As a statistician, have you ever wondered—why are statistical analyses for HTA dossiers differ from those in regulatory dossiers? Aren’t they both just benefit-risk assessments? In this podcast, we review some of the history and background of how HTA and regulatory decision making have common sources of information, but different utility functions for how they use that information. Why have we had a common European regulatory framework, but not a common HTA framework—until the near future. And what is the EU HTA Regulation? Why has it been established, what does it hope to accomplish…..and why does this matter to statisticians working in the pharmaceutical industry? This podcast is the first of a series of 4 exploring the role of statisticians and statistics in HTA analyses, and how the new European HTA Regulation that will first start being applied to medicinal products in 2025, will impact statisticians working with clinical (and non-clinical!) data to support and evaluate HTA decision making. Learn how you can get more involved in shaping the future of EU HTA as a statistician at the local and regional level.
References:
- D4.5 – EUnetHTA
- Antonia Morga, Nicholas R. Latimer, Martin Scott, Neil Hawkins, Michael Schlichting, Jixian Wang, Is Intention to Treat Still the Gold Standard or Should Health Technology Assessment Agencies Embrace a Broader Estimands Framework?: Insights and Perspectives From the National Institute for Health and Care Excellence and Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen on the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use E9 (R1) Addendum. Value in Health. S1098-3015(22)02148-9. doi: 10.1016/j.jval.2022.08.008.
Never miss an episode!
Join thousends of your peers and subscribe to get our latest updates by email!
Get the
Learn on demand
Click on the button to see our Teachble Inc. cources.
Featured courses
Click on the button to see our Teachble Inc. cources.
Anders Gorst-Rasmussen
He received his PhD in statistics from Aalborg University in 2011 and worked with pharmacoepidemiology and RWE in academia prior to joining the Development organization in Novo Nordisk in 2015. At Novo Nordisk, he has held roles as project statistical lead on various late-phase projects. For the last 2 years, has been working in the HTA area as functional lead for a small team of dedicated HTA statisticians. He is an active member of the PSI/EFSPI HTA Special Interest Group.
Lara J. Wolfson
She is Executive Director and Head, Health Technology Assessment (HTA) Statistics in the Biostatistics and Research Decision Sciences (BARDS) organization at MSD Switzerland. Lara holds a PhD in Statistics from Carnegie Mellon University in the US, and an undergraduate degree from Simon Fraser University in Canada. Lara has held academic appointments at University of Waterloo in Canada and Brigham Young University in the US, and has previously worked as a scientist covering health economics, epidemiology, and statistics at the World Health Organization, as well as in the pharmaceutical industry in outcomes research and Market Access at both the Janssen Pharmaceutical Companies of Johnson & Johnson in Belgium and the Merck Center for Observational Research and Real-World Evidence in the US. Lara has authored more than 70 peer-reviewed journal articles, and currently chairs the HTA European Special Interest Group of EFSPI/PSI (a professional organization for Statisticians in the Pharmaceutical Industry). Lara currently lives in the Greater Zurich Area with her husband, skiing-obsessed twin sons, and a cat named Hokey Pokey.
Transcript
Interview with Lara J. Wolfson and Anders Gorst-Rasmussen
[00:02:38] Alexander: Welcome to another episode of The Effective Statistician, and this is a really special one because it is the starting episode of the little series of a couple of episodes that I’m doing together with the HTA Special Interest Group, and I’m really happy to have Lara and Anders here today to speak about this.
[00:03:06] Let’s first introduce ourselves. Lara, do you wanna start?
[00:03:11] Lara: Thanks. So my name’s Lara Wolfson and I work for MSD and I’m based in Zurich, Switzerland. I lead the HTA Statistics Group at MSD, which is a group of about 80 dedicated statisticians who work with an equal number of programmers, specifically to reanalyze clinical data for health technology assessment submissions.
[00:03:32] I’ve been in this role for about four years. I have a PhD in Statistics from Carnegie Mellon [00:03:38] University in the US, but I’m actually Canadian and I’ve been working in the pharma industry for about 10 years now. Prior to that, I spent 10 years with the World Health Organization and also time in academic positions in the US and Canada.
[00:03:51] Alexander: Cool. Yeah. Very good. Eighty statisticians, dedicated statisticians. That is, by far, the biggest number that I’ve ever heard. And of course, MSD is not a small company, so that puts it a little bit into perspective, is it? There’s probably other companies that have less than 80 statisticians overall, but that’s quite a significant department.
