A petition to the British Medical Journal (BMJ) asking for Makary’s paper on medical errors to be retracted has received over 100 signatures. I, amongst others, criticized Makary’s analysis. But I shall not be signing the petition.
I applaud the editor of the BMJ, and the section editor, for publishing Makary’s analysis. The analysis was in the right journal in the right section. To be clear, I believe the editors did nothing wrong publishing the paper. The role of medical journals is not to tell us how to think, but what to think about, which the BMJ achieved.
Makary’s analysis was provocative. I almost developed laryngeal edema reading it. The analysis advanced the discourse about medical errors. The analysis made doctors think, think beyond dull platitudes. The analysis broke the taboo of questioning the number of deaths from medical errors. Whether or not this was the intention of the authors is immaterial. The BMJ achieved a vigorous debate about medical errors. Not since the NEJM series on physician relationship with industry has any publication led to such intense discourse.
Some say that the paper should be retracted because it was a scientific paper, not an opinion, and as science it is even more dubious. This is not true. The paper was an “analysis.” Look up the definition of “analysis.” But were it research, so what? Methodology for this this type of research is flawed. If all research with fallible methods are retracted, that’ll end medical research.
Some say that the paper should be retracted because the media runs with the story amplifying the error. The media is fickle. The media runs like decerebrate pigeons every time a nutritional study is published. The media doesn’t know its posterior from elbow and I, for one, find this state of affairs comical. I find it hilarious when someone, who follows nutritional advice as punctiliously as a celibate nun listens to her pastor, is later told that it is all rubbish. Nothing gives me more pleasure than seeing mirthless optimizers taken down the garden path.
It is true that the media will be quoting “medical errors are the third leading cause of death” for years. However, the root cause of this chicanery is not Makary’s analysis but the Institute of Medicine (IOM) report: To Err is Human, which stated that jumbo jets of hapless patients were crashing daily. The report created a new science and a professional empire. Should the IOM retract that report?
The problem with retraction was highlighted by my 10-year old son recently. Arguing with his younger brother he said “you must be feeling light because you left your brain back at school.” When I heard his gratuitous remark I was enraged, “Take your words back, now,” I yelled. After deep thought, and tears in his eyes, my son said “what good will that do? I already said them.” I gave him a time out for that reply. But his point was well taken – what is said can’t be unsaid.
Retraction has become like corporal punishment. Spectators would flog to the stadiums in Taliban-controlled Afghanistan to watch with glee when adulterers were being stoned. When a medical journal retracts a study there is glee amongst spectators, the glee of an aspiring saint watching a sinner unfold. Read the comments in Retraction Watch, a consumer watchdog which keeps an eye on retractions.
Inquisitive scientists seeking the objective truth embrace uncertainty. There is, however, an emerging rank of science aficionados who treat objective truth like religion. These are the Truth Jihadis. They’re indistinguishable from religious fanatics on PET scans. Truth Jihadis are not seeking the truth. They’re seeking control. “I am for the truth” is the dullest of dullest platitudes. Science, an enterprise which grows by curiosity, not control, cannot be chaperoned by the incurious.
Calls for retraction have become yet another avenue of redress for an affluent society, which has exhausted the legal system, and which increasingly yearns for perfection, is intolerant of uncertainty and can’t handle variation. Americans might introspect this November how they became more hyper regulated than the French, indeed more hyper regulated than Stalin or Mao thought it might have been possible to be hyper regulated.
When Bleyer and Welch published in the NEJM a study which showed that mammograms may have overdiagnosed breast cancer, the reaction amongst my peers was brisk and predictable. It started off with “how could this pass peer review?” – the incredulity cleverly according the skeptic with unchallenged sophistication. I asked a few, what should pass peer review? Was there a normative frame? Who developed the norm? I was met with silence.
Some asked the NEJM to retract the paper. Many called the study “fatally flawed.” One said, in a tone perfectly capturing sincerity and gravity, “this paper will kill women.” Nothing quite inspires righteousness more than believing you have God or statistics on your side.
If science has been hijacked by special interests it won’t be rescued by social justice warriors. I recall an open letter signed by UK academics asking to boycott Israel’s academia. Scientists and politicians, like pigs and farmers in Orwell’s Animal Farm, have become indistinguishable.
The most famous retraction is Wakefield’s discredited paper in the Lancet about the link between autism and vaccination. It must be understood why the paper was retracted. It was not retracted because the link was later shown to be wrong. As the editor of the Lancet, Richard Horton, reminded with frustration – science is self-correcting. It was not retracted because it unleashed havoc, including a congressional hearing. I’m always amazed how congress chooses the most dubious entities to debate over. It takes skill to be so consistently and exceptionally useless. The paper was retracted because Wakefield was struck off the General Medical Council, because he obtained the tissue samples of the children through deceit. The study was retracted because Wakefield had been dishonest.
The most disturbing part of the autism-MMR debacle is not that the Lancet published the study. It is that the Lancet had to retract the study before people believed the dubiousness of the link. Wakefield had to be discredited (he deserved to be), shown to be a bad man before people believed in vaccines again. It was not the scientific evidence that persuaded people, but the (bad) character of the researcher.
