If there’s’ one theme, one cause, that this blog has emphasized throughout the four years of its existence and the three years of its having resided on ScienceBlogs, it’s been to champion science- and evidence-based medicine over pseudoscience and quackery. Whether it’s refuting the lies of antivaccine zealots, having a little fun with some of the more outrageously bizarre forms of pseudoscience, railing against cancer quackery, lamenting how easily pseudoscientific quackery has infiltrated medical academia, complaining about drug companies rigging clinical studies, or trying to educate my readers about the complexities of cancer screening, support for science- and evidence-based medicine has been one of the two or three constants on Respectful Insolence since its very beginning. Indeed, I personally have even lamented at how little of the scientific method is taught in medical school, far too often emphasizing facts over critical thinking, leaving most physicians woefully unversed in the scientific method. This shortcoming has led some doctors to become creationists and some to embrace other pseudoscience. If anyone thinks that more science in medicine would be a good thing, it’s me.
That’s why I feel entirely justified in calling out medical correspondent Sharon Begley as being full of shit for her exaggerated attack on doctors entitled Why Doctors Hate Science.
Thank God doctors in the United States are free to treat patients as they deem best, free from interference by faceless bureaucrats. If bureaucrats were in charge, physicians might have to prescribe the newest hypertension drugs as a first-line therapy, do MRIs to diagnose back pain and give regular Pap tests to women who have had total hysterectomies. Oh, wait–they do. All these medical practices are common, despite rigorous studies showing how useless or wrongheaded they are. Definitive studies over many years have shown that old-line diuretics are safer and equally effective for high blood pressure compared with newer drugs, for instance, and that MRIs for back pain lead to unnecessary surgery. And those Pap tests? Total hysterectomy removes the uterus and cervix. A Pap test screens for cervical cancer. No cervix, no cancer. Yet a 2004 study found that some 10 million women lacking a cervix were still getting Pap tests.
Well, well, well, well. Begley appears to be trying to match Orac for sarcasm and insolence. Note the contempt dripping from her every sentence, contempt for doctors, contempt for medicine. The main difference is that Orac (usually) gets his facts right. For example, she takes what she thinks to be a slam dunk example of the pap smears after hysterectomy and uses it to indict doctors as “hating science.” However, let’s see what the Mayo Clinic website, for example, has to say about the matter:
If you had your uterus and cervix removed (total hysterectomy) for a noncancerous condition, you may be able to stop having Pap smears. However, if your hysterectomy was for a cancerous condition or you had your uterus removed but your cervix remains intact (partial hysterectomy), you still need regular Pap smears. In either case, regular pelvic exams and mammograms are recommended.
Indeed, I found the paper to which I believe Begley is referring, which came out of my alma mater ( the University of Michigan) in 2003. It refers to only the case of pap smears after hysterectomy for benign disease. Begley may indeed have a point that too many pap smears are still done after hysterectomy, by simplifying and mocking she completely undermined her point–not to mention showed that she doesn’t understand the issues involved. Either that, or she does understand them but decided to score cheap points against physicians instead of adding three words after “hysterectomy”: “for benign disease.” Not surprisingly, several doctors took her to task for her distortions. At the very least, Begley should have acknowledged that her blanket statement is more than a bit over-the-top and that the issue, at least in the case of hysterectomy after malignant disease, is not as cut and dried as she thinks. Even if it were, understanding why doctors have been too slow to give up this practice is not what she is about. Bashing doctors is.
Of course, I am not arguing that any of the shortcomings of medical practice mentioned by Begley are a good thing or that they shouldn’t be changed and, presumably, improved. I’m not even arguing that she doesn’t have a point. What I am arguing is that she chose to make that point in the most simplistic and inflammatory way possible. Regular readers of this blog know that the lack of science behind all too much medical practice is something I work to change every day and that I complain about it right here (and elsewhere) on average at least once a week, if not more often. Rather, what bothers me about Begley’s article is the way she takes some observations about how medicine is practiced and then makes the leap to conclude that doctors hate science:
It’s hard not to scream when you see how many physicians, pharmaceutical companies, medical-device makers and, lately, hysterical conservatives seem to hate science, or at best ignore it. These days the science that inspires fear and loathing is “comparative-effectiveness research” (CER), which is receiving $1 billion under the stimulus bill President Obama signed. CER means studies to determine which treatments, including drugs, are more medically and cost-effective for a given ailment than others.
