Who’s Afraid Of Science Studies?

A very interesting article was published recently on the New England Journal of Medicine, with a historical view of the development and importance of randomized clinical trials in medicine: Bothwell LE, Greene JA, Podolsky SH, Jones DS. Assessing the Gold Standard — Lessons from the History of RCTs. N Engl J Med 2016; 374:2175-2181 DOI: 10.1056/NEJMms1604593

The abstract: “Over the past 70 years, randomized, controlled trials (RCTs) have reshaped medical knowledge and practice. Popularized by mid-20th-century clinical researchers and statisticians aiming to reduce bias and enhance the accuracy of clinical experimentation, RCTs have often functioned well in that role. Yet the past seven decades also bear witness to many limitations of this new ‘gold standard.’ The scientific and political history of RCTs offers lessons regarding the complexity of medicine and disease and the economic and political forces that shape the production and circulation of medical knowledge.

Three of the authors have medical degrees, I know one of them – Greene – who is a young, brilliant historian of medicine (his book Prescribing by Numbers made me rethink my own considerations on the definition of disease in contemporary medicine), who cannot therefore be accused of ignoring the subject matter they write about. In effect, they produced a solid, scholarly review of the subject. The article seems to me as non-controversial as may be, considering past considerations from mainstream authors about the whole enterprise of conducting clinical trials, such as the proposal by Richard Smith, former editor of the BMJ, of an outright moratorium in publishing their results since they became, in his words, a mere marketing tool for the pharmaceutical industry (see Smith R. Medical journals are an extension of the marketing arm of pharmaceutical companies. PLoS Med. 2005;2(5):138. DOI:10.1371/journal. Pmed.0020138).

But this well-founded discussion was met with a conniption in some quarters. A blogger wrote a piece criticizing the NEJM for having published it (The NEJM and Clinical Trials: What’s Going On?), from which I would like to quote just a tiny bit: “That, to me, is the social science worldview in a nutshell – that everything, simply everything, is deeply entangled in social conditions, economics, and politics. Take an NMR spectrum? A political act. Weigh out some copper sulfate? Politics. I sometimes think that that’s my vision of Hell. (…) Behind some of this stuff is a worldview that holds that there isn’t such a thing as ‘knowledge’ at all, just power struggles and wishful thinking.

This is in my opinion as willfuly ignorant as the blogger who stated that she wouldn’t eat anything she could not pronounce. The whole field of science studies is reduced to this ridiculous caricature by someone who clearly refuses to engage seriously with the discussion proposed by those who work in that area. Yes, science does have political and economical underpinnings, and failing to acknowledge that can lead to scientism and excessive medicalization. I really believed that the “Science Wars” of the nineties were over, but apparently not everyone got the memo.

The issue here is not that there is no such a thing as “knowledge”, but that we have to examine what we call “knowledge”, how it is produced, how it circulates among different social groups and what are the consequences of its use. Redefining the threshold for a given clinical parameter, for instance – say, blood cholesterol – is not an act of “pure science” (if that even existed), it is a negotiated process, an idea that seems to give that blogger the heebiejeebies, but that was apparently supported, for instance, even by a relatively conservative author such as Karl Popper, who wrote that “scientific objectivity can be described as the inter-subjectivity of scientific method” (“The sociology of knowledge” in Stehr N, Grundmann R. (orgs) Knowledge: critical concepts. Volume V: Sociology of Knowledge and Science. New York: Routledge, 2005, p. 68).

And that redefinition has very concrete consequences, like almost tripling the number of candidates to be treated with statins, something with obvious economic and political implications, which are present, even if implicitly, in the process of negotiating the “new normal”. At every corner multiple interests intersect in the path of the production of scientific “knowledge”, especially in Medicine, and ignoring this can be naively dangerous. Examples of how clinical trials have been manipulated over time abound, and even when they are done with the appropriate methodological concerns, they are not Revealed Truth, but a relevant guide for clinical action with important limitations. Acknowledging such limitations and the complex interactions that go into making science, on the other hand, does not mean denying that reliable “knowledge” that can be acted upon does not exist, but quite the contrary. Two exponents of the field that was simplistically panned by the author of that blog post, Harry Collins and Trevor Pinch, have a position that I subscribe wholeheartedly (I have been using this citation so much – with due credit, of course – that I fear one of this days they will ask me for copyrights…), expressed as follows: “Science may be wrong (…) but this does not make the opposite view right. In the absence of careful research about the opposite view, science is probably the way to bet. This is even more likely to be the case if science is continually put under scrutiny.” in Colins H & Pinch T. Dr. Golem: How To Think About Medicine. Chicago: The University of Chicago Press, 2005 (p. 202)

And while we are at it, what is “knowledge” anyway? Answering this is in itself a philosophical question of the type clearly despised by our colleague, but I’ll leave that for another day.

