Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 September 29; 335(7621): 629.
PMCID: PMC1995493
The Gillberg affair

Profound ethical issues were smoothed over

Aubrey Blumsohn, consultant

The issue at the heart of the Gillberg affair concerns the relation between raw data and the representation of those data.1 That issue is central to most recent scandals that have damaged the scientific basis of medicine and the trust of patients. It also concerns the availability of raw data to journals, fellow scientists, consumers, those who claim to police matters of integrity in science, and even to authors themselves. In short, it concerns the safety of the entire scientific enterprise.

Gornall's article skirts around every one of the principles while making conjectures about the personalities involved.1

The only fact of the affair that is relevant to a serious discussion of ethics is straightforward. The Gillberg team destroyed raw data, having faced an accusation of research misconduct pertaining to those data. They destroyed those data despite a court order that the data should be made available for scrutiny.

The argument about confidentiality is entirely spurious and could be made about practically every bit of clinical research that has ever been carried out. Is it really being suggested that no one (regulatory bodies, courts, bodies investigating research misconduct, trial participants themselves, coauthors, journal editors, research councils, or even authors of the science itself) should ever be allowed to scrutinise any aspect of research? This is not science, and the article that originated this discussion is not part of any form of legitimate scientific debate.

Journals such as the BMJ may request raw data from human studies when fraud is suspected, as do a variety of other bodies. There is nothing at all special about the Gillberg study that makes it an ethical outlier exempt from the usual norms of science. At least no such reason has been provided in anything I have read. In the well publicised case of Singh, which also involved the destruction of raw data (in his case termites provided the excuse), the failure to provide raw data provided grounds for suspicion of scientific misconduct—not congratulations.2 The apparent moral of the report by Gornall1 is that future researchers faced with questions about the plausibility of their findings should simply destroy their data.


Competing interests: None declared.


1. Gornall J. Hyperactivity in children: the Gillberg affair. BMJ 2007;335:370-3. (25 August.) [PMC free article] [PubMed]
2. White C. Suspected research fraud: difficulties of getting at the truth. BMJ 2005;331:281-8. [PMC free article] [PubMed]

Articles from The BMJ are provided here courtesy of BMJ Group