And it’s hard for me not to scream when I see hysterical journalists leaping to the conclusion that doctors “hate science” and then likening them to “hysterical conservatives” on the basis of what, let’s face it, is her own personal political views, the existence of regional disparities in health care, many of which are systemic in nature and whose causes are not nearly as well known as Begley would lead you to believe; and shortcomings by doctors in applying science to medical care. Again, I call opportunistic ideological bullshit on this article. Unfortunately, instead of trying to understand why there might be regional disparities or why doctors might have difficulty adhering to science- and evidence-based treatment guidelines, Begley decides to go for cheap political points and use these disparities as an easy excuse to attack physicians and conservatives (which, although Begley may have trouble believing this, are not necessarily one in the same).
In fact, disparities in the use of various services and medical procedures are a highly studied area of medicine. There are usually many, not few, reasons for such disparities. Begley correctly identifies one possible contributor as medical culture, but that is only one of many potential reasons, and it is likely that a combination of multiple factors are at play. Perhaps one of the more prominent reasons is that the evidence supporting different interventions over others is not as clearcut as a journalist would like it for many conditions. Where ambiguity exists, clinical judgment rules, and different regions may indeed develop different norms based on regional differences. It’s not ideal, but it’s not necessarily due to physicians “hating science,” as Begley so histrionically puts it.
Moreover, patient desires have a profound effect on how doctors practice. I’ll mention one example of a surgical procedure that was adopted long before they were validated by science because of a combination of patient demand and surgical–shall we say?–entrepreneurship. I’m referring to laparoscopic cholecystectomy. Indeed, I started my residency in 1988 and went into the laboratory to work on my PhD in 1990. Pre-lab days, I learned how to do old-fashioned, “open” cholecystectomies, and I did a fair number of them, even though I was only a second year resident. Then I went into the lab for over three years. When I came out, no one was doing “open” cholecystectomies anymore. Laparoscopic cholecystectomy had taken over in a mere three years, at least in Cleveland. What had happened in three short years in Cleveland?
Patients demanded laparoscopic cholecystectomy, that’s what, and surgeons with an entrepreneurial bent gave it to them. One can argue whether surgeons stoked the demand or simply responded to it, but there’s no doubt that nonscientific factors came into play that did involve a significant, if not major component, driven by patient demand.
What happened is that the procedure had started in a few centers, and it spread like wildfire through a combination of patient word-of-mouth, some of which was stoked by advertising by surgeons looking to distinguish themselves from the pack, but that didn’t explain the speed with which the new procedure supplanted the old. Suddenly, older surgeons were finding themselves forced to learn the new procedure rapidly or risk losing all their gallbladder business, as patients wouldn’t go to surgeons who did the “old” procedure anymore, and referring physicians wouldn’t refer to them. Yet, had science and clinical trials validated laparoscopic cholecystectomy as the equal of the “gold standard” procedure? Not at all! In fact, in the 1990s, it was noted that injuries to the common bile duct were considerably more common after the laparoscopic procedures, requiring referral to expert biliary surgeons for repair. The rate was still very low, but it was several times higher than the rate after the open procedure had been. This was attributed to the “learning curve,” but I’m not sure that accounts for all of the difference. After all, it took at least a decade before the disparity in biliary injury rates shrank to the point where it is only somewhat higher after laparoscopic surgery. This was due to more and more surgeons doing more and more laparoscopic procedures more than anything else.
Another example that comes to mind is sentinel lymph node biopsy (SLN) for breast cancer, although this was not driven quite as much by patient demand as laparoscopic cholecystectomy. SLN biopsy involves injecting dye into the breast and following that dye to the first lymph node under the arm to which it drains. That lymph node is then removed and examined. If there is no tumor there, with a high concordance, there is no tumor in the rest of the lymph nodes. From the late 1990s into the early 2000s, SLN biopsy rapidly supplanted the older procedure, the axillary dissection, which involved removing all the lymph nodes under the arm, even though it was unproven.