[00:04:11] Okay. Anders, what about you?
[00:04:14] Anders: Yeah, my name is Rasmussen. I’m also part of the HTA SIG, and I’m working as an HTA statistician within Novo Nordisk. So I’m based in Copenhagen, in Denmark, where I’m part of, I would say also a much, much smaller group than Lara’s. I think we’re about five, five people now working within HTA Statistics.
[00:04:34] And it’s something that I think my group was established [00:04:37] two years ago. And to my background, I’ve been within pharma for seven years. And actually before I started working more dedicated within HTA, I was working as a regulatory statistician so I have five years of experience also working with that.
[00:04:51] And before that, I was actually working at somewhere completely different, which was more on in academia working within pharmacoepidemiology, so a bit of exposure also due to real world evidence. And I have a PhD from University here in Denmark.
[00:05:05] Alexander: Cool. Very good. Yeah, it’s actually quite helpful to have this mixed background in terms of both HTA and clinical and yeah, this episode will actually speak quite a lot about this, I guess.
[00:05:21] So let’s talk about European HTA, EU HTA, because there’s a couple of really significant changes going on and that started some time ago. And [00:05:35] yeah. Lara, do you wanna give a little bit of a high level intro in terms of what’s happening here at the moment?
[00:05:44] Lara: So I think here it’s helpful to take a step back and say what, what’s going on with HT, Health Technology Assessment itself?
[00:05:51] And I think one of the challenges we’re facing in our world is that there’s all these fantastic new medicines and vaccines that pharma companies are developing and bringing to bear. But health systems have challenges in saying, “How do we choose between them? How do we afford them?” And so they’ve started making this more and more formal as there’s more and more options on the table.
[00:06:13] And so regulatory approval, which is where we start in the pharma world, says, “Is this product safe and effective? Is it suitable for use in humans compared to, you know, there’s other options on there. Is it at least as safe and as effective as everything else out there?” In HTA, we’re asking a different question, but we’re trying to use the same data and we’re asking this question of
[00:06:33] “How good is it?” And that’s a different question. [00:06:36] And so we actually need to look at the data in a different way. And so more and more countries around the world, not just in Europe, have started saying, “Okay, how are we gonna interrogate the data to ask these questions?” And not just interrogate the clinical data, but you’ve got your regulatory or pivotal trial that you’ve used to get regulatory approval.
[00:06:57] What other sources of data do they need to bring into the picture? And so how do we decide if this is worth paying for, how much it’s worth paying for? And so you start to look at things like, are there particular subgroups or subpopulations who benefit more than others? How, how do you financially value some of the trade-offs between some of the side effects?
[00:07:16] So with Health Technology Assessment, we’re asking a different question than we ask in the regulatory context. Not just is it safe and effective, but how safe? How effective is it? How does it compare to the other alternatives we have? Can we quantify that? Can we add some economic dimensions so that we can determine how much we’re willing as a [00:07:36] society to pay for it?
[00:07:38] And are there particular subpopulations who benefit from a new treatment more than others or less than others? And so we start to segment it a lot more. And sometimes, as statisticians, we talk about we’re over interrogating it because we designed these studies to get medicines to patients as fast as possible.
[00:07:57] We think there’s a new treatment that could benefit people. So we design a regulatory trial to say, “What do we have to do to demonstrate that it’s safe and effective, get that regulatory approval and make it available to patients?” But then on the Health Technology Assessment side, we’re taking it apart in a different way.
[00:08:13] And maybe we’re bringing in alternative sources of data, real world evidence data from other trials where we might have to do indirect treatment comparison. But putting it all together to answer questions, and then this becomes very contextual. So in the European context, we have the European Medicines Agency, which does a pan-European regulatory approval.
[00:08:33] But most of the countries in Europe have developed their [00:08:36] own processes to say, “How are we gonna take that same data apart and evaluate whether or not this medicine should be reimbursed and at what level it should be reimbursed at?” And that has to be contextualized in terms of what they have approved before. And so, you’re looking at the standard of care and you’re saying, “How does this compare to the other treatments that are available?”
[00:08:58] And so, maybe there’s a treatment where there’s a guideline, for example, that says if you’re under 75 years of age, this is not your first route of treatment, but if you’re over 75, it is the first route of treatment. Well, now the data has to be analyzed that way because we have to look at how it performs in over 75 versus under 75, and this varies across Europe.
