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

Hello, everyone Welcome to today's live broadcast, Clinical Trials Moving From Site to Home—Lessons Learned from Digital Health Technologies. I'm 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. I would now 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. So before we begin, we have a few important announcements. This webcast is designed to be interactive, and we encourage you to ask questions during the event. [Operator Instructions] I'd now like to introduce today's speakers. We are pleased to be joined today by Dr. Tapan K. Raval and Tracy Smith. With the background of medicine and clinical research education, Tapan has more than 18 years of experience working in the CRO industry, managing centralized cardiac safety operations, data management as well as creating end-to-end solutions for using data generated using medical devices as part of the safety and efficacy endpoints for clinical trials. And with over 12 years dedicated to preclinical and clinical trial research, Tracy has supported hundreds of trials in various capacities. Tracy has been with IQVIA for 4 years and is currently managing a team in project management as well as supporting key clients as a strategic operations lead. In her current role, Tracy supports both strategic conversations around device selection as well as streamlining and developing operational processes. So thank you all for joining us today. And Tapan, would you please get us started?

Thank you, Lisa. Start sharing my screen. Good morning, good afternoon, good evening, everyone. We, from Connected Devices' team have had the opportunity as well as the experience of applying lessons learned that we have learned during the execution of the trials while applying digital health solutions into the clinical trials. Briefly covering the agenda that we are going to talk today, we'll very quickly go through the historic updates of around how technology adoption within clinical trials has occurred, past, present and what is a probable future. I would like to share some updates around important recent industry trends when it comes to Digital Health Technologies as well as clinical trial perspective. I will be providing a brief insight into what really are digital biomarkers, what are digital endpoints and how do they move from -- how do they apply within clinical trial zone, then we'll focus on clinical trials from moving from site to home using the DHT and best practices. When it comes to the next portion, we'll cover a certain amount of -- digital amount -- digital biomarker endpoint strategy, data generation as well as collection strategy and data management. The whole aspect or how we are trying to divide this particular discussion today is to provide nuances and to provide certain amount of good application of how Digital Health Technologies have been successfully implemented within clinical trials and what else we have learned from our journey of being completely present at the site to moving to the patient's home. Coming on to the scenario around technology adoption within clinical trials, there have been instances, and I wanted to share certain scenarios over here that in the past, we did have small devices, medical-grade devices, collecting physiological data, which ranged from ECG devices to spirometers to glucose monitoring to blood pressure monitors. All of them have seen evolution where we are seeing devices being shrunk in size as well as most of these devices having the capability to transmit -- collect and transmit the data remotely from being at subject's home. And by -- when the patients themselves collect this data. So when it comes to the scenario, we have very well-established devices as well as technologies around for collecting data for ECG, glucose, blood pressure as well as spirometry. What we are experiencing now -- of course, we have seen from COVID perspective as well that there is an uptick in getting vital parameters collected from home, actigraphy collected from home. And while these are being utilized more and more, we are also seeing evolution in the name of precision actigraphy and specific data, which is individualistic in nature and not a generalized one or -- and this is something which -- again, the benefit is that it can be collected from home and it can be transmitted via Internet so that anyone present looking at the data centrally, remotely can easily access it and make [ informed ] decisions. Having said that, when it comes to the future, from Connected Devices' perspective, what we have observed and are experiencing is that there are multiple voice biomarkers. There are facial imaging devices. There are handheld imaging devices as well as EEG devices and so many such newer technology that is coming into market and it is getting evolved by the day. And I literally mean that because there are multiple devices that are being approved and launched within the industry, within the marketplace on a day-to-day basis. Having covered this, I think I wanted to share certain important industry updates from a clinical trial perspective. When it comes to using the wearables and Digital Health Technologies, I would like to say that at this moment, we have seen that there are more than 130 studies that are actively using wearables in some form or the other. That's something that is really encouraging. And when it comes to this particular aspect, I wanted to share some nuances around how the manufacturers, regulatory agencies as well as pharma companies are moving hand-in-hand towards this particular evolution. From a regulatory point of view, there is a large deal of acceptance of data from new technology that is coming through. This is something that we are seeing from increased direct from patient data collection avenues as well as Digital Health Technologies for remote data collect acquisition and clinical investigations. There are these kind of guidance documents that are being released by different, different regulatory agencies. This is where it is very clear and evident that acceptance of this particular data is there. And having said that, all the regulatory bodies are also moving towards encouraging for development of novel digital technologies because there have been instances where it has been identified that this is something that is going to be extremely beneficial for us in the future. While we talk about encouragement, the regulatory agencies are looking at diversity, looking at inclusions. They are focusing and promoting creation of new algorithms where high volume of data is collected. And at that point of time, there are also nuances that are coming across on once collected from across the globe, how do we time synchronize the data, how do we aggregate the data from multiple devices and try to bring out meaningful outcomes? With the encouragement perspective, I think from the next bit of line from pharma as well as CRO industry, this is something which is also important that while the development of the technology is going at a rapid pace, the CRO as well as data collection and eSource-related platforms do not lag behind. So that's where there is an increasing demand for having software or having solutions which are device agnostic in nature, and which directly collects data from the medical-grade devices and which in turn will reduce the patient as well as site burden. And while we do that, of course, because there is so much of data that is coming up, we have to make sure and we have to keep in mind that -- which is there is a very clear line between where the software will become a normal data collection software versus