findings such as procedures missing from the screening visit, which could lead to potential issues in identifying your target population and a misalignment in the procedure scheduling where the endpoint specified it being measured at day 27 versus the scheduled procedures indicating a measurement at day 42. The sponsor subsequently updated the protocol to remove all identified inconsistencies. In addition to driving consistency within a protocol, assessing potential patient burden can also be helpful to optimize trial design. When a trial is more patient-centric, recruitment and retention are likely improved. 65% of implemented amendments require some sort of modification to patient selection withdrawal and treatment. As part of our protocol assessments at IQVIA, we address patient centricity and design in multiple ways. Direct patient interviews and focus groups are always valuable methods to gain patient perspective. But these approaches can be consuming and expensive to undertake for every protocol, especially during the design stage. As such, for every protocol assessment, we might published patient and advocacy group feedback to uncover qualitative insights to complement our design analytics. An example I have for you is on decentralized clinical trials, a very hot topic, as we all know. Our insights show both pediatric patients and their [indiscernible] may consider home visits and ePro technology to be an important aspect of participation. Incorporation of these decentralized central trial solutions may positively impact their willingness to participate and even protocol compliance. Qualitatively, we assess protocol-specified procedure frequency compared with the patient's standard of care using real-world data. We also scored the protocol via patient burden algorithm derived from a survey of patient willingness to participate in a clinical trial with specific design elements. In this Phase III neurology example, we applied our patient burden analytic to the draft protocol, generating an overall patient burden score of 25. As shown to the right panel, most Phase III neurology trials revealed with our patient burden analytics have a burden score between 10 and 19 inclusive. The sponsor subsequently removed many of the PK visits and other general visits, thereby reducing the burden score by 40%. After the changes in sponsor made, the patient burden is now more aligned with similar trials. Another popular topic in our industry right now, diversity and inclusion. Last year, we enhanced our patient burden analytic to highlight differences by race and ethnicity. Interrogating the perspectives of diverse populations during the trial planning stages can help validate the study design plan and help make adjustments early to increase enrollment potential of these different groups. And the example I have on the slide today, the Hispanic respondents expressed more change than did other groups in their willingness to participate, if a daily diary was required. Knowing this allows us to think through how we can make the diary more accessible for this population to complete and also help them understand why the diary is needed and how it contributes to the success of the trial. Acknowledging differences in participants' preferences and being prepared to discuss and reassure participant groups that they have been considered in the trial design is just one way to address diversity and inclusion in clinical trials. You'll notice that everything that I'm pointing out about these analytics is that findings are brought forth for a discussion to help you during the trial planning stages of your trial. Another analytic that often promotes robust design discussions is our competitor trial analytic. 70% of implemented amendments require modification to the trial design. In this analytic, we use proprietary and public databases of design characteristics. We're comparing our protocol design with those of recent trials in the same indication and phase for similarities and differences in things like choice of endpoints, eligibility criteria and even some of the design aspects of the study whether to use a placebo or not. Additionally, we review clinical outcome assessments used in approved labels in the indication of interest against those in the accessed protocol. This provides insights on patient-reported outcome measures that can often differentiate the product from others that have similar clinical benefit. In these examples, we compare endpoints to competitive trials and highlight where there are differences. In a Phase II non-small cell lung cancer study, the competitor trial analytic identified that neither co-primary endpoints specified in the protocol were like those in comparable trials. The sponsor subsequently revised the protocol from BLR to ORR to better align with similar trials. To the right panel, for a Phase I prostate cancer study, our analytic identified that PSA response was included in nearly 70% of competitor trials, but was not being used in the protocol. By adding PSA response to the final version of the protocol, the sponsor became fully aligned with some of the trials and choice of all their endpoints. Nonetheless, sometimes differences from competitor designs are deliberate and differentiating. While other times, they may not be. The goal of the analytic really is to drive exactly that discussion. Up until this point, I've shown you examples of how each analytic can impact design decisions. Well, what happens when you combine them all together? At IQVIA, we apply a combined set or a package, if you will, of design analytics in the form of protocol assessments to deliver meaningful insights tailored to the protocol at a single point in time. On the left, you can see that IQVIA's set protocols, 96% have had areas in protocols that were unclear or inconsistent. 83% had design elements that increased burden and 40% differed from competitors on trial design. On the right are 2 examples that I want to share with you today. In the first case study, a Phase III in rheumatology, both anti-TNF experienced and naive patients were planned to be enrolled. The trial design kept the percent enrolled of anti-TNF experienced patients to 30%. Whereas when we look at the real-world data, it indicated over double the number of patients that would be anti-TNF experienced. The sponsor then revised their protocol and the study design by separating it into 2 separate protocols, one for the anti-TNF experienced patients and one for the anti-TNF naive patients. These new designs were submitted to the FDA and subsequently approved. In the Phase III CNS study, patient burden was lowered by removing extraneous procedures from the screening visit and other visits during both the double-blind treatment phase and open-label extension. Missing objectives, endpoints and safety procedures were also added. The cost savings of these design changes was potentially about 10% per patient for the double-blind phase and 30% for the open-label extension phase. If the trial did enroll all the planned patients through the open-label extension phase, the potential cost savings here could be about $7.8 million. So you can see that combining the application of multiple design analytics increases the potential over impact in optimizing protocol design. In summary, a high number of protocols require at least one substantial amendment, which increases costs and timelines. And these protocols are typically more complex in their design. Protocol assessments have proven to reduce patient burden like when a large global pharma company removed extraneous procedures that resulted in not only a decrease in patient burden, but a decrease in trial costs by $100,000. It's able to improve accuracy of study budget like when an Asian-based emerging biopharma company had 2 key safety procedures missing, which could have led to an underestimation of the budget by as much as $2.2 million. And it helps sponsors prepare for regulatory submission like when 13 regulatory authority concerns were identified for a sponsor prior to submission. Protocol assessments are very much like my 12-point car inspection, but instead of fine-tuning a car, you're fine-tuning your trial design. If there's one key takeaway I'd like for you to remember from today's webinar, it is this. Investing and optimizing your protocol front by appliance design analytics can reduce risk and ultimately increase the probability of success of your trial. If you have any further questions, my contact is here, and I would love to further expand upon our discussion from today. I'll hand it back over to Andy to get us started with Q&A. Thanks.

