 Okay, so Todd was going to come in and give you his impression as a geneticist about economics and how it impacts the Ignite program potentially. I'm taking his place, so I'm going to pretend to be a geneticist pretending to be an economist. So I feel a little bit like Julie Andrews and Victor Victoria knowing that I don't have a tenth of her talent. That said, these slides are the result of a collaboration of at least four different groups. We had people from Maryland, Vanderbilt, Florida, and NIH contributing to them. So if you don't like them, there's lots of directions you can send the blame. As an overview, we're going to talk a little bit about what economics could do for Ignite, some of the challenges we've encountered, and then briefly summarize what the current projects are doing and what the future potential for projects could incorporate some economic evaluation in the future. So obviously, I think the two previous speakers have made a very good case for why economics is important. I think the sort of wet lab scientists tend to focus a lot on clinical validity, but if this is going to be successfully implemented and sustained in practice, we also have to show social value. And we need to figure out how to incentivize the entire process care chain so that they all buy into doing genetic testing. I think one of the issues that we've come up with is if economics is so important, then why aren't we seeing more of it done? One of the problems we've run into is I did a quick PubMed search, looked up all the papers that had subject heading for pharmacogenetics, there were about somewhere over 20,000, looked up all the papers that had a subject heading of health economics and outcomes research, and there were like 35,000. And then the intersection of those two, and there were, it was a double digit number. It was a relatively small number of papers that have looked at both subjects together. And so Todd and I have talked a lot about why this might be the case. I think part of it is a language problem. So I've spent the last two years at IU trying to figure out what my genetics friends are talking about. They throw around these three-letter terms. I don't know what they mean. Sometimes they throw around three-letter and two-number terms. Those are worse. I didn't realize until a couple of months ago that Todd had exactly the same problem with my three-letter terms. And so there is a communication gap there. Some day I'm hoping that I'll figure out what some of those things mean, but it's going to take some time. I was not aware Todd had animated this. Todd put together a slide to show that some of the communication problems he's had with me, one of which an endless source of frustration is when I say cost-effectiveness, I mean something very specific. When he says cost-effectiveness, he's using it in lieu of the term economic evaluation. And I'd hold this against him except for the following, I've been using pharmacogenetics synonymously for genomic medicine for the same period of time he's been making this mistake. So I think what this really points to is a need for interdisciplinary, rather than multi-disciplinary team science, and the need to develop teams that can work and understand each other better than they have up until now. So one of the first steps is to figure out what questions we should be answering. And I think one of the challenges we encounter, and I certainly discovered this working with the folks from Florida, is that when we start listing the questions that might have an economic dimension that we might want to answer, they far exceed what we have available in terms of our resources, in terms of data, in terms of manpower. And then you complicate that with the fact that this science is moving so quickly that if we don't concentrate our efforts, we're not going to get an answer until it's already redundant and out of date. And I think some of the previous speakers have already alluded to that. So one of the questions for me is what questions will we answer now? But just as importantly, what questions do we need to answer, say, in the next five years? How can we get ahead of that curve in some way? Up till now, my reading of the literature in economics of pharmacogenetics and sometimes even branching out into genomic medicine, now that I know those are different, is that we've been playing a lot of catch-up. We tend to do economic evaluations of companion diagnostics after they've occurred. And this is a model that has served economic evaluations of prescription drugs very well, but that's because prescription drugs take seven years to hit the market. These companion diagnostic techniques are going much faster, and so it's much more important to try and play an anticipatory game and to try and come up with models as were previously described and as Nita Limdi is working on that can try and come to conclusions with partial information in a timely fashion. So then that leads to the second problem is once you figure out what you want to answer, we're also limited in what we can answer. And these problems, again, come in sort of two batches. Economic evaluation in the United States in particular is heavily reliant on using administrative records, particularly for the claims or cost information. And we're becoming more and more aware of the shortcomings. I think with the Florida group, we realized, oh, we've got really great information here on what it probably costs in the inpatient setting for clupidural and its effects. Gee, what happens when the patient gets discharged? What happens when the patient ends up in a long-term care facility because of a bleed? Those pieces of information we don't really have or we have, but they're very hard to get, and that slows us down, and we can't get the answer before you invent some new diagnostic companion thingy that makes everything redundant again. We also have provider training. I think that's earlier this morning. We were talking about how important a component that is. We don't really have the full costs. And we also know that the IT information system is going to be critical, and we don't really