A month ago, several colleagues and I wrote to BMJ about the problematic Norwegian study of CBT-plus-music-therapy for treatment of adolescents with chronic fatigue post-mononucleosis. After I sent a nudge two weeks later, I received a response from a research integrity coordinator, promising an update within two more weeks. That two-week period ended yesterday. Today, I sent BMJ this follow-up:
Dear Ms Ragavooloo--
It has been more than two weeks since our previous exchange, so I wanted to reach out again and check in on the status of the cognitive behavior therapy/music therapy study conducted by Norwegian investigators and published earlier this year by BMJ Paediatrics Open. I have cc'd several colleagues from Berkeley, Columbia, and University College London; they also signed the initial letter to BMJ and are also eager to see this matter properly resolved.
As we indicated in that letter more than a month ago, a proper resolution would mean withdrawing the study pending a second review by someone who promises to read the paper--unlike the original second reviewer. The new reviewer should be asked to compare the published work with the trial protocol and statistical analysis plan. Should the reviewer and BMJ confirm the identified methodological and ethical lapses outlined in our letter and on Virology Blog, the study should be retracted.
In reality, it should take an experienced reviewer no more than an hour to scrutinize the trial materials and assess how the paper itself diverged from what was laid out in the trial protocol and statistical analysis plan. It might take longer to determine why the investigators made the decisions they did and where BMJ's peer review process fell apart.
In the meantime, you might have seen this excellent article in the Norwegian press about the BMJ Paediatrics Open study. The article mentions the unusual downgrading from full-scale trial to feasibility study, which the investigators themselves chose not to mention in the published paper. The article also highlights the negative results for daily steps taken--the primary and only objectively measured outcome--and the key fact that the intervention group did even worse on that measure than those who did not receive the intervention.
The Norwegian article does not mention that post-exertional malaise, an outcome highlighted in the paper, was not included in either the protocol or statistical analysis plan. Nor does it mention that one of the two BMJ Paediatrics Open reviewers acknowledged not having read the paper. Overall, however, the article raises troubling issues that remain unanswered. (I have also cc'd the journalist, Ingrid Spilde, on this message.)
Before touching base with BMJ about this study, I wrote directly to the senior investigator and invited him to respond to my concerns, offering to post his full comment on Virology Blog. He did not respond.
I also wrote to the first reviewer. Unlike the second reviewer, she apparently read the paper and raised smart, pertinent questions about whether the research was designed to be a full-scale trial or a feasibility study. I wanted to find out whether she realized that the investigators, in response to her request for clarity about the goal of the research, falsely presented it as having been designed as a feasibility study. I did not hear back from her either.
My colleagues and I look forward to learning what BMJ plans to do about this problematic paper, which unfortunately has the potential to influence the treatment of children with serious illness--a highly vulnerable group. Given the paper's multiple methodological and ethical anomalies and the fact that BMJ published it after a peer review process that clearly violated recognized standards, we urge you to reach an acceptable decision quickly.
Best--David
David Tuller, DrPH
Senior Fellow in Public Health and Journalism
Center for Global Public Health
School of Public Health
University of California, Berkeley
