The Times (London) reports:
THE doctor who sparked the scare over the safety of the MMR vaccine for children changed and misreported results in his research, creating the appearance of a possible link with autism, a Sunday Times investigation has found….
The [original] research was published in February 1998 in an article in The Lancet medical journal. It claimed that the families of eight out of 12 children attending a routine clinic at the hospital had blamed MMR for their autism, and said that problems came on within days of the jab. The team also claimed to have discovered a new inflammatory bowel disease underlying the children’s conditions.
However, our investigation, confirmed by evidence presented to the General Medical Council (GMC), reveals that: In most of the 12 cases, the children’s ailments as described in The Lancet were different from their hospital and GP records. Although the research paper claimed that problems came on within days of the jab, in only one case did medical records suggest this was true, and in many of the cases medical concerns had been raised before the children were vaccinated. Hospital pathologists, looking for inflammatory bowel disease, reported in the majority of cases that the gut was normal. This was then reviewed and the Lancet paper showed them as abnormal….
Thanks to InstaPundit.
UPDATE: By the way, I should note that if there was misreported data here, the failure was not a failure of “peer review” (at least as the term is, to my knowledge, generally understood). Peer review involves reviewers checking the written article to see if it has errors in its reasoning, or fails to consider important and already known contrary data or arguments, or is duplicative of preexisting research.
Peer review, as I understand it, generally doesn’t involve source-checking the article to make sure that it accurately reports the underlying data. Law review cite-checking in some measure does that, but it’s unusual as far as scholarly journals go (partly because it relies on the unpaid labor of student editors, who are common only in law; faculty editors and peer reviewers aren’t willing to do this sort of thing). And I doubt that even cite-checking would usually go so far as to demand that the scholar provide the underlying medical charts on which the author based his data. Usually the tables from the author would suffice, especially if the author says — accurately or not — that the underlying charts aren’t available (for instance, because of privacy concerns at the institution that maintains the charts).
To be sure, The Lancet and other such journals could institute systems in which all the underlying data is checked as much as possible, though even that can’t catch certain kinds of errors (and especially certain kinds of deliberate falsifications). But it would require a tremendous amount of extra work. Perhaps some learned journals do invest this kind of work, but to my knowledge that is not and has not been the norm; much of academic publishing is built on trust of the authors, and even if some verification is done, it is not the sort of comprehensive verification that one does when one suspects that the data is badly wrong or even deliberately falsified.