Thus, we cannot debate facts on their own merit but we must involve the person behind the facts. If he is sincere, “a good guy,” we must believe what he has to say. If the evidence is dubious he must be dubious, too. The religious undertones in the scientific enterprise are ironic. Pope Urban VIII is probably laughing in his grave.
It is a small logical step from saying that because dodgy scientists produce dodgy science, that dodgy science is produced by dodgy scientists. And an even smaller step from concluding that because dodgy research is later contradicted, that research which is later contradicted is dodgy. Retraction is a Spanish inquisition, of sorts. Calls for retraction are reminiscent of pitchfork-carrying righteous puritans in Salem. I’m getting tired of puritans.
It is precisely because Makary’s analysis is flawed that it should not be retracted. For it to be retracted would be to set a standard for medical journals that is neither achievable nor desirable. Knowledge grows by error, not perfection. Perfection merely picks out the lowest-lying fruit which everyone can agree about. Makary’s study received several online comments on the BMJ website. Voices were heard, which would not have been heard but for this paper. Were I the editor of the BMJ I’d open a bottle of champagne and celebrate a job well done.
“I asked a few, what should pass peer review? Was there a normative frame? Who developed the norm? I was met with silence.”
A great point. Right now peer review allows all sorts of garbage to be published without any standards of what that means, other than the ones publishing the journal agree this would be a good article. It seems like some publishers are getting into the game of remaining provocative rather than scientific or informative.
I enjoyed reading this despite my disdain for the term ‘jihadi’. It makes my skin crawl. But I also loved the sentence: “Nothing quite inspires righteousness more than believing you have God or statistics on your side.”
Anyway, the way we recoil with medical errors coverage, I’m sure people from law enforcement/judicial system recoil reading about prisoner exoneration. I could come up with more analogies but you get the point. There are many things we could do better as individuals and also as a system. Seeing the skepticism with media coverage of science please read this from the New Yorker http://www.newyorker.com/news/news-desk/the-mistrust-of-science
Lastly, a ‘one size fits all’ system for retraction is a disservice to research and its goal of advancement of human knowledge. Retractions have a place but never when our dislike for the results or its media coverage are the only talking points.
I wholeheartedly agree with you that the retraction is our equivalent of the book burning. The practice points to a wider crisis in science brought on by a gross misunderstanding of what the scientific enterprise unfolds.
In his 1962 essay “The Republic of Science,” Michael Polanyi states that “The professional standards of science must impose a framework of discipline and at the same time encourage rebellion against it.” The dynamic tension between discipline and rebellion cannot bear fruit if we brandish the specter of retraction at every turn.
“If science has been hijacked by special interests it won’t be rescued by social justice warriors.” You hit it on the nail. People should reflect on the broader relationship between the funding of science and the crisis in peer review. As I have pointed out recently, peer review is a relatively modern invention that underscores how far we have strayed from Polanyi’s ideal. http://alertandoriented.com/snippets/sciences-funding-problem/
I going to have to think about this one.
I see your point, which is well stated. If the world were an orderly research laboratory or a badminton match and everybody played by the rules, I think I’d agree completely with you.
The problem is that the stamp of the BMJ or NEJM or JAMA has unfortunately become a validation of scientific proof. The label is placed on information in the same way that the logos of Harvard University and Stanford are used by marketers to sell their products.
One only has to look at the long term impact IOM report to guess at what the potential impact of the Hopkins report could be over the next twenty years. Badly thought out government rules will be based on the assumption that this finding is accurate and that innocent people need to be saved from the grips of organized medicine. (The words will be uttered with the same tone as “organized crime.”) Future “analyses” will build on Makary’s conclusions. Lawsuits will be filed that will drive unlucky physicians from medicine. Costs will rise as we attempt to reduce an unknown error rate to zero.
I’m going to have ponder this one, but I’m doubtful .
If we’ve allowed the stamp of JAMA, BMJ and NEJM to substitute for scientific validity, then we must agree with Saurabh’s general point: science is now an activity of signaling!
OR my friend, we must argue that journalism needs to rethink both how it covers science and how to manage its own scientific lens in the age of reporter driven research and reporter-generated data …
Every time someone dies there is someone else–who is exactly the same age and sex and has exactly the same disease–who lives a few more days or weeks,etc. The person who dies must have had an error in his treatment compared to the patient who lives. How else to explain it?
Therefore medical errors must be the leading cause of death because half of the patients dying are explained by the above and the remaining living patients from the above eventually have to go through the same logic when they die at a future date, so that eventually 100% of us die from medical errors.
This must make sense because anyone who dies could have eked out a few more days with better treatment.
Therefore, we all die from medical errors.
Bill, good response.
Good medical care might have more deaths than great medical care which might have more deaths than the best of medical care. We strive in the direction of the best of medical care, but good medical care is not a cause of death. The disease is the cause otherwise the physician need not be involved.