 I'm giving the perspective today from the Blue Cross Blue Shield Technology Evaluation Center. I am the Executive Director of Tech and Director of our AHRQ Evidence-Based Practice Center, and I am disclosing that I am a Salard Employee of Blue Cross and Blue Shield Association. I'll briefly describe our center, talk about how we evaluate evidence, and touch briefly on value affordability. In other words, who are we, what do we do, and what do we worry about? Value and affordability, but we'll only have a little time for that. So there are 39 Blue Cross and Blue Shield plans. They are independent plans making their own coverage decisions, but united by a series of principles and standards. There are 99 million beneficiaries, almost one in three in the U.S. carries a Blue Cross and Blue Shield card. We have been supportive of evidence as the basis for decision making, health plan decision making with the support of the Technology Evaluation Center since 1985. Principles of Tech are rigorous evaluation of clinical evidence. We use systematic review methodology, and the ultimate question always is, does this technology improve health? And so with diagnostics in general, not just genomics in particular, questions do revolve around clinical utility, because that is the crux of improvement in health. Our tech assessments can be accessed online at bcbsa.com slash tech. We consider that our contribution to public understanding of evidence and to leadership in evidence-based medicine. That work is done under the clinical and scientific authority of an independent medical advisory panel, primarily clinician and clinician researchers nationally known, including an appointing of the American Academy of Medical Genetics. We also, and this is not publicly available, but we provide for our plans who have to make day-by-day constant decisions around contract administration and benefits for what interventions are medically necessary, a medical policy reference manual. This is, again, supportive of our plans. We do not dictate coverage decisions. We don't make this publicly available because we don't cover anybody at the association. We don't make payment. We don't make reimbursement decisions. We don't make coverage decisions. We provide the clinical and scientific analysis to support these. But if you go to Plans' websites, you will see those policy manuals displayed online. And to a large part, they are resourced from our compendium. And we have been an evidence-based practice center since 1997, most recently designated as the Comprehensive Comparative Effectiveness Review Center in Cancer and Infectious Disease. I think it would be helpful to distinguish between these elements in the planned vocabulary. Medical policy is based on scientific evidence, on evidence of clinical effectiveness. It does not take into account costs and coverage. All of this analysis is conducted independent of those decisions. I will also add that, for the most part, blue plans are not for profit. Of the 99 million individuals who are beneficiary of blue plans, only 30 million of them are in for-profit plans. The rest are non-for-profit plans. What is covered is determined by the contract between typically the account, the employer, and the plan that circumscribes the benefits. And what is paid is a matter of the contract with the provider. So we're focused strictly on the clinical, not on the cost. We've been very active in genomics. This is what we have done in the decade before 2007. And in the two years after, you can see that the volume of genetic tests of interest and our devotion of resources to them has exponentially increased. We consider this a high priority for our work, for our systematic reviews, and are devoting about 25 percent to a third of our capacity in genomics, because we see this as an area which is calling for that kind of analysis. I remind you of the framework for evaluating evidence of the ACCE, where we look at analytic validity, clinical validity, clinical utility. But I also want to remind you that this is not unique. This is the model of the continuum for diagnostic efficacy put forth by Fryback and Thornberry in 1991. And it reminds us that it's not about technical efficacy, that is, pretty pictures are not enough in diagnostic imaging. It's not even about the accuracy of diagnosis, but it's about, can you ultimately improve health? And all of our questions are, can we improve outcomes? Quality of life, length of life, ability to function. This is the ideal, this is the model of the randomized controlled trial. Test versus no test, treat accordingly, measure the outcomes, compare. But I will caution now, and I'm going to remind you, very little of the literature and diagnostic technologies about randomized controlled trials. And this is not necessarily a lack or a gap, it is frequently not necessary, because one often has a reference standard, a model of disease, a model of what intervention into the disease means, that understanding performance is often sufficient to make inferences about outcomes. But not always, in particular, not when one is by the fact of a diagnostic technology, for example, defining a new disease state or looking at a new spectrum of disease. And I would suggest that this is probably occurring more frequently in genomics than it might in imaging, but I would not