a software, which will be related to or which will be tagged as software as a medical device. So those kind of considerations have to be kept in mind. From industry perspective, again, to summarize, there is acceptance of this data from new technology. All the regulatory agencies are encouraging pharma companies as well as manufacturers and CRO industry to team up together, try to create different, different novel endpoints and then apply it in clinical trials. And having said that, do not forget the benefit to the subjects and how do we create easy pathway for the subjects as well as the sites to execute. With that said, I'll briefly get into the discussion around what are digital biomarkers, what are digital endpoints and what do we need to keep in mind when we are moving across this particular journey. I start with on the left-hand side of the screen with talking about digital biomarkers. So for example, we have to very clearly articulate that any physiological parameter, which is an outcome generated from a device which gives data in digital format, the parameter outcome is what is also termed as digital biomarkers. When we move into the clinical trial space, we call something -- we have some words which are called as digital endpoints. Now when it comes to this, how I or how we are terming it is that whenever we are deriving meaningful outcome or meaningful output by combining different digital biomarkers, those are digital endpoints. And in addition to that, even if we are looking at an individual digital biomarker, whenever there is a data comparison that we do for before and after in order to accentuate or in order to determine how is the disease progressing that is something that are also called as digital endpoints. To give simple examples, any medical device which gives heart rate, respiratory rate, blood pressure, temperature, blood glucose, all those devices are medical-grade devices and the output that they give is termed as digital biomarkers. When it comes to the endpoints, if we are creating a difference or a combination of output, say, for example, when it comes to infectious disease. If we are creating an output, combining the heart rate, respiratory rate as well as temperature and blood pressure to determine if the patient's infection is deteriorating or is the patient's condition -- overall condition becoming better, that is where it is called as a digital endpoint. So if we create a combined output by combining different digital biomarkers, that's one example. And the second example is if, for example, there is a medicine which has an impact on heart rate of a person. And if we do a comparison of heart rate in that person before and after giving the medication that as well becomes a digital endpoint for that trial. While we speak about this, I wanted to also classify these digital endpoints into two basic types: one, which have prognostic value, which we'll talk about how is the disease progressing; and the second is a predictive value, which will lead to some action, and there are instances where it will determine before an occurrence has happened that this particular occurrence may happen, so please take care, take necessary steps to avoid that occurrence. That's how I would classify that. Now both of these types are used in clinical trials. Certain ones for exploratory endpoints, certain ones for commercial use, which have a commercial value as well. And while we talk about all of these, any particular software or any particular platform, which has the capability -- or a mobile application, which has the capability to derive any kind of a behavioral pattern, activity or treatment decision would fall under software as a medical category -- medical device category. So this is something that is on a very high level, I'm talking about, but we'll have to keep in mind all this. Once we get into the zone of -- or have spoken about the digital biomarkers as well as endpoints, we'll get now deep dive into the focus area for the discussion today, where -- when we talk about how are we moving the clinical trials from home -- from site to a patient's home using these technologies, what are the best practices? I would like to start this particular aspect that any such kind of a program, key point to consider is deriving the digital biomarker or endpoint strategy, and we call that as a protocol or a program endpoint strategy. This becomes very important because, one, we need to very clearly articulate which are the biomarkers that are required for the next steps to go in line. In addition to that, I will now come up on -- talk about the second aspect, which is also important, is how do we manage the data generation as well as collection activities. And in that, we will cover today vendor-related aspect, device shipments-related aspects, data collection-related consideration, training as well as oversight. And when we come about the last piece, it is everything about the data. So managing the data, cleaning, aggregating as well as the final analysis become the focal point for a successful trial, and we will be talking about certain best practices across all the 3 -- all these particular points. Moving on to the next one. I'll speak in detail about best practices that can be adopted when you are talking about your digital biomarker on endpoint strategy. In this case, what is very important and what we have learned is that for a study team, it is most important to get your study objectives clearly laid out. Once we have that particular thing completed, mapping it back to the best suited digital endpoints and biomarker will become important and easy. And once we do that, the downstream activities of selecting the best digital endpoint becomes easier. So that's where I would suggest from a study objective perspective, map the objectives with physiological parameters, evaluate the regulatory acceptance of these measures with respect to the therapeutic area indication. So very clearly, identify whether there are any approved endpoints that are available for a particular disease condition, therapeutic area or indication. And lastly, do evaluate the feasibility of what is the optimum quantity, what is the duration as well as frequency of data collection that is required to achieve this output. We need to be very clear with this particular approach in order to back track what will be the most suitable device for us in a particular trial. Having said that, study objectives is one part which is very important. We should not forget what our labeling needs are because determining the data analysis plan based on the labeling needs will have a direct impact on what data is being collected and which is the best suitable Digital Health Technology. So this is again where we need to derive the quantity as well as quality in order to know what is required from my labeling needs downstream as well. And last but not the least, after we do all this particular evaluation, we can hence derive which are the optimum digital biomarker as well as endpoint, which are based on the objectives as well as labeling requirements for my study. So how we found it easy from an application perspective is that we always try to work with the pharma company and the study team to define the objectives first, define the labeling requirements and then directly filtering out or deriving what would be the optimum digital biomarker and endpoint strategy. At this juncture, I would pause my discussion forum and hand it over to Tracy to take us through the operationalization of the entire aspect and share some highlights around what are the best practices when we operationalize this. Thank you.