Thank you, Tammie, for that informative presentation. [Operator Instructions] Okay. So we will jump to our first one. In the patient burden case study, the findings you shared with the sponsor like removal of PK visits may be essential to meet study endpoints. How do you ensure the integrity of the trial remains intact with your recommendations?

That's a very valid point. So thank you for this question. I think there's a few things to consider here. First off, the patient burden analytics that you're speaking of is only one aspect of the protocol assessment. So if we look at protocol assessment holistically, for instance, in parallel with conducting the patient burden analytics, we also say, take a look and assess the protocol with design consistency. We may identify that a procedure, in this case the PK measurement, does not support any endpoint or objective in participating in the trial, but the PK may not serve any purpose in the protocol design or the frequency at which is being measured and in the schedule assessments actually exceeds what is needed to meet the study endpoint, which is the exact scenario in the case study. The next thing to consider is that the patient burden algorithm scoring and surfacing of these burdensome elements is really meant to drive conversations and discussions. That's really the case for all of the analytics and findings we bring forward, in this case the patient burn analytics focuses on how protocol requirements might be seen from a patients perspective, so that we can identify what change might be possible. From my experience, though, by this point, we're thinking through risk mitigation strategies and solutions to put in place. And the third thing is the protocol assessment isn't just a tool that you insert your protocol in, press the button and yield an answer and move forward with that answer. We also rely on subject matter experts, medical advisers, and operational leads to provide their expertise regarding the highlighted findings. So the data and the design analytics is not the end all be all. It really works hand in hand with the expertise and a complementary approach.

Great. Thank you, Tammie. Our next question is, do you have any data that shows influence of protocol assessments on recruitment and retention decrease in the number of amendments, less questions from regulators or anything else further downstream than what was presented today?