suggest that all genomic testing would require randomized controlled trials. So I want to give you some examples of assessments we have done on the continuum of types of evidence available to us. And I'll start with genetic testing for long QT syndrome, which illustrates a chain of indirect evidence going from what we know about performance in a clinical state where there actually is no true gold standard or reference standard, what the links in the evidence about what can be done in change of management in the inferences that outcome would be improved. In this setting, and of course this would refer to individuals who have a relative who either have a mutation or a no mutation status, a clinical diagnosis of disease, or individuals who, by clinical criteria, their status is suggestive but not clear. What is certainly ascertainable is that the genetic test ascertains more cases than clinical criteria. There is no gold standard. This is a lethal disease affecting a young population, highly unpredictable. There is the potential to change management for use of beta blockers, a relatively benign intervention considering the risks. And the conclusion was that use of the test, use of the management improves outcomes and has the potential to avert catastrophic consequences of left unassertained and untreated. So this is an example of inference through relatively, I'm not going to call it low grade evidence. But it's descriptive evidence and it's inferential through a train of what's known. I also point that we have new opportunities and predictive biomarkers for looking at evidence with the retrospective method, prospective method logic methodology proposed and used by Simon et al. Which allows us, it's not simple, but allows, for example, the use of archived tumor materials from existing randomized controlled trials to prospectively, rigorously pursue certain questions in predictive genomics issues, there are always issues, there are issues of stewardship of the archives, there are issues of missing data, but again, potentially alternative designs, alternative approaches. And then at the other end of the continuum, I'm just going to post their genotyping for warfarin dose, very complex set of intervening variables, trials or RCTs are in progress and a question of such complexity that we think it does call for RCTs. So again, a spectrum of evidence from a train of logic to RCT illustrating some of the points on that spectrum. Now I'm just going to touch, in the 27 seconds I have left, on value and affordability. To emphasize once again, clinical effectiveness is this cornerstone of plan, medical and coverage policy. I think there's a thought idea out there that plans are doing things based on cost, that they have some cost effectiveness criteria, but actually, contractually, plans really don't have that capability at all. They are mostly driven by the medical necessity contractual provision and it really makes virtually no provision for consideration of cost. In the last five or six years or so, there has been a provision basically put in there through court settlements that permits consideration of cost when two interventions have the same outcome and one cost more, then the one cost more isn't medically necessary. But real life is interventions rarely have the same outcome and what we're typically faced with is what's the increment of benefit versus the increment of cost. And we have no levers with that. That is a societal issue. Moreover, there is no gold standard for what is the value of a quality. Again, that is a societal issue, that is a political cultural issue. That really is not existing as a public kind of conversation in our society. But I think it is clear that even when you have interventions of high value, there may be limits to what is affordable. That is, Mercedes Benz on sale may be a good value, but it's certainly unaffordable to me. So we really are in a threshold, particularly as we look approaching any implementation of health reform, how are we going to get more individuals under that umbrella? And some of you may have read, I certainly did with interest the IOM report on the essential health benefit. It really costs for a very prudent approach where if interventions are added, then something needs to be subtracted, that you should take the total cost of the benefit to be stable. Will that be adapted? Will secretaries, civilians accept that? I would say, you know, that's speculative. I don't know. Probably not politically problematic, but I think it reminds us that we are dealing with an effort, and Netta Kalanj has mentioned this, that when cost goes high, access goes down, and how are we going to handle all of those things? We don't have it in our power. I believe that the moral compass and the leadership needs to come from the professions, from the clinical scientists, from the clinical researchers. Health plans are really not in a very good position to do anything about this, but people will suffer if nothing is done. And this is just a, this is from Peter Ortzog, Congressional Budget Office, showing that if health care spending continues on the course it has been, it has the potential for entirely consuming the gross domestic product, and obviously that's unsustainable, it's undesirable, so I simply remind you that, that as we talk of evidence, we really, and talk