Administrative staff at PLOS asked me to inform them if I had any difficulties obtaining the data that I had formally requested from the authors of the PLOS article. Here is my reply –
The deadline that the authors’ university has set for responding to my request has passed. The university has not responded to my further query as to when the data will be forthcoming.
The author group has a consistent history of refusing requests to provide their data to other investigators. Their PACE clinical trial that served as the basis for the publication in PLOS has become quite controversial. Critics including myself have raised considerable doubts about their basic design, their changing of the trial’s endpoints after the trial began, and their analysis and interpretation of the data. There have also been complaints that analyses in published papers and lack sufficient detail and transparency to allow independent evaluation of what was done and how the authors interpreted their results.
It is a matter of public record that the authors have denied at least 14 requests for sharing of their data, notably for data are used in articles appearing in Lancet, Psychological Medicine, and Lancet Psychiatry. In each case, the investigators have deemed requests “vexatious” and argued release of data would undermine their personal credibility and that of the study. Richard Horton, editor of The Lancet has gone on record with an estimate that the investigators have spent over 750,000 pounds in fighting requests to share their data.
In a 30 page decision on October 27, 2015, the UK’s Information Commissioner’s Office ordered Queen Mary University of London (QMUL) to release anonymized from the PACE trial data to an unnamed complainant. The PACE investigators have appealed that decision and so any release is delayed. The data set cited in that IOC order overlaps with what I am requesting,, but omits some specific data needed to redo the analyses presented in the PLOS article with appropriate sensitivity analyses.
A cross-sectional study in PLOS One found an unwillingness to share data in response to formal requests is associated with weaker evidence against the null hypothesis of no effect and a higher prevalence of apparent errors in the reporting of statistical results. It is my opinion, after carefully considering numerous publications from this trial, that the PACE investigators seek to hide what would be revealed by release of the data used in the PLOS article. But the matter should not be a one of my versus their opinion , the scientific community needs to see the data. This is particularly so when such clinical and public health policy implications are being attached to the study and the PLOS article in particular.
If the PACE investigators behave consistent with their well-established pattern, they will refuse to release to me the data used in the PLOS article. However, what is different than in past requests is that mine is for data published in a journal internationally known for its commitment to transparency and data sharing. If the PACE investigators refuse, they will be testing the commitment of PLOS to its policies and the scientific community will be watching.
I am unaware of a PLOS journal’s data sharing policy similarly having been tested by investigators refusal to make their data available. In the eventuality that I do not receive their data, I believe appropriate sanctions should be available for immediate application. Otherwise the PACE investigators are making a mockery of data sharing at PLOS as they already have done with the clearly articulated policies in the UK.
I will keep you posted.
Update on my update: When I awoke this morning in Philadelphia with its five hours difference from the UK, I had received no response from King’s College, which is handling my request for the data. I then composedand sent the above email to PLOS. Apparently it crossed paths with the following email from King’s College:
Dear Professor Coyne,
Thank you for your email. Your response will be sent out next week. The 20 day time frame refers to working days and we have calculated that we should provide our response no later than 11th December.
Information Compliance Coordinator
Information Management and Compliance
Governance and Legal Services
King’s College London
PACE investigators tear down the wall of secrecy surrounding your data.
Note that the previous email from King’s College mentioned only 20 days, not 20 working days. And that the new email only promises a response, not the data. Stay tuned.
Note: Although I am one of thousands of academic editors for PLOS One, I have sole responsibility for my blog posts, which cannot be construed as reflecting the opinions of any institutions with which I am affiliated, including PLOS. In requesting the data for an article that appeared in PLOS, I am exercising the same rights to anyone in the world has with respect to this open access journal committed to transparency and data sharing.