[00:09:20] It could be that in one country, that split is at 75 and in another country, that split is at 65. And so you have 27 member states in the European Union. Each of them has their own standards of care. Different medicines are reimbursed in different countries. And so they’ve all [00:09:36] evolved different rules and regulations. And this has been evolving quite intensely, I’d say for the last 20 years, and for a long time at the European level, they’ve been discussing, “Well, does this make complete sense? Do we want it this way? Is it fair that just because you live a hundred miles apart across, or a hundred kilometers apart across a border, that one person can get access to a medicine and another person can’t?”
[00:10:05] And so there’s a question of, “Could we make this simpler as well?” There’s also huge delays in timing of access. You’ll find that there’s some markets where as soon as there’s a regulatory approval, for example, in Germany, they have a process that says, as soon as that opinion comes out from the European Medicines Agency, they’re immediately going into discussions.
[00:10:25] That medicine is immediately available a year later. A price is set because of the HTA process. Other countries have a process where the timing is not as close to [00:10:36] regulatory approval and it can take a few years, and so there’s been lots of discussions. Can you harmonize the processes across Europe?
[00:10:44] Can you speed up access so that there’s equality of access in timing across Europe? And so there’s been these voluntary cooperations that people have tried to experiment with. And there’s been this series of different joint actions sponsored by the European Commission. We hear them talked about as EUnetHTA, but in December 2021, they actually passed a regulation that said, “Starting in 2025, we are gonna have a common EU HTA assessment, and this will form the basis of the country assessments that will follow.”
[00:11:21] There’s still the recognition that each country has its own standards of care, its own reimbursement pathways, different ability and willingness to pay in different countries. So there’s still gonna be a national HTA process, but at least starting from a common [00:11:37] foundation. And it’s interesting watching the evolution of this. And the amount of data reanalysis of the clinical data has changed dramatically in the last 20 years. There’s also a lot of economic modeling, and the economic modeling part of this will still sit at the country level. That won’t be part of the joint assessment, but they did pass this regulation in December 2021, and over the next two or three years, they’re gonna be shaping how this regulation takes shape.
[00:12:09] Exactly what analyses need to be done? How are they gonna do the joint clinical evaluations? Because starting in 2025, at first with novel cancer molecules and ATMP products, it’s going to start, it’ll take five years to fully take effect to cover all medicinal products and vaccines. But by 2030, there’s gonna be a joint clinical assessment that feeds into the subsequent national assessments.
[00:12:37] And it’s a huge undertaking, but lots of opportunities, particularly in the next 18 months to say, “What’s the right statistical way to do this?” Challenge because today, we don’t design the trials to meet HTA needs. And if you think about the pathways of clinical development, the trials that are already in progress that will start to read out in 2025,
[00:13:00] those trials have already been designed, and will already have first patients enrolled. But there’s gonna be a lot of questions that come up a little bit later in the game, which is that now that we know this process is gonna be taking place, should it, will it change the way we design some of those trials?
[00:13:17] So as statisticians, there’s all these different dimensions of how you wanna get involved, right? One is what’s the advice we can give about how they shape what should be in the joint clinical assessment? How are we gonna build capacity to do this type of assessment, not only within [00:13:36] pharmaceutical companies, but all the assessors who have to be involved?
[00:13:41] The assessors are gonna come from the national agencies and then work together at the EU level. I don’t think there’s enough of them. Anders, this is something we’ve talked a lot about this capacity issue and how do you develop it overall. I think it’s kind of an interesting challenge because there’s gonna be this huge emphasis on statistical analysis of clinical data re-analysis that the studies weren’t designed for, and do we actually have enough statisticians in Europe to do all of this?
[00:14:11] Alexander: Well, and in different companies. Yeah. We have just seen that there’s a huge discrepancy between the different companies and about other companies like Roche and Novartis and Pfizer, and they all have different organizations. Some have local people. Some have regionals, some have for global people.
[00:14:30] It’s very, very diverse. Some people say they don’t even report to the same organizations. Yeah. Some of them [00:14:36] report into R and Ds. Sometimes, they report into commercials, sometimes they report into affiliates. There’s all kinds of different things going on and that doesn’t make it easier.
[00:14:48] The number is one thing. The kind of complexity of the organizations themselves is another hurdle. So, and there’s in terms of capacities, that is one thing. And timing is yet another thing. So, as Lara just mentioned, timing will change as well. How will it change?
