Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 Concluding Remarks General Observations Wylie Burke, M.D., Ph.D. Moderator These comments are oriented around some general themes that seemed to emerge from the presentations and workshop discussion, Burke said. One theme is that innovation is important but that most innovations fail, perhaps because they are not good innovations. Both technological and organizational innovations are important drivers of cost and quality improvement. Therefore innovation should be encouraged, even recognizing that many innovations will fail. The environment in which an innovation occurs must be taken into account, as should the barriers that innovation confronts. Many innovations are incremental, but the willingness to pay a pre- mium price tends to occur only for disruptive innovations. Genomics may provide disruptive innovations, but it will also provide incremental ones. Disruptive business models are in play as well. The context for genomic innovation is a complex, fragmented health care system that incorporates a fair amount of uncertainty about the regula- tory environment. This uncertainty makes it difficult for innovators to plan. If the goal is efficient health care for the improvement of the health of the population, then incentives are not aligned well to encourage innovations that will achieve that goal. 81
82 DIFFUSION AND USE OF GENOMIC INNOVATIONS Not only is the U.S. health care system today fragmented and complex, but it is a system that delivers very poor value for the money as compared to health care systems in other developed countries. Innovation is a learning process. Most new technologies are relatively primitive when they get to market, and benefits and risks often are refined over time. Sometimes, unanticipated new uses of technology occur as Âpeople use the technology. There is a question about how the efficiency of that process can be increased since it frequently takes a long time to obtain a good understanding of the uses and value of a technology. One approach might be to think explicitly about capacity developmentâthat is, not just encouraging innovation but encouraging the capacity to incorporate new innovations into the health care system. The theme heard over and over in the workshop was evidence. Every- one agrees, although from a variety of perspectives, that evidence is needed to support innovation in health care. It was suggested that evidence is needed to show that an innovation will make a difference in outcomes that patients will notice. Evidence that shows an innovation will help clinicians do better at something they are already doing or do something helpful and beneficial that they previously were unable to do is also important. From a health technology assessment viewpoint, randomized clinical trials (RCTs) are a gold standard, but they are not the usual way in which technology is evaluated because there are not very many RCTs. Health technology assessment usually needs to use a combination of indirect evi- dence and causal inference, and that leads to a number of serious questions. How does one make optimal use of indirect evidence? When should RCTs be insisted upon? What is the role of the post-market evaluation process; that is, for that early diffusion process where there is learning about the technology? What is the role for evaluation at that stage, either RCTs or observational data? What kinds of infrastructures would promote the best use of post-market evaluation? Throughout the discussion of evidence there was a recurrent idea that further work on evidence standards is needed. One important point made is that test developers face different needs from regulators than they face from payers. There is a need to think about the hurdles for regulation and payment being the same. When is a leap of inference justified? What is the role of cost-effectiveness or other economic indicators in determining whether something comes to market? Is there any way to enable health technology assessment to occur earlier in the process, beginning even pre-market?
CONCLUDING REMARKS 83 Another theme in the discussion of evidence was that methods for evaluation should be transparent. Some models were presented for how one might go about this kind of evaluation. Discussion highlighted a need to ensure that clinicians and health care systems have authoritative sources of information, that evidence evaluation occurs in a transparent way using established standards and methodologies, and that evaluations for clinicians and the health care system are completed by a disinterested source. Ultimately, it is clear that providers and patients want data on out- comes. Even if an innovation enters clinical use without outcome data, there is a need to acquire those data. There are tensions in all of these issues, and those tensions need to be acknowledged. The level of regulation that is appropriate for genomic inno- vation is an area of tension. The level of evidence needed before proceeding to market is a tension, as is the role of conditional coverage as a mechanism to ensure that better data are obtained over time and after something comes to market. Then there is the question of what happens if the data say the innovation does not work. Finally, there was a fair amount of discussion about encouraging inno- vation, but with the caveat that innovation should not cause harm. What harm is and how it can be avoided are issues requiring discussion. Those were some of the big themes in todayâs workshop, Burke concluded.