Okay. So once sponsors have finalized their endpoint strategy, we move into the planning the operational aspects of the trial. So vendor selection, logistics, data collection, data management activities, all have key components that must be discussed, planned and executed appropriately in order for the trial to run smoothly and successfully. So looking at vendor selections, identifying best fit for purpose, right? We're considering 3 aspects here. Regulatory-cleared measures, so we want to ensure that the devices selected are approved for use in the country selected for that particular protocol or they can be imported under investigational use only, right? So intent-of-use needs to be very clear from the beginning as this is how we'll manage regulatory requirements during planning and shipments. So the second aspect here is biomarkers. When the biomarker's collected, should be mapped to the sponsor's clinical endpoint strategy across multiple geographies. This is very important, obviously, because we're working a lot of times on global studies. The third aspect here is most important, so easy-to-use technology. The preference here is for passive data collection. The less the subject has to interact with the device, the better the compliance and the quality of the data. Once we've determined what the device is, we're doing some risk-based evaluation with our vendors, right? So to qualify your vendors, we're looking at security, we're looking at privacy, scalability, right, for those Phase III trials as well as financials. So that you rest assured, right, that for a particular trial, the longevity is covered, and there's a little risk to a sponsor of bringing on a new company or device. So as we bring in new vendors, right, that our work does not stop there. We need to determine the best way to manage our vendors. We need to very clearly articulate in our MSA and our contracts what our liabilities are, who owns that intellectual property and who owns the data. So from a continuous improvement and new technology sharing perspective, we need to determine what the governance standards [indiscernible]. And once that's determined, we can -- we need to ensure that, that partnership with the vendor remains collaborative and transparent, and we can do this by monitoring those predetermined KPIs and metrics. So once then these vendor partnerships are established, we need to ensure the process for device provisioning, resupply, replacements to site [indiscernible] subjects are outlined. So global reach to subjects is going to be key here. So we should work with a provider that has a global reach and that understands required documents and licenses to support import and export of medical devices. So they should understand per country requirements, lead times required to import as these are going to vary by country and are dependent on device approval status. The key here is to really ensure that your study time lines are managed appropriately. So that study teams are very clear from a project planning perspective that startup times are realistic and that they allow for the required approvals to take place before shipments are to begin. If approvals were not in place before shipments are sent, this causes major delays in customs and delays FPI. The next, we need to understand and support disposal of biomedical waste at home. In sending devices and supplies, we need to clearly demarcate the biomedical waste and how it should be handled and then providing clear instructions to subjects, subjects, especially right, as they may -- you may or may not understand how to properly dispose of biomedical waste. So lastly, provisioning of locked down phones and laptops is important because we don't want our subjects misusing devices by exploring potentially dangerous websites. So in order to mitigate any risk to study data or material, we need to ensure that phones, laptops, tablets are locked down. They should only be set up to allow access to particular vendor portals, maybe depending upon your device, specific URLs or websites as needed to collect data or allow for access to device training, things like that. And of course, remote management of these provisioned supplementary devices, phone, laptops, tablets where it will be essential, right, because software over the course of the trial, especially a long one, will require updates and remote management of those devices allows for that. Okay. So once vendor selection and device planning has been established, we'll then start planning for data collection and generation. The key components to this are going to be to ensure proper use, training support and then along with project oversight of patient safety and remote medical monitoring. So first thing, device collection should be easy to use. We need to ensure there's easy calibration checks, device maintenance, explaining to the subject right device maintenance and any sterilization requirements. So the second thing there is easy-to-use device agnostic software. So eSource is preferred here so that subjects don't need to manage multiple logins. We also need to be able to manage infrastructure and connectivity of the devices at the subject's home for seamless and uninterrupted data collection and transfer. So we need to be sure to understand geographic locations and the infrastructure availability in these locations so that Internet connectivity does not impact data collection. If we can anticipate these risks globally, you can ensure that [ Wi-Fi ] devices can be provisioned in advance with the devices with provisioned phones, laptops, tablets, et cetera. And these can be sent to the subjects along with the devices themselves. So moving on to training support. So ensuring that the device is easy to use, does it mean that it's going to be used appropriately. So it's key that our [indiscernible] support are just in time and can be supported globally. Preference should be given to video-based training that's readily available for the subject to review at any time. As training too far in advance presents its own risks, we also need