[00:15:06] Anders: So, one of the things that will happen here is that all of these weights laid out in the regulation is that we will process that pact basically to the EMA timelines.
[00:15:17] So basically, what will happen is that once you do your submission to EMA for a new medicinal product, you will also notify the Coordination Group of EU HTA and they will then set up this process of going out and asking the member countries basically via questionnaire. “Okay, what are the [00:15:35] instead? Is it interesting for you from the perspective of your healthcare system?”
[00:15:39] So, what kind of population would you be interested in that might vary between countries? Exactly what does the intervention look like? Are you adding this on top of some kind of background therapy that we should be aware of? What would be the relevant comparatives? Again, I think you’re also alluding too, Lara, that might again be different between different countries.
[00:15:58] And then finally, what kind of outcomes are you interested in? We might have countries that are more interested in heart clinical outcomes. You might have some that can make do, for example, with surrogate type outcomes. So all of this will basically be pulled together within this Coordination Group with the assessors and the co-assessors that will be appointed to oversee this.
[00:16:17] And then the member countries will send this information back and you basically, as a health technology developer, at some point, I think it’s around when we receive 120 days and from email, you’ll receive the scope with information about what is relevant to look at here. And of course, there’s a concern that has been discussed quite a lot [00:16:35] also in all of this.
[00:16:36] Will we see, I mean, we have 27 different member states. Will we see 27 different questions that we need to address as technology developers? EUnetHTA 21, they’re trying to reassure us that this is not how it will happen. But I think irrespective of how things are, we’ll probably get more than just one question and it’s then up to us to be able to predict that.
[00:16:57] You could also say well in advance because we will not have a lot of time. Because then we need to go into production mode and basically have the dossier ready for submission, collating all these different analyses, but based on the clinical trials. It can, maybe we need to do network meta-analysis. Maybe we need to do real world evidence registry studies, basically.
[00:17:17] And we need to have all that ready for submission no later than 45 days before CHMP opinion. If you do the math, you can see, they’re not counting clock stops. That’s about a nine-week duration that we have to prepare. It’s quite intense, I think, for technology in terms of putting all this together.
[00:17:35] But I think an important fact here that we also need to bear in mind is that I think it could also be quite the challenge for the people that will then subsequently do the assessment of the dossier that we’re providing just basically with breadths of evidence. We might include data from clinical trials.
[00:17:52] It’ll be, NMAs will be real world studies. There will also be a lot of pressure on the people that will actually do the assessment afterwards, because all of this needs to be done, I think it says 30 days post marketing authorization in Europe, where the Final Joint Clinical Assessment Report needs to be done.
[00:18:09] So there’s also a lot of pressure on the assessors and the co-assessors and the statisticians that will be sitting on the other end of the table actually looking through this and assessing the evidence that we’ve submitted as health technology developers. So I think this really calls for statisticians also.
[00:18:25] Not just a lot of statisticians, but I think also for statisticians also, those working in the regulatory space to do have maybe increased their understanding and also of the HTA mindset and the way [00:18:35] that you need to approach evidence when you think in terms of HTA, which I think is different as Lara also explained, than what you do from a regulatory perspective.
[00:18:43] Alexander: Yeah, and it will mean you need to work very, very closely together. So this is what I’ve seen across lots of different companies. We first take care of the submission, of the regulatory submission, and then you get access to the other data that will not work.
[00:19:02] Lara: So the dilemmas here, there’s a funny story.
[00:19:05] I sometimes, one of my first encounters with health technology assessment was when I worked at the World Health Organization. I worked primarily in the vaccines area and there were some questions about that. We worked a lot with GAVI, the Global Alliance for Vaccines and Immunizations. And GAVI was basically functioning a little bit as a health technology assessment agency because they were making decisions about how they were going to allocate billions of dollars? Which vaccines were they gonna give it to?
[00:19:31] Were they gonna give it to measles control? Were they gonna use it to support novel [00:19:35] vaccines for pneumococcal and rotavirus vaccines? And this is one of the big challenges of health technology assessment is that on the surface, it seems really simple. It’s, oh, we wanna fund the most cost effective intervention, but it turns out a lot of other values come into it.
