The Newspeak of comparative effectiveness research (CER)

Few topics lately have suffered more obfuscation in the use of words than CER. Concerning the idea that in today's health care debate there exist persons who oppose the pure notion of CER I have, as readers know, screamed straw man. Though the claim of such opposition is popular no one has yet to provide me with an example. Today I ran across a post at Drug Wonks which, according to some, might qualify. And while it contains some very strong warnings about the unintended consequences of the government's financially conflicted agenda and its manipulation of the use of words concerning CER (many Orwellian references therein) it comes short, way short, of opposition to the pure notion of conducting research to compare one treatment with another.


With that out of the way let's look at what the post says (and there's a lot to parse here). First, after a little background on the Orwellian notion of “Newspeak,” the blog author Peter Pitts seems to imply that a proper rendering of the language of the CER provision of APRA (the stimulus bill) calls for a focus on individualized medicine rather than population based research:

The Recovery Act of 2010 (aka – “the stimulus package”) gave the Agency for Healthcare Research and Quality (AHRQ) $1.1 billion to conduct (according to the HHS press release) “comparative effectiveness research” into various “healthcare interventions.”

Except that’s not what Congress funded. Per the Recovery Act, that $1.1 billion was earmarked for clinical comparative effectiveness not comparative effectiveness research. And this is not splitting hairs. Enter cost-think.

Those in favor of comparative effectiveness research favor large scale trials to "compare" drugs and other healthcare “technologies, striving to show which medicines are most effective for any given disease state. Is there a “more effective” statin? A “more effective” treatment for depression?

But how do you compare two molecules (or three or more) that have different mechanisms of action for patients that respond differently to different medicine based on their personal genetic make-up?

That's a bit of a stretch for me. Was individualized medicine what the politicians originally had in mind? Who knows what the thought process actually was. According to a post over at Science Based Medicine by Val Jones the original bill didn't really define what the research was supposed to be, leaving it instead to AHRQ and the Institute of Medicine (little comfort that is!) to define the agenda.


Pitts goes on to elaborate on the difference between the population based one-size-fits-all approach and individualized medicine:

..even well-funded comparative effectiveness trials are swiftly superseded by trial designs based on better mechanistic understanding of disease pathways and pharmacogenomics. And, since most comparative effectiveness studies are underpowered, they don’t capture the genetic variations that explain differences in response to medicines by different patients. Comparative effectiveness in its current form leads to a “one-size-fits-all” approach to healthcare, which means that it doesn’t fit anyone all that well.

Clinical effectiveness, on the other hand, measures outcomes on an individual patient level. Clinical effectiveness studies help us to understand how to design treatments based on patient variation rather than cost. The very definition of personalized medicine.

Here Pitts seems to be speaking more to the appropriate design for comparative effectiveness rather than the value of CER itself. As knowledge of genomics and disease mechanisms grows so will the tension between population based and individualized medicine. Ultimately, in some areas, individualized medicine will win out. Population based CER simply (and sometimes simplistically) asks how treatment A compares with treatment B for condition X. The potential weakness of population based CER could be illustrated by an absurd example. Imagine a comparative trial that asks the question: How does vancomycin compare with ceftazidime in the treatment of infection? Here's one that illustrates the point equally well but is not so absurd because it's a clinical question often asked today: What's more heart-healthy, low fat or low carb? When the diet question is asked in so simplistic a manner what one gets is liable to resemble a series of scientific flip-flops. As it turns out it depends greatly on your genetic makeup. (Even this is at the risk of oversimplification, but if you happen to have a certain APO E genotype you're probably better off with low fat. If you're genetically predisposed to small dense LDL particles you might do better with low carb).


Some of the tension between individual and population based medicine in comparative effectiveness may be artificial. Just about everyone would agree that CER must of necessity follow the principles of evidence based medicine (EBM). But population based research is only part of EBM. In its classic definition EBM has two other components, starting with the individual patient and incorporating clinical expertise. If CER really is part of EBM, then, it should never lend itself to a series of one-size-fits-all mandates.


Nevertheless that may be just what some policy makers have in mind. One of the mandates of APRA is that comparative effectiveness be disseminated. There have been varying opinions about what that means. Bob Wachter, for example, once said comparative effectiveness research will form the basis for a revolution in health care that will “drag the self-interested laggards along, kicking and screaming if need be...” into, presumably, some type of conformity. You can form your own conclusions about what dissemination might be. Pitts says in his post, as has been noted before, that the AHRQ has funding under APRA to become the promotional vehicle for government health care recommendations. That raises questions about how such promotions might be leveraged:

Will physicians be “academically detailed?” And if so, will they be required to be detailed? Will physicians be given incentives to spend time with AHRQ’s comparative effectiveness angels (i.e., CME credits) and punished if they do not (via Medicare and Medicaid restrictions)?

And how scientifically rigorous will such “academic detailing” be?

Asked another way – how can an “academic detailing” program funded by our nation’s largest payer (Uncle Sam) be considered neutral? Just like detailing programs run by pharmaceutical companies, there is an inherent “interest.” And that’s okay – as long as that “interest” is transparent. Who will be the arbiters of transparency?

Who will decide what these detailers can say or not say? Will these government “reps” have to play by the same rules as their pharmaceutical counterparts? And, importantly, what is the oversight mechanism? If academic detailers stray into off-label conversations, to whom does DDMAC send a letter? Whom does the Department of Justice investigate? Who pays the fine?

Interesting questions, those.