to ensure that the training provided is optimum. So as patients are less likely to read a long document on the use of the device and will likely rely on memory if the training was maybe provided on site at the first visit or potentially not at all. This is strictly a patient -- or device-to-patient trial. So the creation of subject kits and materials should be considered when shipping direct to patient. This ensures that they have easily understood materials and also some guidance for some home troubleshooting. Subjects will also need around-the-clock support. So -- and in their local languages, should they run into issues with devices, right? So it will happen that a device is faulty, needs to be replaced. So the key here for patient support is going to be that 24/7 how [indiscernible]. Patient safety and medical monitoring management direct to subject is going to be key considering global data privacy and other regulations. These things should be considered and discussed. So historically, right, the only mode of data transfer was subject to site and site to pharma or CRO. This now has a potential for bidirectional flow. Traditional one I just mentioned, but now where subject data will flow directly from device to the pharma company or CRO. There are instances where alerts and notifications are being shared directly with CROs where sites don't need to monitor 24/7. And then from a project and data oversight perspective, right? First thing to understand is that we will never gather 100% data collected or required to be collected. During the course of trials, we need to understand what is the optimum data required to support statistical analysis. So once this is defined, we can create those quality and quantity objectives and thresholds. Once these -- once those thresholds or objectives are defined, we can monitor the health of the trial with a set of standardized metrics. So with that metrics review, we can monitor things like compliance. Are they collecting the data as instructed or at specific visits? For some devices where data is stored and then uploaded at a specific time point, we can monitor whether subjects are uploading on time. And obviously, if they aren't, there is a potential to lose or overwrite data. So frequent review of those metrics and monitoring, right, allows us to course correct quickly. In doing that, we minimize the data loss as well as providing that guidance, that retraining to subjects and/or to sites, right? So we can also use this to identify trends over time, update training materials. Frequent data reviews in accordance with study objectives is key, right, so that we can identify issues in the data and mitigate any risk of unnecessary data being collected. So last piece of this is data management. Data cleaning, data aggregation and analysis are key components to data management activities. Ongoing data cleaning, so there's automatic edit checks at multiple levels were going to be important. So at the site, first entry by the site or by the subject, right, ensuring that no subject identifiers are shared. We can maybe only allow for certain number or alphanumeric subject ID or only allow them to enter in a [indiscernible] as opposed to a full data [indiscernible]. And then understanding and creating edit checks to the range of data collection that's required. Are we expecting 5-days data? Are we expecting 2 weeks, right? And the same point here demographic discrepancies after the data has come to us so we can monitor quality, right, and/or potentially device malfunction. Having said that, immediate subject feedback, like what we discussed during data collection is important even from a GM perspective because there will -- this is where it's going to be identified, is the data adequate and of good quality. And then as we move on to data aggregation, Tapan referenced this a couple of slides ago with regard to time synchronization. So this is going to be key, right, because we're working in multiple geographies, right, with multiple time zones. Really just within 1 country in dealing with daylight savings, time changes globally. This needs to be discussed and confirmed during the planning stages, right, because defaulting on this piece is going to result in data that is de minimis. Conversion of data into standardized formats across devices is going to be important in order for downstream activities and analyses to align across all other activities in the protocol. So the last piece of this, and the most important is related to data analysis. So Tapan referenced earlier that statistical analysis [indiscernible] needs to be clear. Interim evaluations need to be completed regarding relevance of that developed algorithm, which are used -- reused or new ones developed for accuracy. So this is where interim evaluations are done to ensure we're on track, right? And lastly, we need to apply validated algorithms to derive those digital endpoints as per protocol strategy. All right. So in conclusion, we can identify 4 major areas that are key considerations for moving clinical trials from site to patients. The primary goal here is to ease the burden on the patient while collecting as much data as possible. So our first one, Digital Health Technology selection, considering ease of use, considering age, population and disease state and planning for the required infrastructure to support data collection and transmission. Training, preference here again, video-based trainings that are easily accessible by the subject. Simple reference documents, as I talked about, direct-to-patient kits, right? So there's simple refence and troubleshooting documents, so they're easily understood. And then again, local language around-the-clock support for those subjects should they need further support. Subject -- I'm sorry, safety and compliance of data, so that real-time data reviewed through remote monitoring. And lastly, device management. Consideration planning for direct patient shipments and/or a hybrid trial, disposal of biomedical waste and the remote management of supplemental devices and supplies.