So I think this is a really good question and one that we receive often, so thank you for bringing it up today. Showing how protocol assessments can make an impact downstream, in my opinion, is rather challenging. Operational variables like the ones you mentioned, I think it was a reduction in number of amendments, recruitment rates and maybe, let's say, even start-up timelines or study duration are not solely dependent on the types of findings, protocol assessments identify. And it wouldn't be completely accurate to draw a 1:1 correlation here. Impact to me is multifactorial, right? So for example, recruitment and retention may be impacted by multiple factors, not just by patient burden alone or not just by eligibility criteria alone. And a correlation does not always equate to a causation. So let's just say for purposes of discussion today, we're able to draw a correlation, so to understand if IQVIA's protocol assessment actually made an impact, studies would need, so this is all an imitation of being a CRO, studies would have need to be awarded to us or executed by us, and us being IQVIA, to have insight into such operational variables. So in fact, we're in the process of correlating protocol metrics right now. For a few clients that we've partnered with and such as patients complexity and patient burden and with operational [indiscernible] such as recruitment and retention. And we're able to carry out this activity because we have partnered with them and we have insight into those operational variables, but regardless, they're still, in my opinion, value and surfacing for discussion before protocol finalization. The findings protocol assessment raises related to, say, unsupported endpoints and procedures, maybe potential causes for patient burden and to some end points compared to competitor trials all have an impact. So I think it's -- right now, we're at over 90% of about 1,000 protocol assessments that we've conducted, where the results that we shared with the sponsor, and we've actually tracked the impact of the final protocol, so just meaning that we are comparing the final protocol to the draft protocol that we reviewed. There were changes related to the assessment findings.

Great. Thank you. Our next question is, is this IQVIA data informed protocol assessment?

Yes. So that's -- yes, these are the examples that I covered today are from the data inform protocol assessment, which is a tech-enabled service. I think a lot of IQVIA on the webinar today and some of the sponsors that we've worked with also known by the acronym of DIPA, or DIPA. Our DIPA typically includes admin among the three analytics that I covered today plus electro-world data analytics that I have alluded to in the introductory slide and more advanced analytics that are bespoke to specific therapeutic areas or study design. I think the DIPA is currently our most popular IQVIA solution for leveraging design analytics but it's not our only method and it continues to evolve and grow as we add more analytics, build out our current analytics and continue to scale up and outwards. And we're actually in the process of -- we're automating some of our analytics right now to make it go even faster than what we're doing in analytics. So if you have any further questions about the design analytics. The ones I speak about today, feel free to reach out to me directly if you'd like to learn more.

Can this be applied to real-world trials? What about for trials with rare diseases?

I think a simple and short answer here is yes. We can apply design analytics to pressure test nearly any type of trial design, regardless of phase indication. But I think the key here is that the level of impact or potential impact will vary depending on the phase and the indication. So let me give you an example. The consistency analytic that you saw today can be applied to any observational study or rare disease trial following our usual practice. So really no change in how we would assess using that analytic there. But when we're looking at the patient burden analytics, it may still be conducted for observational studies as some of the design elements are relevant and may still cause burden. Well, there's other elements in an observational study that might not be relevant in the analytic, for example, discontinuing current treatment because observational trials don't have treatment, and this would be included in the burn calculation. But for any trial design type -- our analytic provides directional insight that we typically also supplement with other patient prudent information. And usually, this is qualitative information. and analytics to support specific to the opportunity at hand and particularly for rare diseases where there might not be that much data out there, we like to complement it with qualitative information. And then again, with competitive analytics similar to the consistency analytic, there's no change in how we would conduct this analytic regardless if it was an observational trial or a rare disease trial, although in the rare disease trial, you may have less competitor trials to compare your trial to.

Great. Next is, do you have insight into why there are differences by race and ethnicity? Why are blacks, Asians and Hispanics more or less concerned about x,y,z?

So our patient burden survey was really designed to determine whether there were differences by race and ethnicity and to quantify that difference. We didn't really focus on so much the why, although we're very interested in this and would like to include it in a future survey for sure. But we did in parallel, look at other published survey results as to why that may draw some hints as to why. For instance, there's this misconception that Black or African-Americans are not interested in clinical trials, and that has been the misconception for a while now. But current research finds that it's quite the opposite and as well as our IQVIA findings. Trust and clinical research has improved in this group, and there's a desire for "people like me" to be represented in the clinical research. On the other hand, though, there's other publications that also highlight concerns of minority communities, which make it difficult for them to participate or lowers their willingness to participate. And this is usually due to the comfort level due to maybe a lower level of education, maybe the socio economic status, maybe not having insurance or a primary health care provider, language barriers and even a lack of clinical trial sites near their home to impact us.

Great. Our next question is, how do you make a thorough critical and objective reading of a study protocol?