of clinical utility, we are hoping to put value into the system to improve care and to do the best we can with the resources we have. And then I'm happy for questions. Thank you. Any questions? Yeah. So, in terms of cost-effectiveness criteria, I guess you basically said you don't do it, but when it goes to extremes, you do consider it? No, no, I didn't say that. We don't have cost-effectiveness criteria. We have one provision in the medical necessity contract language that says if you have two things that have exactly the same outcome, you don't have to pay for the one that's more expensive. That is a rare event that you would have at occurrence. Just to rephrase then, if it's a million dollars per quality-adjusted life year, say you don't. No quality-adjusted life year. Same outcome. I'll give you the only example actually I can think of is virtual colonoscopy versus endoscopy. Virtual colonoscopy is actually quite a bit more expensive. And that is the one instance I can think of where plans have actually applied this provision in five years of experience. I think it's important to recognize, Ken, that again this gets at the perspective issue that qualities and that sort of thing are from a societal national perspective and that's not the way health plans adjudicate. So again, I think this is to reinforce what Naomi is saying. There's this misconception that plans are all about the cost, but the reality is is what they're really looking for are improvements in outcomes and that if you can demonstrate improvements in outcomes, they'll almost always be coverage. Now the coverage is variable in terms of its implementation with early adopters and late adopters, but the bottom line is that if you're improving outcomes, we almost never consider cost in any of the entities that cover health care in this country. And in fact, many of our biggest payers are specifically prohibited from considering cost. I see the French horn section is up again. This time I think I'm coming in on the right note. I wanted to ask you just a question. It's very common among geneticists to believe that what makes genetics different from other fields and what makes it different in the insurance arena is that clinical effectiveness applies to the patient in front of you, but in genetics it applies to the entire family. And I think there's always a concern that coverage plans feel that, you know, if someone, a brother or sister who's not covered by this particular plan, is covered by somebody else, sort of doesn't count and therefore we're not going to take family impact into consideration when considering clinical effectiveness. Could you comment on that? Yeah, that is an implementation issue that we recognize and really can't control because it's tied up in some of the contractual language. But our position is at Tech that the science says in many cases that it's clinically important to test the relative. This has been a challenge for plans. I think it is ultimately, and I think they have come out in various positions as a practical matter, but I think it is ultimately resolvable. I am reminded of HAIL testing for unrelated donors was a real sort of shock. How do you manage this? That's been worked out. I think this will be worked out. I actually feel that given that we are covering one in three of the U.S., we ought to be in a position to give some leadership and find a way to do it. But remember, we are operating under employer-driven contracts and there's not an appreciation of this. But we are promoting it. But in terms of the science, in terms of clinical effectiveness, do you take into account that testing the patient may not be of direct benefit to that patient but may be of tremendous benefit to the relative? I just told you we think it is very important to take a leadership role to understand that the clinical use of these tests may depend on testing an index case that will benefit a member, but et cetera. And we are trying to exert leadership there. That's the clinically right thing to do. And I think our medical staff understands this, but we are operating under contracts that are from another era and may not recognize that. And that's what they are circumscribed by, by what, for example, the employer may have purchased, et cetera, or what may be in the regulated insurance. So you are dealing with the overlay of contractual mechanisms that may not have caught up to current practice. That's why I used the example of HLA matching as something that was, as I said, a little bit of shock to how things were done but was eventually incorporated. Could it go faster? I wish it would. But you're dealing with a legacy system. And I'll just add to that that I think we have an opportunity here because this type of an approach fits within concepts, at least theoretical concepts of accountable care organizations and the fact that the healthcare systems have responsibility beyond just one patient by one patient. And particularly with the innovation center and the opportunity to perhaps explore this and maybe this is something that Jeff will mention in his talk as well not to put any pressure on you, Jeff. But I think this is an area where we should be very proactive in terms of looking for opportunities to really test this out and say this really fits under the concept of accountable care. Great. Thank you. Thanks.