[00:19:52] And so one of the funny things when I did this work with GAVI was that we started re-analyzing all kinds of data. It wasn’t so much focused on the clinical data because the effectiveness of these different vaccines was well understood. But the question then became linking it to how many patients
[00:20:08] were you targeting? And in this case, you weren’t targeting people who had a disease, right? Vaccines are preventative. So what was the potential number of kids that you could vaccinate? Well, that’s not so hard to get to the total patient population. It’s the number of kids that are being born each year.
[00:20:23] But then how likely were they to get the disease and potentially suffer severe consequences? And so you started mapping all of this out, and it turns out that you start factoring in different things. What’s the [00:20:35] morbidity? What’s the mortality outcomes? Which outcomes are most important?
[00:20:40] And then you start to bring in some things that aren’t obvious because it also has to do with, wait a minute, measles vaccines, they’ve been around for 40, 50 years, right? They’re not novels. They’re being produced. You don’t need to incentivize that. But at this time, it was 2004, 2005, rotavirus, pneumococcal vaccines were just coming into play, and there were other values that came into play, like wanting diversity of supply, wanting to encourage manufacturers.
[00:21:07] And so one of the things you saw was that the people who were making the decisions, they thought they knew what they wanted, what information they wanted. And so they gave us these table shells and said, do all of these analyses, fill these all out. And we’re like, and I started working on them and putting them together and I kept thinking, how on earth is anyone going to be able to make sense of this because of a thousand pages of tables.
[00:21:30] Like who can process that kind of information? And then you produce it all and you start to [00:21:35] systematically evaluate it. And it kept pointing to investing in measles control, and everyone kept resisting it because they’d already made up their mind. They actually really wanted to make sure that they were allocating money to rotavirus and pneumococcal vaccines because it turned out that one of the core underlying values was stimulating production of new vaccines.
[00:21:56] And in this case, this was a global context, not a country-specific context. They wanted to make sure they brought developing country manufacturers in and created incentives for them. But the analyses they asked for were never designed to answer the question. And so this takes it back to one of the critical things about statisticians really starting to understand the health technology assessment system you’re working in.
[00:22:19] So it seems weird that I went back to this story that doesn’t have to do with European HTA, but it’s really, to me, it was really meaningful because at the end of the day, it turned out they fundamentally had a utility function that they weren’t expressing. And so HTA, whether it’s gonna be European HTA or the [00:22:35] country specific HTA,
[00:22:37] a lot of it is about articulating or interrogating that utility function. What is it that you’re really trying to assess and how are you defining value? So in the regulatory context, we have a utility function that is implicitly well understood. In the HTA context, I think that utility function is gonna continue to evolve.
[00:22:59] And we’re gonna have to work with the national HTA agencies and with the European HTA agencies to understand the information that really helps them make their decision. But their challenge also is that they haven’t fully articulated a utility function, and so they go after everything. We’re likely to see a system that’s very similar to the AMNOG system in Germany
[00:23:24] where an AMNOG dossier can, people are still submitting them as PDF files and if you actually printed them, which I’m sure some people are doing, I worry about the environmental impact of that. [00:23:35]
[00:23:36] Alexander: You need to concentrate for that, yeah.
[00:23:38] Lara: Yeah, I have been involved with AMLOG dossiers that run to thousands of pages.
[00:23:43] And again, I struggle with how do you really make sense of this much information and what information is actually driving the final decisions? And over time, can we work with these systems to focus the analyses they need on what really impacts their decisions? Because I think right now, these HTA systems have a challenge that they know there’s all these different aspects they wanna go after so they just ask for everything.
[00:24:13] Alexander: Yup, they’re all subgroups versus all endpoints versus all time points versus all these methodologies by study and combined study and everything. Then you end up easily with, yeah, as you said, thousands and thousands of tables.
[00:24:29] Lara: And how cognitively, how do you process that? How do you say, how did [00:24:34] get to a decision?
[00:24:35] How am I sure that my decision is reproducible and traceable and people can understand it? And so on the one hand, as statisticians, we’re going to have to meet the needs of these systems. But then, there’s this other question which is, how can we help to make these systems better and more effective and more efficient and really understand what they’re trying to get at?
[00:24:58] Alexander: That’s a really good point. And the HTA SIG has really started to get involved there. Anders, do you wanna talk a little bit about what you have done already there and what’s currently going on there in terms of how the HTA SIG helps them?
[00:25:16] Anders: Yeah, so I just want, before we go on to that, I think I want to also add a comment to what Lara was speaking about, mainly this idea that we have this utility function lying around also out in the different member states, that they’re not always very explicit about.
