August 16, 2016, 8:51 a.m EST: I’ve only just received a scan of the 48 page judgment. Because the scan is not searchable, I dictated the closing Majority decision.
[Update 8:00 pm: A searchable full PDF of the judgment is now available here.]
But to get to the punch line:
The Tribunal, by a majority, upholds the decision notice dated 27 October 2015 and dismisses the appeal.
After some brief introductory comments from the blog of Attorney Valerie Elliott Smith, I present some details from the majority decision. I will continue to review the whole document, but I think the decision represents a smashing defeat for the PACE investigators, in which their detailed arguments were turned back on them. It is also defeat for the many foes of routine data sharing, including those hastily marshaled by the Science Media Centre.
In terms of my own quest for the PLOS One PACE data, recall that I never framed my request as falling under the Freedom of Information Act rubric. I simply asked that the data be provided to me as the PACE investigators had promised. There has been no consultation with me nor other stakeholders in the many months of the Senior Management of PLOS One’s negotiation with the PACE investigators. Maybe the Tribunal decision will now break the stalemate and encourage some transparency on the part of the senior management of PLOS One. However, I’m continuing with plans to go to Cambridge in November or early December and march on the PLOS headquarters and demand an open dialogue. I welcome you to join me and we will party outside the headquarters. Or maybe not – I hope, but I cannot expect the matter to be resolved by then.
Finally, I find particularly delicious a section of the majority decision that I reproduced below:
In any event there is a strong public interest in releasing the data given the continued academic and for so long after the research was published and seeming reluctance to Queen Mary University to engage with other academics they thought were seeking to challenge the findings (evidence of Professor Chalder).
Trudie Chalder used the scarce time allocated to her appearance before the tribunal specifically to continue her attack on my character. She expressed alarm over the threat my criticism posed to the reputation of the PACE investigators. Trudie, vexatious, my ass. You have been rude and unprofessional. You have consistently refused to debate, but you can look forward to a future detailed critique of your miserable, p-hacked mediational analysis.
Attorney Valerie Elliott Smith immediately blogged with some background, to which she promises to add more later.
The First-Tier Tribunal judgment in this case has just been published. QMUL’s appeal has been roundly dismissed and therefore the Tribunal has decided that the requested data from the PACE trial should be released.
I have just skimmed the 48 pages of the judgment and so have only taken in a small amount so far. However, it appears that this is a defining moment for the international ME community and the PACE Trial. Alem Matthees (the original requestor of the data) has done an extraordinary job.
However, it is important to remember that, in theory, QMUL could still seek leave to appeal against this judgment to the Upper Tribunal so it will be a bit longer before we can be absolutely certain that this judgment will stand.
Background note for new readers
In March 2014, Mr Matthees sought some of the data from the controversial PACE trial, using the process set out in the English Freedom of Information Act (FOIA). This information is held by relevant public authority, Queen Mary University of London (QMUL). QMUL refused to disclose the data.
In due course, Mr Matthees complained to the Information Commissioner (IC) who, in October 2015, ordered that the information be disclosed. QMUL appealed against the IC’s decision; that appeal was heard by the First-Tier Tribunal on 20-22 April 2016 in central London. QMUL and the IC were legally represented and QMUL called witnesses to give evidence. Mr Matthees had been joined as a party to the proceedings. He was not legally represented and did not attend the hearing but made written submissions. Judgment is awaited.
[Note: the PACE trial, which was published in 2011, relates to certain treatments for the condition known as “chronic fatigue syndrome” (CFS). CFS is often conflated (confusingly) with myalgic encephalomyelitis (ME) and referred to as CFS/ME or ME/CFS, to the detriment of genuine ME patients. This is the situation in many countries and has been for decades; it is the cause of significant confusion and distress to many patients worldwide.
The results of the PACE trial appear to promote psychosocial treatments which many patients find either ineffective or actively harmful. As a result, some patients have been using FOIA to try to obtain the trial data in order to understand how these results were achieved. However, most requests have been denied and, five years on, most of the data is still unavailable.]
The Tribunal decision [Please alert me to typos in this section that was hastily entered with voice recognition software.]
The Tribunal, by a majority, upholds the decision notice dated 27 October 2015 and dismisses the appeal.
The majority decision
The Commissioner accepts that the question turns on whether anonymization is possible, and he argues that in this instance identification is an extremely remote possibility. Professor Anderson accepted that the information alone cannot identify participants, and his hypothesis that identification is possible to combine that information with NHS data (involving an NHS employee both having breachedtheir professional, legal and ethical obligations and also having the skill and inclination to do this) is implausible.