Excellent. Thank you so much for the informative presentation. Audience, before we get started on the question-and-answer session, I'd like to welcome IQVIA Senior Director of Connected Devices and Strategic Solutions, Sara Pawley. Sara has joined us today to answer any of your questions that you would have had for Tracy. So welcome, Sara.

[Operator Instructions] And let's just get started with our first question, what is a significant lesson from Digital Health Technologies from your point of view? Tapan, did you want to take that one?

Sure. Thank you. I think I would kind of respond to that one in the sense that what we covered through the 3 main aspects that we need to keep in mind when we are talking about a digital endpoint strategy, that is one particular aspect that we have to very clearly keep in mind. And the second thing that we should never forget from our perspective is compassion. What we are looking at is we should keep ourselves in the foot of the end user. So if the end user here is, say, a subject who suffers from cancer or a pediatric person or a geriatric population. So considering all of these particular aspects what is the feasibility of using the digital endpoint at the end stage is very important. And it is equally important, I would say, to the fact that how beneficial is the data collection to the sponsor or to my study. I would definitely lay most emphasis again, on compassion and deriving decisions on the basis of what the end user will -- who the end user will be and how good or bad or how feasible or difficult it will be for them to utilize this particular technology. So that's how I would probably look at it. That's the most significant and important aspect from my perspective.

Excellent. Thank you, Tapan. Our next question, what is an example of passive data collection?

I can take that one. This is Sara. So when we talk about passive data collection, we're typically talking about a device that the patient will wear where the data is just collected automatically. So if you think about something like a continuous glucose monitor, where the patient wears the patch and the data is collected. When you think about actigraphy, sensors and vitals -- continuous vitals collection, they are all examples of what we refer to as passive data collection, where again, the patient will wear a device, typically a patch or a watch, and then the data will be collected continuously without them having to actually take any actions. Thank you.

Thanks, Sara. The next question, how to ensure that patients are well -- how do you ensure that patients are well trained and confident? And are they supported by any HCPs or nurses at home?