[00:25:30] So basically, I think as you also started out, Lara, this is really about assessing [00:25:35] the value of new treatments. And I think in that respect it’s also a massive undertaking. What, what is going on with the EU HTA, right? Because the regulation is quite specific to the joint clinical assessment, so the dossier that will be producing the actual assessment that will come afterwards is not something that will in any way dictate how the value should be looked at, I mean, here in the individual member.
[00:25:57] That will be up to the member states themselves also because that would be very much dependent on how the health system is set up. So basically, what is lined up now is that we need to produce a dossier that needs to be an assessment that’s broad enough to fit into the value perception of all the different member states.
[00:26:13] And it should be fully objective. It should only be about, you could say, validity, and then put the value assessment out with the member states. And I think it’ll be interesting to see how this actually goes on in practice. I think it also very much easily can end up in a situation where it becomes quite a bit of information overload, right?
[00:26:30] Because there are just so many different perspectives on this, but I think this [00:26:34] aspect of trying to make a clean cut between value and validity, it’s really difficult. I think we also see it in terms of what we see some of the draft methodological guidelines that are coming out now. When are you actually with your recommendations about particular statistical analysis?
[00:26:51] When you’re making a value judgment, when is it just a case of validity? I think that’s also a place where statisticians have something to say. So, in terms of what’s going on in the HTA SIG, you can say what’s happening right now also in terms of the preparatory phase leading up to the implementation of the HTA regulations. So until 2025, there’s agreement of the tender agreement with the so-called it’s an involvement of the EUnetHTA that Lara was alluding to at the beginning. But we have this EUnetHTA 21 consortium that’s basically looking into doing even more preparations and making sure that we’re in good shape for actually being able to implement the regulation come 2025.
[00:27:31] And one of the things that they’ve been doing is that they’ve been grafting. It’s [00:27:34] a very extensive methodological guidelines for how the processes should look. What kind of, if you do, for example, a network meta-analysis, what are methods for doing that? What should assessors look for when looking through the dossier? If you do a registry study, what should they look for and so forth?
[00:27:52] So we’ve actually been quite busy over the year, over the last couple of months. I think the EUnetHTA 21 consortium has been even busier. But there’s been commenting going on. So public consultations basically where there’s been input also from industry organizations. So at PF, we’ve also commented within the HTA SIG.
[00:28:11] And apart from that, which you can say, it easily becomes a bit of a one-way communication. We’re also doing things like, for example, participating on your podcast, Alexander, but we’re also looking into various other venues of really sort of getting some engagement also within the statistician community and also making people that are not necessarily deeply involved in HTA statistics realize [00:28:34] that this is something that will also affect them, that this is really about being able to also to want to end up with actually having faster patient access, we really need to think about how do we integrate the HTA perspective also into our pivotal trials so that we’re not, when we get at around the same time as the email 120 day questions, we get the scope that we are not completely surprised by what we see, but it is actually to a certain extent, at least predictable.
[00:29:02] Alexander: Yeah, I noticed that for example, at the PSI conference in 2022, there was this lunch meeting where there were lots of different people sitting together and talking about this topic. I can absolutely encourage people, check out the HDA Special Interest Group homepage, check out if you’re a PSI member, there’s a newsletter that is coming out.
[00:29:25] There’s also for SPIs has very often things coming out about this. Watch out for that. If your company is not [00:29:33] yet involved there, make sure you have someone that is involved there. If you don’t know who is in your company as a statistician involved, reach out to them. That will be really, really important for if you work on phase two, phase three, absolutely important.
[00:29:52] Yeah, really critical. As Anders just mentioned, if you’re surprised by what you’ll get, then it’s too late. Getting a high level understanding of disease will be the critical point. Then, oh, here’s another H cut point like Lara mentioned, or here’s another kind of cool medication to take into account.
[00:30:10] Okay, is it, you can take it into account when you do this. But if you haven’t collected certain endpoints or you can’t cut your data because you haven’t collected these core variants, which I have seen as well, then you’re in trouble, then you are really in trouble. And then maybe millions or billions of dollars that you know your company has invested in [00:30:33] is at risk.
[00:30:35] So really an absolutely important point. And as I said, this is just the first of a series of episodes. With this one, and I’ve actually, we are not recording them in sequential order so I can already tell you, “Yes, those would be really good.” So really look out for them. Lara, Anders, any kind of final points you wanna make and the key takeaways for the listener?