I say “unproven,” because there was a small rate of false negatives, in which a “negative” SLN missed a lymph node or lymph nodes with cancer. Theoretically, that meant that SLN could lead to some patients being undertreated because they are thought to be node negative when they are in fact node positive. Where I worked at the time, we did the right thing, viewing the procedure as still unproven, and only did it under the auspices of a clinical trial, namely the NSABP-32 trial until that trial’s accrual was complete. During that time, many women asked for SLN and ended up going elsewhere when informed that we would only do it as part of a clinical trial that could lead to them being randomized to undergo axillary dissection. Fortunately, short term results reveal the results of SLN to be equivalent to those of axillary dissection, but the procedure is too new for us to have long term data equivalent to what we have for axillary dissection. No matter. SLN has become the standard of care. You can look at it as either accepting that a small percentage of women who undergo SLN biopsy will be undertreated as being worth the decrease in morbidity, such as lymphedema, from removing only one to five lymph nodes under the arm instead of nearly all of them, or you can look at it as most women having decided that they would prefer not to risk nerve damage or lymphedema from an axillary dissection. Either way, the result is the same.
What Begley doesn’t seem to understand is that, although medicine should be primarily science-based, it is never going to be pure science. Too many factors other than science impact how medicine is practiced. These include the doctor-patient relationship, availability of resources, third party payors, patient preferences, the influence of pharmaceutical company advertising and P.R., the individual circumstances of each patient, and, yes, to some extent physician culture, the latter of which was pretty much the only factor to which Begley attributed all manner of evil to. Add to that the frequently conflicting and unclear evidence on so many clinical questions, and it’s no wonder that there are regional variations. And, yes, I don’t deny that there is a certain resistance among many doctors to being told what to do that sometimes leads to resistance to anything they perceive as mandating how they must take care of patients, as well as a desire to “fit in” among one’s colleagues. Where Begley errs is in emphasizing these two traits over everything else and ignoring systemic factors. In essence, she takes the easy, simplistic explanation over the more accurate, nuanced reasons.
A lot like the conservatives resisting science that Begley castigates.
Also, it’s not that doctors “hate science.” There may be a small minority who do, in fact, “hate science” because it restricts their actions (the physicians who belong to the Association of American Physicians and Surgeons immediately come to mind), but in fact most physicians actually believe that they are practicing based on science. They really do. One problem is, as both Steve Novella and Val Jones have pointed out, is a lack of resources to help them keep up with the latest scientific literature and have access to the latest science-based recommendations at the point of patient contact. Another problem, as I have pointed out before, is that most physicians are not trained in the scientific method, at least not very well. They truly do not know how to separate the wheat from the chaff, and are easily swayed by anecdotal evidence, be it their own “personal clinical observations” (shades of Dr. Jay!) or those of their colleagues, not realizing how easily anecdotes can mislead and how easily a series of anecdotes in a practice or region can lead to groupthink. Also, as Steve points out, the more certain the scientific literature is with respect to the correct treatment for a condition, the less regional variation there is in the use of that treatment. Ambiguity in evidence leads different regions to develop regional standards.
Unfortunately, not only does Begley spew oversimplified drivel as if it’s the Gospel Truth; she sees dark conspiracies against her favored solution to the problem, comparative clinical effectiveness research (CCER), painting those who are skeptical of it as either driven by money, ideology, or just downright hubris. I’m disappointed in her because I say this as someone who in general likes the concept behind CCER, namely testing different treatments for a condition head-to-head against each other and letting the chips fall where they may. What’s not to like about that, if you’re a booster of science-based medicine like me? Even so, apparently Begley is going to have to add me to her list of a money-grubbing, unscientific, clueless, conservative wingnuts, because I actually do share some of the concerns she heaps scorn upon in her article, namely that the fruits of such research could become mandates. Val Jones explains why:
Although comparative clinical effectiveness research is distinct from comparative cost effectiveness research – it is likely that payers will use CCER to build their formularies. This means that even though the government (at this point in time) is not mandating coverage decisions based on CCER, health insurers are going to be using the information liberally to justify coverage preferences and even potential denials of coverage.
