Following briefly reviews some specifics of the fraud allegations made against Christopher Gillberg.
Details of the fraud allegations are contained in the petitions of Elinder and Kärfve to the Ethics Committee. Those petitions, together with the responses and rejoinder from Gillberg and Rasmussen, comprise over 60 pages; the real length is actually greater, because full consideration requires also reading the cited works of Gillberg and colleagues. Herein, I present a summary of the main allegations. For more details, please see the petitions.
The primary concern here is whether or not Gillberg committed fraud. In addition, it is also of concern whether or not the central conclusion of the Gothenburg study is true. The central conclusion of the study is that if a child is diagnosed as having DAMP/ADHD (at age 6–7), then that child is much more likely than usual to experience severe problems later on in life, e.g. with drugs, alcohol, schooling, the law, and in relationships.
First allegation: examiners
There were roughly 100 participants in the Gothenburg study. Each participant was examined at age 7, when the study began, and in follow-up investigations at ages 10, 13, 16, and 22.
There is a simple requirement that helps to ensure that the examinations were unbiased: when a child was being examined, each examiner had to be unaware of what other examiners had concluded about that child. Such examinations are said to be “blind”. Gillberg and Rasmussen acknowledge that blindness is a requirement (e.g. in their response to Elinder's petition).
Such a large study, done blindly, implies that Gillberg must have found a fair number of examiners who were willing to do a substantial amount of unpaid work. Who were those examiners? Kärfve has repeatedly requested the names of the examiners—so as to confirm that the examinations were actually done, and done blindly. Gillberg and Rasmussen do not deny that this is a reasonable request. But, they have not supplied the names.
There is a bizarre twist to this. In the correspondence with the Ethics Committee, Gillberg and Rasmussen claim that they have supplied the names to Kärfve. All the correspondence between the various parties is in the public record. Indeed, in Sweden, copies of all such documents are stored in the National Archives. (A National Archives document number can be seen, for example, on the first page of each of the original Swedish-language versions of the correspondences with the Ethics Committee.) None of the correspondence with the Committee contains the names of the examiners. Moreover, Gillberg obviously could have addressed Kärfve's repeated requests by supplying the names.
Gillberg's claim to have supplied the names of the examiners is untruthful. Gillberg's refusal to supply the names seems inexplicable, if the examinations really were done as claimed.
Second allegation: subjects
There were 112 participants in the study initially, when the children were 6–7 years old: those were the children who were to be followed for the next 15 years. There was also a separate group of 29 children, who were suggested to have psychiatric problems by the original questionnaire (as answered by the children's teachers), but for whom examination by psychiatrists, etc., indicated no such problems.
Kärfve alleged that if a child in the group of 29 developed serious life problems during the next 15 years, then that child was claimed to be from the group of 112. The effect of this would be to make it seem that there were more children with life problems in the group of 112 than there actually were. In other words, the effect would be to make it seem that having DAMP/ADHD as a child was much more serious than it actually is. Because the grave seriousness of having DAMP/ADHD was the central conclusion of the Gothenburg study, this would mean that the conclusion was exaggerated, possibly even unfounded. I.e. Gillberg had come to his conclusion only by falsifying the data.
The petitions of both Kärfve and Elinder provide various arguments to substantiate this allegation. The main—but not sole—argument is that the drop-out rate is very low. Indeed, during the follow-up examinations (at ages 10, 13, 16, and 22), the number of people participating does not decrease much from the original 112: the numbers were 106, 96, 101, and 101. Kärfve and Elinder considered those number to be highly suspicious, particularly the increase (from 96 to 101). This was especially so because in principle the subjects received no treatment by participating in the study—so what would motivate them to go through all the trouble of detailed psychiatric examinations?
To me, these numbers appear to be somewhat suspicious. They are however, nowhere near proof of fraud. (Kärfve and Elinder did not claim them to be proof though—merely enough to justify a detailed investigation of the study's files.)
There is also a counter-argument to support Gillberg. In Sweden, each resident has an identification number; using this number, it is possible to find out where a person resides. That makes it feasible to track down previous study subjects (after they have moved, say) to participate in follow-up examinations. Additional support for Gillberg is summarized in the letter from the Ethics Committee rejecting Kärfve's petition. The letter tells that the Committee interviewed a nurse who assisted Gillberg, and from the interview concluded that “the researchers have put in a lot of effort to make the drop-out rate as low as possible”.
Seeking to learn more, I contacted Ove Lundgren, who was Chairman of the Ethics Committee at the time. As discussed elsewhere, Lundgren had spent four hours one day inspecting some of the files. Lundgren e-mailed me back (on 21 November 2006), stating the following, concerning that inspection.
I got a printed list of the participants of the 16 year long study. The list was said to include the participants of the study. When I looked into the first file some of the participants that apparently had been part of the study according [to] the records in the file were not included on the printed list.
Lundgren does not recall how many names were on the printed list. Moreover, he did not examine the file thoroughly, and so he did not know the total number of names that were in the file and not on the printed list. Even so, Lundgren's evidence seems to be very supportive of the fraud allegation. (It is unfortunate that Lundgren did not appreciate the significance of what he found at the time.)
Additional problem: subject awareness
A fundamental aspect of the Gothenburg study is that the study subjects did not know whether or not they had been diagnosed as having DAMP, i.e. they were blind about their diagnoses. In other words, DAMP, if it existed, was allowed to run its natural course. Gillberg and his colleagues have stated this many times in their writings.
In the correspondence with the Ethics Committee, however, Gillberg and Rasmussen state that the results of the various examinations “were presented to the parents”. (This is not something that was stated in any prior publications, to my knowledge.)
Gillberg and Rasmussen claim that telling the parents did not violate the “blindness” of the study, because the parents were not told whether or not the examination results meant that their child had DAMP. It is obvious, though, that if parents are told that their child has been assessed poorly on some psychiatric tests, then the parents will tend to be concerned. Hence the parents might well treat their child differently. Moreover, some parents would be expected to tell their children about the results. A child to whom it has been strongly hinted by a person in authority and with prestige (as the examiners were) that she or he has some serious psychiatric condition would be expected to behave differently. All this means that the condition of DAMP was not allowed to run its natural course.
Kärfve says that this could be due to lack of competence or of integrity. Whatever the case, the lack of subject blindness alone seems to make the central conclusion of the Gothenburg study questionable.
|•||The Gillberg affair|
|•||The Gillberg affair—Gothenburg study|