have a good handle on how much it costs incrementally to add genomic information to an existing IT system. At IU, we may be getting an insight into this because one of our sites is shifting from a custom-made IT system to EPIC, and so we're learning what it costs to change from one to the other. The second set of problems has to do with modeling. Most of the models that have been developed in economics, like the decision model seen earlier, are incremental, and they're designed and well-suited for looking at single companion diagnostics. What we're discovering at IU is the importance of things like multimorbidity, polypharmacy, and the socioeconomic factors that probably act as moderating effects that were also discussed this morning. And so one of the ironies I think is that when we do economic evaluations of stratified medicine, which is sometimes what we call it in economics for some reason, we don't actually stratify the economic evaluation itself, and that's something that probably needs to be done. So looking at Ignite and what we've done, we have three sites, again, that have incorporated economic evaluation originally in their proposals. IU uses economic costs as a primary outcome. Maryland and Mount Sinai both use costs as a sort of a juvenile outcome. That's a good start. What I think is more promising has been something that we only got because we're part of a network, and that's been in the work groups and the discussion of the role economic evaluation could play and how that has infiltrated into the other sites in addition to those that may have included economic evaluation at the beginning. So this is something that we wouldn't have achieved if we had funded those six sites as independent R01s. This is something that's only come about because of the network aspect. And I'll point out two in particular, the PGX work group, Laurie Cavallari and Nita Limdi are playing big roles there in incorporating economic evaluation in the second presentation, the one that immediately preceded mine. But as well, Tony Polin and Vicki Pratt have been working just as hard on the other side of the economic equation, which is how do you set up the reimbursements and incentivize the providers throughout the care process change to implement and adopt these different programs that we may find to be worthwhile. For future potential in Ignite, I think there are, again, the potential here that's unique is leveraging the network aspects and the cross-pollination across the groups, and it goes beyond just pooling data. I think there are now, thanks in part to Dan Mullins, who's pulling in more and more people from the corners of the country that I didn't fully realize we're operating. We're in a unique position to sort of establish methodological standards in economic evaluation. Points that were raised earlier, perspective, analytic horizon, how you handle uncertainty, and these are all needed to increase the reliability of the economic evaluations that are done, to improve the comparability of studies that are done at different sites, and probably most importantly, to ensure the transferability of results to other sites. So if we're talking about sending genomic medicine into the practice field, not just in the Academic Medical Center, we have to get a handle on this transferability issue. The other thing that economists can do here is we talk more and more about, gee, what are some of these moderating socioeconomic factors that might come into play? That's something that's very hard to do in an experimental setting. You cannot randomly assign somebody to race. I'm sure it's been considered, but it's never been done. Economists are really good at doing observational analysis methods, and so they can bring that expertise to the table as well. A second issue we could talk about is building the evidence base. I think a classic example of this is the effort down in Florida to pool data across different sites, in that case with respect to one drug gene pair. It's important to do that, not just to pick up very rare events or very small treatment effects, but also if you want to pick up heterogeneity of effect. And we've talked about that earlier today, where we think there may be some socioeconomic factors that come into play, adherence might come into play. And you can't do that just by increasing the sample size at a single location, because your population may just be too homogeneous in that one location. So the best way to do that is probably to pool data across multiple sites, as has been facilitated by the Ignite Network. And then the last thing is that the Ignite Network provides an incubation opportunity for cross-training some expertise. Again, I think it's going to be critical as we think about how you implement from the wet lab situation all the way down to the primary care office. We have to think outside of the Academic Medical Center. We have to think outside of the lab bench. We need to anticipate where the economic evaluation is going to be most useful. I just noticed the doctor at the end there is up to something. I don't know what. From this angle, it's kind of interesting. Yeah, I did not put the slides together. And so anticipating where economic evaluation will be most needed, that's something, again, where it will be very helpful to have people like have been gathered here today to tell us, you know, where are the most salient economic questions? Where are the biggest barriers? Where can the economists be the most helpful? Because there are, I get, economists study scarce resources, and in this field, they do seem to be a relatively scarce resource. Those people who can talk both sides of the science, the person who knows what an ROI is and what an ADR is, those are relatively easy, but I've figured those out. And there may be people who can do even the five letter and five number combinations as well. So that's my spiel for why we should continue with Ignite. Thank you, Anne. The discussion will be moderated by Howard McLeod from the Moffitt Cancer Center with the discussant Stan Roden of Vanderbilt University and Gail Jarvik at the University of Washington.