By David Tuller, DrPH
I have posted a batch of letters about the Lightning Process study that have been sent to Dr Fiona Godlee, editorial director of BMJ, here, here and here. I have been impressed with how direct these scientists and clinicians have been in expressing their dismay at BMJ’s failure to adhere to its own editorial standards. I get the feeling some of the writers have been inspired by the earlier messages to Dr Godlee.
Anyway, here are three more: from Steven Lubet, a law professor at Northwestern University in Chicago; Robert Garry, a microbiology professor at Tulane University in New Orleans; and Philip Stark, a statistics professor at UC Berkeley. Thanks to all of them for their support. That makes 14 letters so far.
Dear Dr Godlee: I am a legal ethicist, rather than a physician or scientist, so I will not weigh in on the various methodological and other problems that have been identified regarding “Clinical and cost-effectiveness of the Lightning Process.” I do want to comment, however, on one aspect of the “Correction” recently published in Archives of Disease in Childhood, which accepted “an assurance from the authors that the change in primary outcome was not influenced by (positive) findings in the feasibility phase.” Having spent many decades studying and evaluating conflicts of interest in both academic and professional settings, I find it extremely troubling that a respected scholarly journal would discount acknowledged research irregularities based upon such an obviously self-interested conclusion by the investigators themselves.
To put it bluntly, conflict of interest principles exist because it is impossible to rely upon affected individuals to provide objective assessments of their own decisions. In this instance, the “Lightning Process” authors have a quite evident stake in avoiding retraction of their article, which they only now admit “was not fully ICMJE compliant.” Thus, no matter how sincere they may be in defending their work, they should never have been the ones to determine whether the change in primary outcome had been influenced by their earlier positive findings.
In legal and judicial ethics, the applicable standard for disqualification is whether an individual’s “impartiality might reasonably be questioned.” Let me kindly suggest that the standards of BMJ Archives of Disease in Childhood ought to be no less rigorous.
In this instance, the authors of the “Lightning Process” article cannot possibly appear impartial regarding the validity of their own study, much less the question of whether their outcome swapping had been influenced, consciously or otherwise, by the pre-registration findings. Applying the most basic conflict of interest principles, the decision to issue a correction, rather than a retraction, should not have been based, even in part, on the authors’ after-the-fact assurances.
Williams Memorial Professor
Northwestern University Pritzker School of Law
Chicago IL 60611
Dear Dr. Godlee,
I am writing to urge you to reconsider the decision to republish the pediatric study of the Lightning Process conducted by investigators from the University of Bristol and published in Archives of Disease in Childhood.
My own perspective is as a virologist, with a long-standing interest in chronic illnesses such as chronic fatigue syndrome. Although I am not a physician, I have met many individuals with these conditions, and have at least one family member that is afflicted. Chronic fatigue syndrome is not a state of mental confusion that can be corrected by the pseudo-scientific incantations of the Lightning Process.
I am sure that you are aware that the International Committee of Medical Journal Editors has stated clearly* that, “The integrity of research may also be compromised by inappropriate methodology that could lead to retraction.”
If there was ever an instance when there was “inappropriate methodology,” this is it.
My recommendation is immediately publish an “Expression of Concern” regarding this multiply flawed study in accordance with COPE guidelines.
I then urge you to consult with other experts outside of the Bristol sphere of influence and follow their guidance.
I am confident that an independent group of experts will come the very same conclusion as the distinguished group copied here – this manuscript must be retracted.
Robert F. Garry, PhD
Robert F. Garry, PhD
Professor, Department of Microbiology and Immunology
Assistant Dean, Graduate Program in Biomedical Sciences
Program Manager, Viral Hemorrhagic Fever Consortium
Director, Tulane Center of Excellence, Global Viral Network
Chief Scientific Advisor and Co-Founder, Zalgen Labs, LLC
Tulane University School of Medicine
New Orleans, LA 70118
[From Philip Stark, Professor of statistics at UC Berkeley]
Dear Dr. Godlee–
I hope you reconsider your decision.
Leaving this paper in place is a disservice to science and to public health–although it does provide a lovely [(negative) example for teaching ethics in science and illustrates how skirting proper statistical practice may lead to erroneous conclusions.