As I described in chapter 2, a similar complaint was made against me. A group of readers and authors complained that the BMJ presented a one-sided view of the condition known as chronic fatigue syndrome or myalgic encephalomyelitis. The BMJ, they alleged, published only material that supported the idea that the condition is psychological, used the same reviewers (most of whom are psychiatrists), and refused to publish studies that show that the condition is a physical condition. The complainants conducted an analysis of what the BMJ had published on the subject, and their anger was increased when we declined to publish the analysis. (One objection to the analysis was that it included information only on papers we published and not those submitted to us — so how could it show bias?)
My response was that the BMJ did not have a position on the nature of chronic fatigue syndrome or myalgic encephalomyelitis. We selected the best research studies submitted to the journal and published them. Often studies contradicted previous studies that we had published. Bias seemed to arise because we were more likely to accept randomized trials of treatments (including a psychological treatment called ‘cognitive behavioural therapy’) than we were studies reporting a small series of patients who had evidence of a previous infection. We thought of this as a bias towards rigorous science, a bias that it was right for us to have.
In addition, we would ask a variety of people to write commentaries for us, selecting those who know about the condition, argued well using evidence and wrote clearly. It did so happen that most of the experts thought that the condition does have a substantial psychological component and that it is highly unlikely to be entirely a physical condition. Indeed, most of them believe that most conditions have a physical and psychological component. They are also distressed by the implication of the complainants that a physical condition is somehow more ‘real’ than a psychological condition. This seems to perpetuate the stigma against those with mental health problems.
When it comes to letters anybody can comment, and the BMJ carries many letters from people who disagree with most of what the BMJ publishes. When I left the BMJ received well over one hundred rapid responses (electronic letters to the editor) each week and we were able to publish fewer than 10% in the paper journal. This arrangement provided a rich test bed for studying editorial bias: what were the characteristics of the fewer than 10% that we selected? We’d like to think that they were more interesting, sound and readable, but no doubt other forms of bias were at work.
This is a question of where editorial judgement ends and misconduct begins. Editors are expected to discriminate, but they should discriminate on grounds of evidence, importance, relevance, quality and clarity rather than on personal foibles. But it is also widely believed to be the job of the editor to give a publication a ‘personality’ — and that’s likely to be related to his or her personality. So some personal selection seems desirable.
The complaints against me alleging misconduct in relation to selecting material on chronic fatigue syndrome — to the British Medical Association (BMA), the owners of the BMJ, and the Press Complaints Commission — were dismissed, but the complainants saw this as an establishment cover up. Perhaps it was.