That is precisely why many physicians are concerned about government-mandated CCER. It’s disingenuous in the extreme of Begley–and Dr. Elliott Fischer, whom she quoted–to deny so vociferously that CCER has anything to do with cost control and to paint those who oppose it as a bunch of right wing ideologues. True, some of them are; but many of them are not. It is also true that there is language in a Congressional report about CCER that raises a bit of concern even to me:
By knowing what works best and presenting this information more broadly to patients and healthcare professionals, those items, procedures, and interventions that are most effective to prevent, control, and treat health conditions will be utilized, while those that are found to be less effective and in some cases, more expensive, will no longer be prescribed. Substantially increasing the Federal investment in comparative effectiveness research has the potential to yield significant payoffs in reducing health care expenditures and improving quality.
Maybe it’s not being such a “scaremonger” (as Begley so quaintly put it in the subtitle of her article) to wonder how CCER will be used, after all, as Val Jones did:
There’s also the question of stifling innovation. Blockbuster drugs are rarely discovered in a vacuum. They are the result of incremental steps in understanding the biology of disease, with an ever improving ability to target the offending pathophysiologic process. The first few therapies may offer marginally improved outcomes, but can lead to discoveries that substantially improve their efficacy. If an early drug is found to be only marginally better than the standard of care, an unfavorable comparative effectiveness rating could kill the drug’s sale. Without sales to recoup the R&D losses and reinvestment in the next generation of the drug, development may cease for financial reasons, and the breakthrough drug that could cure patients would never exist.
A theoretical concern, I agree, but it’s not an unreasonable concern at all.
That being said, because I support science- and evidence-based medicine, I still support CCER. My support for only the most rigorous science and clinical research does not, however, keep me from some mild concern about whether the results of CCER will evolve from useful research results that help to guide treatment choices in a science- and evidence-based manner into a government straightjacket on medical practice. Like Steve and Val, I’d love to see the results of CCER serve as a means of helping physicians to become more science-based, not of telling them what to do and reducing treatment algorithm’s to “cookbook medicine,” although for some conditions for which the evidence is particularly clear I also agree with Steve that strong guidelines based on the scientific literature and CCER are not a bad thing. The problem is: How to blend the two approaches, top-down and bottom-up, optimally?
Another issue I’d love to see addressed but likely never will comes in the wake of Senator Harkin’s little woo-fest last week in the Senate. Personally, I’d love to see CCER applied to “alternative” medicine versus conventional medicine in a rigorous way. Indeed, if CCER is implemented, it would be one way of driving home once again how ineffective the vast majority of “alternative” medicine is and, if CCER is ultimately used to decide which treatments to reimburse, might blunt the drive “integrate” woo with scientific medicine. A guy can dream, can’t he? Unfortunately, what’s more likely is comparing woo to woo, such as head-to-head trials of homeopathy versus reiki, for instance.
In the end, though, for all the worship of CCER as some sort of panacea for the ills of regional variation in treatment practices, the continued use of treatments that are inferior, or wasteful treatments, I don’t see CCER as the be-all and end-all of scientific medicine or even its next stage. Indeed, Begley’s faith in CCER and her castigation of physicians who remain skeptical of it show that she’s utterly clueless about ever having actually tried to use guidelines such as the ones that are likely to derive from the results of CCER. After all, in oncology, we’re far ahead of the curve in developing and using science-based treatment algorithms. Indeed, just peruse the algorithms in the NCCN treatment guidelines for breast cancer. It’s 121 pages, and at our cancer center we try hard to practice within its guidelines. Even so, at our weekly breast cancer tumor board, almost every session we encounter a case that does not fit well into the treatment pathways therein.
There’s a reason for that. It’s because good medicine should be based first and foremost on science, but science alone can never completely dictate how medicine should be practiced in many cases. There are too many other factors apart from science at play. If Sharon Begley’s article is any indication, she probably thinks that my saying so must mean that I must “hate science,” just like those doctors she castigates as lazy, money-grubbing, clubby, close-minded individuals who don’t need no steekin’ science telling them what to do. Unfortunately, her simple-minded and misleading conflation of criticism of CCER with wingnuts, pharma shills, ideologues, greed, insular groupthink, and just plain hubris throws a lot more heat than light on the discussion.
Maybe that was her purpose all along. On second thought, strike the word “maybe.”