[00:31:03] Lara: Following on what you said, this opportunity to get involved. So if you go to our SIG webpage, you’ll see that there’s the schedule of all these deliverables that the EUnetHTA 21 Consortium is producing. These are the first draft. They’re not just first drafts. These are the guidelines that they’re going to use for the assessments
[00:31:22] they do in 2023 and 2024, but they’re the foundation of what they call the implementing acts of the EU HTA regulation. And so they’ll be, they’ll be coming back next [00:31:34] year and in 2024 for finalization. And so it’s an opportunity to post the links to the deliverables as well as coordinating input across all these companies on the deliverables.
[00:31:46] One other angle that you’ll have available to you within your company is that most of the pharmaceutical companies are participating in EFPIA or the European Federation of Pharmaceutical Companies and Industry. And EFPIA is also providing input. We work closely with EFPIA. We provide input to them. We look at, are there statistical issues that we need to particularly emphasize in our feedback that we give to the consortium?
[00:32:10] But this process is gonna be going on for the next three years. It’s incredibly important to understand it, to shape it, to bring it back into your own company in terms of how you’re thinking about the design of your trials and looking at all these different dimensions. One thing we didn’t talk that much about but, quality of life measures.
[00:32:28] We collect them in our studies. We summarize them briefly in our regulatory dossiers. They play a [00:32:33] huge role in HTA assessments. They come in much more there. Now, there’s gonna be this European level scrutiny and one thing to also realize about EU HTA and one of the reasons you wanna get involved and think about it, even if you’re a regulatory statistician, is that the EU HTA assessments are most likely gonna be in the public domain.
[00:32:55] So a lot more of the data and the way that you’re gonna have to take it apart is going to become much more transparent over the next five to ten years. And so again, thinking about what does that mean for the design of your trials, what does it mean for your choices of methods and analysis? So one thing I also just wanna give a brief plug for is that on November 14th of 2022, we’re gonna have a webinar with a chance to have a panel discussion with some of the people who are the architects of the EU HTA system and talk about some of the statistical issues and ways for the statistical community to get more involved and to think about what are the skills that we’re gonna need to [00:33:34] be able to do this.
[00:33:35] How do we contribute? How do we upskill ourselves? How do we upskill within our companies? How do we contribute to the bank of knowledge among the assessors? And where are the methodological gaps? Because there are some. There are places where there’s no clean cut answer. There’s a lot. It’s a great opportunity.
[00:33:55] It’s gonna really bring some of the statistical issues to the forefront. In our conversation today, we focused a lot about the joint clinical assessment. There’s also gonna be joint scientific consultations where you can get input on the HTA aspects of the design. And it is incredible to see how much of this whole European HTA system is really founded in statistical analysis and statistical principles.
[00:34:19] Lara: Huge opportunity for statisticians to be involved in shaping the future of the healthcare system in Europe. And so don’t miss out. We’re really looking forward to getting more and more people involved. The more smart [00:34:32] people that think about this and contribute to this, the more successful the overall system will be. And I think we all want to see a successful European healthcare system.
[00:34:41] Alexander: Yeah, but actually as you point out the European part of it, as you mentioned earlier, lots of other companies in other countries are looking into this. As we have seen with the kind of NICE and how it’s had impacted, kind of globally how HTA has done, pretty sure that lots of other countries around the world and I think even including, we’ll have a look into this. Those that might sound, oh, that’s Europe. Don’t think that way.
[00:35:10] Lara: Completely agree.
[00:35:11] Alexander: Yeah. Anders, how about you? What are your kind of key takeaways for the listener?
[00:35:17] Anders: Yeah, I can only echo Lara’s call for involvement. I think there is also a bit of confusion also around what we can do. Is it now that the methodological guidelines are being developed, is that it? Once they’re done and finalized from a EUnetHTA 21 perspective, does that mean that [00:35:32] it’s set in stone? I think the answer to that question is probably not, because this is just the work of EUnetHTA 21.
[00:35:39] At some point, there will be implementing parts of this and there’s at least been discussion also that once the coordination group gets into to work and the coordination group of the EU HTA will comprise all of the 27 different member countries, not just the smaller hands-on group with, I should say also EUnetHTA 21, a very large representation of some of the German institutions.
