Yesterday came the news of a precipitous drop in the breast cancer rate in the U.S. Todd Ackerman, one of the Houston Chronicle's fine medical journalists, reports:
The most significant decline ever recorded in U.S. breast cancer rates may have been the result of millions of older women stopping hormone replacement therapy, according to a new analysis.
Researchers at the University of Texas M.D. Anderson Cancer Center revealed Thursday that the rate dropped 7 percent in 2003 and suggested a striking reason: women's abandonment of menopause hormones after a large study was stopped the year before because the pills were increasing health risks.
To be frank, this makes me feel a bit like Marvin the Martian (.wav file), and it has everything to do with evidence-based medicine. EBM, as it is often called, is roughly defined as "the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patient." My own reaction, as well as a common reaction of laypersons when they hear this definition, was something along the lines of, 'What do you mean? What else have physicians been doing?'
I have personally been writing and thinking a great deal about EBM in the last month, and though this blog will always reflect my own voice, it also is not a chronicle of my own academic pursuits. I will not, therefore, go into too much detail about the controversy over EBM, but suffice it to say that while I think that awareness of some of the practical problems and pitfalls of EBM is merited, I also think such problems do not negate the conceptual importance of EBM, and that avoiding any commitment to engage the problems of EBM can result in dire consequences.
Jay Katz documents some of these consequences in the course of breast cancer treatment in the 20th century, but the saga of hormone replacement therapy ("HRT") is another sad piece of the puzzle. From my own research, I found that in 2000, physicians issued over 45 million prescriptions for the most common HRT, in the staunch belief that HRT would help prevent a range of diseases for postmenopausal women (including cancer and coronary disease). Such beliefs -- in terms of outcomes and efficacy -- had not been confirmed via rigorous scientific studies. When such a study was finally done by the Women's Health Initiative, the "unexpected" findings were that HRT actually seemed to increase the risk of certain kinds of invasive breast cancers. In fact, the study was halted on the recommendation of the data safety and monitoring board because it appeared to be doing more harm than good.
Thus, providers had issued prescriptions for HRT for millions of women in the belief that it might help reduce the risk of certain kinds of cancer, which belief had simply not been confirmed by the best kinds of evidence (which are not randomized controlled trials in every instance). And now we hear the news that the most plausible single factor explaining the highly statistically significant drop of 7% in the 2003 U.S. breast cancer rate is the fact that providers ceased writing the prescriptions and women stopped taking them.
Please do not misunderstand this post as an attack on physicians or providers. I both understand and agree that there is immense value to a physician's practical experience, training, and intuition. But I think that such sources of knowledge, unaccompanied by the best evidence (quantitative and qualitative) can be exceedingly dangerous when used as the sole criteria for assessment of an intervention. Often, there is simply no way to know if a given intervention will be therapeutic or not. It's not the uncertainty that is the problem, IMO; it's the willingness to plow ahead, to avoid incorporating that uncertainty into clinical practice, that concerns me.
Anyway, I obviously have a great deal more to say about this topic in general, and what I have said here about EBM is merely the tip of a very important iceberg. But I do think the news about HRT is an important theme in the overall discourse, and I wanted to mention it.
(You can find a short piece I wrote on EBM and the recent news involving stents here (.pdf), and you can find an article on another sad chapter in the breast cancer saga here, which I wrote before I had ever heard of EBM, but which fits in pretty obviously with some of the points I am concerned with here).
I can also recommend some excellent sources on ethics, policy, and EBM, so if anyone is interested please feel free to email me.
AMENDMENT: I have had some discussions with mentors, who have pointed out that it is much easier in hindsight to question the wisdom of prescribing HRT given what we now know, but that the observational evidence was quite consistent and was based on an entirely reasonable physiological model (women generally had much less problems with coronary disease when they were premenopausal).
This is obviously legitimate, and I did not mean for the tone of the post to be construed purely as a retrospective critique of the prescribers. Though I stated that "Please do not misunderstand this post as an attack on physicians or providers," I realize I was not as clear as I could have been in addressing this point.
I understand that many decisions that seem questionable in hindsight seemed quite reasonable at the time. This reflects the problems of induction (that sometimes regularities we have grown accustomed to linking causally are linked by entirely different causal relations or not at all), and it is obviously unfair to fault clinicians for those problems.
I suppose I am more concerned with the process by which these kinds of "inductive mistakes" seem to happen quite frequently, especially in the treatment of breast cancer. I am also concerned with whether prescriptions were issued to patients with the caveat that the long-term effects of augmenting levels of substances as powerful as hormones is unknown. That might be worth investigating, because the extent such caveats were generally issued is part of the breast cancer story as well.
In any case, I appreciate my interlocutors' suggestions and corrections.