So I think the key here is that not one design analytic will provide you a thorough review of the protocol, but it really is about the combination of all the design analytics together. So objectively, let me give you an example. We use data like real-world claims, electronic health records, et cetera, to provide insights on perhaps key inclusion or exclusion criteria may impact patient availability of [indiscernible] here or looking at how familiar patients may be with a required procedure and to quantify the added burden to the frequency of that procedure against their usual care or standard of care. But when you look at something like, say, the design consistency that I mentioned, we're only focusing on key areas like the objective end point procedures and eligibility criteria and not the entire protocol. So it's not a full examination of your entire protocol. Like we wouldn't be looking at the statistical plan. And like I mentioned previously, we do rely on subject matter experts, operational leads for their expertise in regards to the objective findings we bring forward. So there is a little subjective element here, but as I mentioned before, it's -- I see it as a hand-in-hand approach of something that's a complementary approach.

Next is, are protocol amendments the same as change orders?

So when I think of change order, I think of it as something that you're updatings to the protocol after you've received a regulatory approval. So if that's the case, yes, I would consider a protocol amendment a change order. And in the slides that are reported from the TAS, I'm talking here about substantial amendments, and TAS defines substantial amendments in their report as any change to a protocol on a global level, requiring internal approval followed by approval from regulatory bodies.

All right. Next one up is, what is the turnaround time for analytic to run on a draft amendment and receive back the results and recommendations?

So that's really -- I mean, I do quote over time here, but generally, it takes us about 2 weeks once we actually start the assessment process to go through the process of doing the analytics, reviewing with the internal team and then having the output to a sponsor and reviewing it with you. Now there may be iterations. We might have a kickoff meeting to better understand the needs and that all adds to the time as well. So I generally say, we try to aim for 2 weeks, but it could be depending on the added elements, about up to 4 weeks.

Our next question is, is this a standalone service? Or does it need to be clubbed with others?

So there's 2 ways I see this. Is it a standalone service as far as can you pick and choose the analytics that you want to apply? The answer is, yes. So there are cases where we might do select analytics, pick 2 or 3 analytics rather than from the full of the menu of design analytics. But you're going to get more impact, I feel, if you do club all the analytics together because they each bring something different to the table when we're looking at design consistency versus looking at the patient burden or evaluating the eligibility criteria using real-world data. The other way I see this question is, can this be clubbed with other services upstream or downstream? And the answer is, you can use it in silo or you can actually add it and have even more impact. Let's say, we're working with, let's say, feasibility downstream. You can assess your protocol before approval, and then we can punt it out to feasibility to do a more detailed look at some of the things that we brought forth or coming up with risk mitigation strategies based on the burden that we've identified using the data informed protocol assessment.

Great. Our next question is going to be, is this a tool that we use or a service that you offer? How long does it take to assess a protocol?

So I think this is similar to our previous question. As far as duration, I would say, again, we like to aim for about 2 weeks to do once we start assessing the protocol, but it could be depending on the complexity, how many iterations that we do, if there's a kickoff meeting may extend out to an additional week or 2 after that.

Great. Now our next question is going to be, how do you handle protocol confidentiality when doing the assessments?

So that's a great question. As far as confidentiality, we don't share any of the information that we've done for one sponsor with the other. We do have -- let me give you an example here. With the patient burden, we do have, as you saw on the screen, benchmark data. So we're comparing your protocol against other, say, 30 or 40 CNS studies. And so the scale -- we do have a limitation on the number of sponsors in that mix and the number of studies before we can share that information. So we've actually, for the benchmark data, had to take some time before we could collect as much information to keep it blinded. So as long as we meet those restrictions and those limitations, we can do benchmarking data such as you saw the patient burden, but we do not share your information or anything like that of the findings in our DIPA unless we have met those restrictions and limitations for confidentiality.

Great. Thank you so much, Tammie. With that, we are going to wrap up. I want to thank the audience for attending and for participating in today's event. I'd also like to thank our sponsor, IQVIA, for making today's webcast possible. We would like to ask everyone in the audience to participate in a brief survey, and this survey will appear on your screen after today's presentation has ended. You will receive an e-mail, alerting you when this webcast will be available for replay. We invite you to forward that announcement to your colleagues who may have missed today's live event. We hope to see you all next time. Enjoy the rest of your day. Goodbye.