[00:35:59] So G-BA and IQWiG, for example. But once the coordination group gets into work, I think things will also continue to evolve. We will have the implementation starting in 2025, but come 2028, there’ll also be a review of, okay, has it actually worked? Does it help out in the individual countries? So I think there’s, it’s not just, okay, it’s 2022 and then everything is done and over with, and then we just wait for the implementation.
[00:36:25] I think there’s actually also a longer window of opportunity where we can raise engagement and where statisticians can really raise their voice in terms [00:36:32] of making sure that we have a robust system, and importantly that we ensure faster access for patients to new treatments in Europe.
[00:36:41] Alexander: Yeah, completely agree. If you see, think about NICE or if you think about G-BA and IQWiG, they have continually evolved their guidelines. I don’t know, in Germany, I think we are now at Methods Paper number six, Versions Paper number six. And if you look into the NICE technical summary documents, they have grown and evolved.
[00:37:03] There’s lots of different versions of these. So, I don’t think that this will be done and finished when people see how that actually implements and how these dossiers actually see and look and feel and the volume of them, I’m pretty sure that will evolve. Because then people will see, “Oh, maybe we underestimated its complexity or we underestimated how much pressure that we’ll put into the system.
[00:37:32] And as Lara mentioned, maybe we underestimated how many statisticians we actually need for it. I think there will be a lot of learning both on the HTA, the EMA side and the company side. And I also think that other parties, patient advocacy groups will get more involved. Yeah, because they will potentially drive these kinds of utility functions that you mentioned by saying, “Oh, for us, this endpoint is much nearer.”
[00:38:01] Anders: And I think you’re touching on a very important point here because will it just be a massive PDF file with lots of super complex statistics, right? It’ll be super transparent also. That’s also something that’s built into the HTA regulation, and I think we really need to think hard about the statisticians also.
[00:38:16] How do we make this, it’s super complex statistics, it’s not accessible in any way for anyone without a PhD degree in statistics, is that being transparent or do you just collate that into 50,000 pages of analysis? I think there’s also an opportunity here to think a bit about how we actually succeed with [00:38:32] transparency?
[00:38:33] Because transparency is not necessarily just putting a lot of stuff out there. It’s also thinking very carefully about the presentation.
[00:38:39] Alexander: Yeah, completely agree.
[00:38:42] Lara: I worry. I think the transparency initiatives are great, but then I put myself in the position of what if I was a cancer patient? And all this, I’m diagnosed and I start Googling things and I, it takes me to this HTA assessment filled with hundreds of pages of PDF tables and what happens if I get confused about whether or not the treatment my physician is recommending is actually right for me? And so at some point, that’s also something we have to think about is how do we make this information digestible to different consumers?
[00:39:14] Alexander: I completely agree. That is so much of an important area and yeah, we need to learn not to just to speak to other experts, but to speak to treating physicians, to patients, to parents, or to the kids of the patients [00:39:31] because they might take care of their grandparents or their parents that are really old and have dementia.
[00:39:39] So all these kinds of things really play a role. Thanks so much. That was an outstanding discussion and it’s such an important point. So again, head over to the HTA Special Interest Group on the PSI homepage at psiweb.org and there you’ll find all the other things. And also show notes are in the block here on theeffectivestatistician.com.
[00:40:08] So check that out as well. There you will also see that EUnetHTA is actually spelled EU big NET small and HTA big again. So you say if you search for EUnetHTA, that’s how you search for it. Thanks so much, Anders, Lara. Have a great time and it was an honor to have you on the show.
[00:40:30] Lara: Thanks.
[00:40:31] Anders: Thank you for having us.
Join The Effective Statistician LinkedIn group
This group was set up to help each other to become more effective statisticians. We’ll run challenges in this group, e.g. around writing abstracts for conferences or other projects. I’ll also post into this group further content.
I want to help the community of statisticians, data scientists, programmers and other quantitative scientists to be more influential, innovative, and effective. I believe that as a community we can help our research, our regulatory and payer systems, and ultimately physicians and patients take better decisions based on better evidence.
I work to achieve a future in which everyone can access the right evidence in the right format at the right time to make sound decisions.
When my kids are sick, I want to have good evidence to discuss with the physician about the different therapy choices.
When my mother is sick, I want her to understand the evidence and being able to understand it.
When I get sick, I want to find evidence that I can trust and that helps me to have meaningful discussions with my healthcare professionals.
I want to live in a world, where the media reports correctly about medical evidence and in which society distinguishes between fake evidence and real evidence.
Let’s work together to achieve this.