POST-SCRIPTUM: After having posted this, I recalled reading a post by PZ Myers that addressed a similar problem, and I found it: Emotionally invested in despising philosophy. He writes “Yet again, people are asking why are so many smart people such idiots about philosophy? I have a different answer than you’ll find at that link. It’s because so many smart people are idiots about psychology. I deal with a lot of atheists, and one of the many flaws in that group that have been coming to the fore lately is the obliviousness they have to their own motivations. Atheists are all about the scienceyness. Good people are rational, objective, and unemotional, which whether they are aware of it or not, is a value judgment built on emotion. There is a lot of self-esteem-building going on, centered around who is smarter than who, who can build the most logical argument, and who is best at being aloofly superior. It’s all very annoying.

Very annoying, indeed.


3 thoughts on “Who’s Afraid Of Science Studies?

  1. I can understand your being annoyed at my post, but I’d be interested in seeing you take on the Vinay Prasad criticism of the NEJM article that I linked to, since he doesn’t veer off into criticism of science studies. And I can tell you, for what good it may do, that I don’t despise philosophy – although I do wonder about some of the avenues it’s been traveling down in recent years.

  2. Hi, thanks for stopping by! As for Prasad’s criticism, I think it is predicated on a flawed premise: that the authors are “against” RCT, and from that follows an unnecessary and at times foolish defense of those. I do not have the time or patience to go over all his claims point by point, but just to give one example, his argument about metanalyses being the same thing as RCTs is a poor one, and his analogy is even worse; yes, French fries are different from raw potatoes, I love the former and would never even think of biting the latter. The fact that metanalyses (and systematic reviews, for that matter) take data from RCTs as a starting point does not negate them as a new and innovative methodological step. By the same token, RCTs are a collection of data from individual consultations, so it would be the same as a case series? His wave of hand with regard to psychotherapy and RCTs is equally arguable; the issue at hand is: are there truly *individualized* issues with regard to therapeutics? If the answer is yes, RCTs will not apply in such cases, and will at most provide ballpark guidelines. Which is in the end what it boils down to when the rubber hits the road; no matter how many RCTs are done, the translation of the information they provide to the bedside (or across the table) decisions is not automatic, and experience does make a lot of difference. If it were not the case we (I have left practice a long ago but am still a MD) would have long been replaced by computers, which are way better at storing and retrieving information. Finally, as for other methods, he quotes himself pragmatic trials, and just consider the panoply of methods that the people who work with technology evaluation have developed. RCTs are an important tool, but by no means the end-all for medical reasoning. Another issue is that if we consider the use of placebos, the RCTs depart from the implicit admission that placebos do work – so we could devise a research programme that instead of doing away with that effect would aim to exploit it to the maximum, something like is being done at this place: http://programinplacebostudies.org/ (I don’t know if the quality of their research is good or not, but the idea is very interesting, to say the least).
    But the main point for me is that while criticism is a vital part of science and academia, so is respect for different disciplines and methodological approaches. What Prasad did, and to some extent your post as well, was to dismiss beforehand a body of scholarly work because it doesn’t fit the model he seems to be more comfortable with (the initial sentence stating that the historical review is “boring” is rather childish, in my view). And in addition he seems to attribute to the authors of the article ulterior, mischievous intentions (what a friend of mine quite aptly dubbed “the hermeneutics of suspicion”) with no good reason for that. We can discuss the substantive issues without resorting to accusations that are tantamount to ad hominem arguments.
    In the end, the gentleman doth protest too much, methinks.

  3. Oh, and another thing: RCTs have an important limitation, which they share with their siblings, risk factor epidemiological studies. As the global population ages, multiple pathologies with multiple concurring therapeutics will be the norm, rather than the exception, and as the number of concurrent interventions goes up, the complexity of the possible interactions make it increasingly harder to keep the “one factor, one outcome” model… There has been a lot of discussion on how to apply systems theory to epidemiology, and this is something that clinical epidemiology should consider as well.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s