Saturday, April 4, 2009

Comparative Effectiveness Research - The Emerging Opportunity

Industry's initial response to the health care system reform plans being talked about (a lot in public) and worked on (largely out of the public eye) in Washington has been, on the whole, cautious. No one really wants to defend the status quo, but there is widespread fear that whatever changes are made will have a negative effect on the commercial viability of life science companies. This should not be surprising - change is always threatening, and the threat seems greatest when, as is the case for health reform right now, the nature and extent of the change is uncertain.

Much of industry's current concern is focused, as I discussed in my last post, on the potential threats associated with increased funding for and utilization of Comparative Effectiveness Research (CER). It is easy to image CER being used to prevent market entry for some new technologies, or to stop payment or require market withdrawal for established products when new and better alternatives emerge. At the same time, the fundamental conceptual virtue of CER is so clear that it would be silly to flatly oppose conducting the research. Industry is reduced, because it is so focused on the potential threat of CER, to supporting the principle while opposing the use of CER findings by regulators and/or payers - i.e. trying to stuff the genie back into the bottle.

This is a losing strategy; CER studies are going to be conducted in increasing numbers, funded by private payers as well as government agencies. This would have been true independent of the Obama initiatives, but will certainly be accelerated by current reform efforts. The information CER studies generate is going to inform emerging clinical practice patterns and, ultimately, public and private insurer coverage policies. And there are substantial commercial opportunities to be exploited by companies that understand, embrace, and plan for the coming changes.

The most important thing to understand about CER is that it is a critical element in a larger ongoing systemic change - the move toward what is commonly referred to as Personalized Medicine (the right treatment for the right patient at the right time), and what Clayton Christensen has conceptualized as the evolution from "intuitive medicine" toward "precision medicine"(see The Innovator's Prescription: A Disruptive Solution for Health Care). Christensen's formulation is instructive because he provides a framework for identifying the opportunities that will be generated - indeed, required - by the health care system that will emerge, whether gradually on account of natural change processes or rapidly because of government-driven reforms, in the coming years.

I won't try to reconstruct Christensen's entire analytic schema; his book is too good, and his argument too subtle, for effective summarizing in this space. But I can point out a short list of the changes he sees as inevitable and the opportunities implied by those changes:
  • The move toward precision medicine will be driven by the development of increasingly accurate and refined diagnostic technologies - diagnostics that allow assignment of a particular patient to a therapy that is virtually certain to be successful; these diagnostics will include genetic and molecular clinical laboratory tests, but also increasingly refined imaging technologies and other diagnostic tools. Overall, opportunities in the diagnostic sector will be greatly enhanced by the system changes that are coming.
  • Realization of the promise of new and more discriminating diagnostic tools requires better and more comprehensive information exchange between and among institutions and practitioners; CER findings need to be incorporated into clinical decision support tools that can interface with individual electronic health records and that are accessible to practitioners but secure within the framework of loose organizations that Christensen refers to as "value networks". There is, therefore, an enormous emerging opportunity in health information technology (HIT).
  • Implicit in the growing commitment to CER is an opportunity in clinical research management and evaluation. Recent years have seen the emergence of innovative clinical research data management and evaluation technologies or systems, typically as the proprietary solutions offered by contract research organizations. FDA's (tentative?) embrace of adaptive clinical trial designs is the leading edge of a broader class of innovative tools or extracting meaningful information from clinical experience. As examples, look at the value propositions offered by PharmaPros, Cytel, and others.
  • For therapeutic technologies, the emerging opportunity is inextricably tied to embracing the CER paradigm - proactively identifying the patient subsets and/or circumstances for which the technology is most likely to be effective, accepting a smaller target market in order to achieve a higher efficacy rate. This dynamic is well understood by biopharmaceutical companies adapting to the business model implications of genetic "companion diagnostics"; broader CER initiatives will extend the dynamic more explicitly to device technologies. As a practical matter, this will frequently mean designing research studies that allow adequately powered comparison of outcomes across different clearly defined diagnostic subcategories. At the end of the day, companies that get out in front of the CER wave will prosper, and those that lag behind will not.


  1. To draw an analogy before commenting, a good baseball statistician can tell you the odds of a batter popping a low pitch when presented a curveball in a specific area coming from a left-handed pitcher.

    For precision medicine to work, we would need to have similarly well-differentiated data on hand so practitioners could make decisions on a particular situation. Has anyone proposed a comprehensive schema for such a knowledge base? Does anyone have the data to fill such a schema?

    To a casual observer such as myself, it feels like CER needs to be based on the fruits of a multi-decade data collection project but I would be interested in what the expert opinion in the field says.

    How long has baseball data taken to become an effective tool? Data for personalized medicine, being a bit more complex, could take far longer.

    In other words, could we actually expect CER to deliver results any time soon? I agree it is the holy grail but its capacity to pick winners and losers in the next decade feels distant.

  2. Easystageit is right to urge caution - but the baseball analogy is not particularly apt. Think rather about automobile repair. It used to be that you'd bring your car into the shop and tell the mechanic about troubling symptoms ("There's a rattle somewhere in the left front, the steering wheel vibrates when I go over 55 mph on a hot day..."). He'd then use his knowledge and experience - perhaps amplified by a test drive - to diagnose the problem, and would then get to work under the hood. The best mechanics were often right, but not always; when the first diagnosis turned out to be wrong, attention would be turned to the next likely answer. Today, this process of diagnostic evaluation, trial and error has largely been replaced by a computer that is far more accurate and which requires far less experience and diagnostic capability on the mechanic's part. What Christensen calls "intuitive medicine" is analogous to old-time auto repair; "precision medicine" is founded on diagnostics sufficiently refined as to precisely identify the underlying medical problem (the disease, not the symptom)and point clearly to the required treatment. Yes, it will be a long time before most medicine achieves that level of precision, and there will always be rare or complex cases that will require the intuition of skilled and experienced physician diagnosticians. But piece by piece, segment by segment, medicine is already moving into the realm of precision, driven by advances in diagnostic science.