Last week the Times of London revealed inside information from the General Medical Council (UK, responsible for physician licensing) of an ongoing investigation of Dr. Andrew Wakefield and from its own investigation. This revelation recalled other instances of fakery from reports of sectarian medicine (“CAM”) successes. The Medical Council information contained evidence that the data from the now famous Wakefield cases used to claim an association of the MMR vaccine with childhood autism and inflammatory bowel disease were misinterpreted, altered, and to some extent, faked. The report and history were reviewed by David Gorski last week. In eight of twelve cases, the dates of autism onset were “rearranged” to fit the needed time association, and many small bowel biopsies were “reinterpreted” to show inflammation.
The Wakefield claims were long suspected by reputable medical scientists and skeptics as being erroneous or fraudulent. Note: the Council hearings are still in progress, and the Times report is subject to legal complaint. The original details can also be seen in the Times articles.
This expose’ adds to a growing list of reports with erroneous and faked information in medical journals used either for economic reward, undeserved fame, or to promote ideological claims for medical sects and cults. Although history of erroneous or false claims goes back hundreds of years, the altering or synthesizing of data passing the recently conceived peer review system is new, illuminating defects in the journal peer review and editing system.
If journals were invented in the 18th century, and operated like journals of today, Mesmer’s demonstrations might have been published, and the Ben Franklin and Antoine Lavoisier and French Academy’s disproof might have been rejected. (Laughter here.) But journals continue to make major goofs in publishing implausible results despite the popularity of a famous specialty journal for that purpose.
Examples vary from acceptance of language manipulation – “alternative,” “healing,” “integrative,” etc., to the fakery of recent papers showing effects of prayer. The two famous studies of prayer in the cardiac care unit ranged from the unadmitted breaking of the blind in the Bird study (So Med J 1988; 81:826-826) to unadmitted imbalances of subject and control groups (Harris, Arch Int Med 1999;159:2273-2278.) And from those to the likely fakery of the distant prayer study of in vitro fertilization (Cha, Wirth, Lobo; J Reprod Med 2001:46;781-786) in which three separate prayer groups on two continents improved pregnancy rates in a group of women on a third continent by an implausible 100 percent. In all of these cases, the papers passed peer or editorial review despite the methodological defects that were picked up by us skeptics (K. Atwood, K.Courcey [an RN] B. Flamm, and others.)
Adding insult to the above, Annals of Internal Medicine published a systematic review of intercessory prayer (Astin et al, Ann Int Med, 2000;132: 903-910) containing not only the Bird and Harris studies counted as positive, but also the Targ study on brain tumors, found by reporter Po Bronson to have had its end point altered by the authors when the primary one showed no effect.
Topping that, the same review contained four more papers by the same co-author of the of the in vitro fertilization study, counting two of them as positive. We have evidence that those four studies also were never performed. The author of the four was just released from Federal prison where he had been for four years on a fraud conviction.
Topping even that, the review claimed to have found 15/23 papers that met their criteria to have been positive. Yet many articles were not from commonly recognized peer reviewed journals, but nine were from “CAM” advocacy journals, six were from “alternative” nursing journals, and one was from a medical news magazine. (Sampson W. Are systematic reviews…sufficient as well as necessary for assessing medical effectiveness of prayer? Sci Rev Altern Med, 2000;4:(2)12-16. ) The review was made famous by national TV reporter Timothy Johnson of ABC, and by Hillary Clinton, who proclaimed it to prove the power of prayer.
More recent famous cases of misinterpretation or misrepresentation include the Benveniste homeopathy lab tests (I will present those data later), the several clinical trials of psychological support for cancer (Lancet again and still going on) and even the original survey of “CAM” use in the US published in 1993 in the New England Journal. All of these and more led to increased false belief in the populace, and may have cost a number of lives indirectly by furthering belief in .he “CAM” cause.
Beyond the factual details, we are faced with several meta-phenomena (or, are they more basic underlying phenomena?) One is the inability of any system to control the activities within it; and to decide which of several solutions works best.
How should journals deal with fraud and misrepresentation? If they detected misrepresentation perfectly, there would be few positive articles on quackery and sectarianism like “CAM“. Richard Horton, editor of The Lancet stood by the Wakefield paper through much of the criticism. Even though he vouched for the report’s authenticity, as I recall, he is of the school that favors research papers and commentaries having maximum freedom for publication, and that the normal give and take of scientific replication and debate will in the end determine validity. Others continue to develop stricter guidelines and safeguards against fraud and misrepresentation. At times the International Committee of Medical Journal Editors policies look like a stimulus bill or the military budget. Nevertheless, as per John Ioannidis, even when working ideally, most research findings are “false,“ or as I’d prefer, erroneous. But there is a wide difference between unrecognized bias and error and scientific fraud.
An entirely laissez faire policy is a cop-out, producing little to no need for editors or reviewers other than copy editors. The present situation is a matter of peer review failure, in my opinion. No doubt there are many reasons for the failure, but failure it is, because peer review advocates promise more than can be delivered.
Behind the failure lies human nature. Who has never tried to rationalize inclusion of an inconsistent finding , or the inverse, discard a finding that did not fit what one had in mind? (Rather than investigate and try to explain it rationally.) Psychologists have names for the ways the human mind leads to wrong conclusions and wrong beliefs. They are universal.
Then, what makes people who submit falsified manuscripts publish tick? That’s minor mental error gone berserk. Reading Dr. Wakefield’s protestations recalls the arguments for krebiozen, oxygen therapies, Laetrile, arguments against HIV as cause of AIDS, complete with imagined conspiracies, paranoid arguments of persecution, railing against unjust laws and rational interpretations. Not long ago, Val Jones suggested looking into quacks’ personalities for answers. Personalizing the necessary terms produces legal problems, especially in the UK and Canada, and also problems of how to study subjects who reject study. Yet Val opened a vast area for questioning. I’ve called the phenomenon “localized intellectual deficit,“ and “intellectual sociopathy.“ But so far it doesn’t help much. maybe the media haven’t caught on?
Two last questions: How much fraud is recognized on peer review? Perhaps this has been studied…I don’t know. Does the allowance of fraudulent and misrepresented material augment anti-sciencers’ arguments against rationality and science? Speculate among yourselves.
For the time being, we live with a need for after-publication recognition of fraud and misdirection, and with satisfying ourselves living within the limitations of our observations. And also recognizing when reviewers and editors also fail.