Maybe I can take that one. Thank you. So from our perspective, what is crucial is that, yes, we need to -- there are times where in executing these particular trials, we try to create training material, which is such a huge document that eventually, we want to give them or train them on actual complete use and have everything covered and all of that. So while that is very important, we need to clearly keep in mind that just-in-time training, which is going to be very appropriate at the time. I mean, something which is too early or too late will not help. So just-in-time training, videos or simulation-based trainings and materials which can be hand-carried like a single-page document or a couple-of-pages document are the most important aspects that we have learned are really helpful when it comes to training the end users because we need to know -- keep this in mind that they are not the ones who are using these devices on a day-to-day basis. So they are pretty naive when it comes to use of these devices. So just-in-time and video-based, simulation-based training and very small leaflets or smaller documents are most effective in this particular scenario. Thank you.

Excellent. Thank you, Tapan. And we have a related question. What are best practices regarding correct usage by patients and monitoring? So for example, the attendee says, we had a patient who kept the plastic cover on a digital blood sugar monitor that had a number on it, and she kept reporting the same number every day until someone checked into it and found out that she was not using it correctly. So how do you -- what can you do in those situations?

Do you want to take that? Okay.

Go ahead, go ahead. No problem.

No, I was going to say, I mean, that again, certainly at the start of the trial, one of the aspects to clearly discuss is what the edit checks that are considered important. So obviously, if we're looking at data coming in and we continuously have to review the data then making sure things that just would not make sense are flagged and caught automatically so that we can see something like that. I mean, clearly, in that example, you would not expect the same value every single day. So starting to be able to program those edit checks into the data ingest mechanism allow us to flag those and identify those early on and then contact the site and then thus the patient to start to course correct. So again, having those discussions early on to make sure we set the parameters and set up the rules to automate that inspection of the data is key at the beginning of the study.

Tapan, do you want...

I would just like to add -- yes, I think I would just like to add one point. I think like -- emphasize one point, which is that eSource platform, getting the data directly from the device without manual intervention is something that will also be really very important for us because these plastic covers, I think we have seen that there are many such instances, not just blood glucose, but there are many others where it is important for us to avoid transcription errors or data entry errors that are happening at the site. So directly connecting the device and collecting the data from the device itself is also very important. Thank you.

Thanks, Tapan. Thanks, Sara, for that answer. Our next question is, which apps are better for muscle recovery and mental health and how do we approach these issues?

Tapan, I'll let you take that one.

Sure. I'll take. Yes, sure. Thank you. This is, again, an interesting one, and I think kind of connects back to the first point that I was referring to that whenever we are trying to select the best digital technology, we should not kind of go with the flow to select the best that is available and move ahead with that. Because the question or the challenge that has been posed here is a definite one. For people who have muscle-related issues and people who have mental health-related challenges, there are 2 main aspects that we need to look at. One, have devices which are most suitable from a site perspective or site orientation perspective. And there has been an increase of scenarios where caregiver related or caregiver managed devices or technology are also available, but that has to be moved with caution. Like I mentioned earlier, compassion is what we'll have to look at, whether the benefit of this particular output outweighs the frustration or irritation or problems that we may create for the end user or caregivers is what we need to definitely look at. There are very few technologies to be specific with respect to the muscle movement as well as muscle recovery or mental health, but there are devices that are available. However, the caution that I just mentioned about is something that, from my perspective, is really important for us to look at.

Excellent. Our next question is -- here we go. How do we ensure data privacy considerations while using these technologies?

Yes, data privacy. I'll take that one, Tapan. So data privacy is a pretty large discussion we will have with the study teams when we are doing our device selection. So for example, if we can recommend a device for the specific endpoints where we do not need to enter any personally identifiable information into any of the apps, and that would certainly be something that we would recommend so that the data actually on the devices can really anonymize with just using the subject ID at the time of data collection. However, that's not always possible. And so then if we are talking about using devices where, for example, they have to register in a cloud for a device manufacturer, for example, then certainly, we make sure that in the informed concern, the understanding of the consenting of the data is key when we do the data acquisition into the actual study database, we only pull through the minimal amount of data that is needed to support the actual study endpoints required so that we do not pull any superfluous data that might compromise PII into the study database. But all I have to say is it's absolutely a key point of discussion as we start to go through the digital endpoint strategy and then the actual device strategy to determine what are the risks and the benefits associated with any specific device.