In his article about the Lancet retracting the Wakefield paper, Steven Novella brings up a major problem in the scientific research community that I wasn't even aware of. I had assumed that taking 12 years to retract a flawed paper was an anomaly, due in part to the celebrity standing that the paper has, and its wide acceptance by the public prior to (and after) the retraction. But his example of the Cha/Wirth/Lobo study is even worse, with more blatant fraud, and yet at first glance, to the common layperson, is still held on the same pedestal as every other article from a scientific journal.
Someone should have already noticed that the statistics are too perfect in this article. The patients with intercessory prayer (IP) had a 50% pregnancy rate, compared to 26% in non-IP patients. The patients with IP had a higher implantation rate as well, 16.3% versus 8% in the non-IP group. These figures are almost exactly double, +/- 1%. What are the odds of that happening? Why would intercessory prayer increase the success rate by exactly 2 times? It makes no sense, unless you throw in someone with no MD, no affiliation with the university, and an agenda. Daniel Wirth, aka John Wayne Truelove, who had been charged with dozens of counts of mail fraud, bank fraud, and holds an MS in parapsychology, was for some reason allowed to lead the study and get published. Why and how did Cha and Lobo even manage to get involved? Were they paid, coerced, or just equally shady?
And yet these papers survive after inadequate peer review, inability to replicate results, and clear evidence of wrongdoing, while the Wakefield paper took 12 years to be completely retracted. I used to have much more respect for articles printed in scientific journals, but now, especially if results seem surprising or counterintuitive to me, I'm going to have to look more closely at what was done, by whom, and why.
Wirth/Truelove info: http://www.csicop.org/si/show/columbia_university_miracle_study_flawed_and_fraud/
Article Summary: http://www.ncbi.nlm.nih.gov/pubmed/11584476
Emma,
ReplyDeleteI think you bring up some good points here that hark back to Gould's idea of the blind faith we place in numbers. I think it's more common than not to place a certain degree of trust and authority in the information published in medical journals as being accurate and published with public welfare in mind. It seems that we also still maintain some of that blind faith in statistics when we really ought to be more skeptical. This is a good reminder that just because something is published online, or in a medical journal (even if it's peer-reviewed) doesn't mean its one-hundred-percent accurate.
Unfortunately it seems peoples names get slapped onto things that they themselves don't review, I believe Lobo's case (or so he says). This means that being thoroughly certain about what documents you are signing is very important, especially when it comes to being a secondary author on a publication (Be wary, I sense a disturbance in the journals). People can put words under your pen that may lead to your defamation and thus the possible end of your career. Sadly, I think this happens all too often in the cases we have covered in class thus far. Interestingly enough this also happened (in a way) to the Earth and Planetary Sciences last key note speaker, Dr. Surovell. I only offer a word of warning, so stay famous and a real scientist.
ReplyDelete