In short we accept and adopt the Commission’s wider submissions and reasoning as set out in his Skeleton Arguments and Written Closing on this issue. In all circumstances and on the evidence before us we are satisfied that the risk of a dedication has been anonymized to the extent that the risk of identification is remote. Incomes this conclusion we’ve also taken into account:
- The nature of the information, which did not contain any fixed or direct identifiers;
- The evidence of Dr. Rawle that the anonymization methodology followed the guidelines at the time and would still comply with current guidelines although they were said to be under review for the future;
- The evidence of Dr. Rawle that none of the identifiers were contained in the disputed information, (the anthropometry measurement issues were cleared up);
- The evidence of Professor Anderson that third parties could not identify participants from the information alone and that, when pressed, he said the chance of an “activist” being able to discover information that would lead to individual identification was remote. It was clear that his assessment of activist behavior was in our view, grossly exaggerated and the only actual evidence was there an individual at seminar had heckled Professor Chalder. The identity of those questioning the research, who signed an open letter or supported it, was impressive. While we accept that Professor Anderson was an expert witness, he was not a Tribunal appointed independent witness but appointed by the Appellant and clearly, in our view, had some self-interest, exaggerating his evidence and did not seem to us to be entirely impartial. What we got from him was a considerable amount of supposition and speculation with no actual evidence to support his assertions account the Respondents’ arguments;
- That even on Professor Anderson’s evidence for identification to take place there would have to be a breach of medical ethics and law and there was actually no evidence to quantify the risk of this occurring in the circumstances and on the facts before us in the appeal. In fact there was no tangible evidence of an example where such steps had led to a identification of the individual in any circumstances.
- That even on Professor Anderson’s it was only the walking scores that were likely to lead to identification (if all of his other suppositions and speculations came about);
- That the Fine research had been released, albeit t by accident, and there was no evidence that although it contains similar data, that there had been any individual identification or problems arising from it.
- We do not accept the commercial interest arguments of the College. There was very little evidence of the withdrawal of consent and where it happened it was only directly related to the issues before us. Funding had been obtained for new trial in the knowledge that the tribunal may not allow this Appeal. New confidentiality guarantees, or perhaps more explicit ones, could be given to new participants.
- In any event there is a strong public interest in releasing the data given the continued academic interest so long after the research was published and seeming reluctance to Queen Mary University to engage with other academics they thought were seeking to challenge the findings (evidence of Professor Chalder).
- There is insufficient evidence before to persuade us that the disclosure of the disputed information would cause sufficient prejudice to QMUL’s research programmes, reputation and funding streams.
- Professor Anderson when cross examined as to whether or not the patient identities (HESID) that would disclose to anyone with access were encrypted. He was unaware but has since checked and is now confirmed through submissions on behalf of the Appellant that this does not affect the wider point he made about care data disclosures, that they would potentially allow different health events to be linked where it could be established that they relate to the same patient identifier. This, we would have preferred to explore further in evidence but it seems to us that encryption makes the chance of identification even more remote in any event and strengthens our view that the speculative assertions of the occurrence of possible events actually taking place in a way that could lead to identification individuals, by Professor Anderson are indeed remote.
- We do not accept the speculation that the chanmce of a determined person with specialist skills could make the link, while less than probable, is more than remote. There’s no tangible evidence before us to persuade us that is less than remote. Professor Anderson accepted it would be extremely difficult to identify individuals even from the collective information, which as the Commissioner submits “given his approach to this would indicate the near impossibility of reidentification”. We are not persuade the risk of identification is more than remote.
- Generally, regarding the Commissioner’s discretion, we heard nothing, and are not persuaded on the evidence before us, that would lead us to question that the Commissioner had not apply himself correctly , that his decision was not properly arrived at or should be set aside.
[The majority Decision – (Mister Watson dissenting);
- Each role in the spreadsheet is unique and refers to one person in the trial.
- The information necessary to link the data to an individual is available to a large number of people due to the way in which security has been implemented and the NHS and the quantity and nature of the information is now available on social media.
- I believe Professor Anderson is correct when he gave evidence that the chance they determined person with specialist skills could make the link, while less than probable, is more than remote.
- For this reason the information contained in the spreadsheet is personal data and should not be disclosed.
Finally, at paragraph 5 c) above: “Would disclosure cause sufficient prejudice to QMUL’s research programmes, reputation and funding streams to refuse disclosure?”
We unanimously accept and adopt the Commissioner’s wider submissions and reasoning has set out in his Skeleton Arguments and Written Closing on this issue. In all the circumstances and on the evidence before us we are satisfied that the perceived risk of prejudice as submitted by the Appellant’s has not been substantiated or demonstrated evidence before us. Such minimum risk as has been expressed would not now view outweigh the public interest in disclosure of the disputed information as defined in the specific request in this appeal.
- The Tribunal wish to thank all parties for the helpful manner in which they have presented their arguments and submissions. We’ve been provided with an extraordinary amount of anciliary and background information on and about the important subject matter under consideration and have considered all of it. They can be no doubt about the Public Interest in the subject matter which is evident through the course of this appeal, and beyond, and we are grateful for the assistance has been given to us in this regard.
We have considered all of the above arguments, submissions and evidence together with a significant volume of supporting evidence in legal precedents and for the reasons given above we refuse the appeal by majority decision for the above reasons, and the Commissioner’s DN stands.
Brian QC 11th August 2016
Promulgated 12th August 2016