Peter White et al: Releasing patient data from the PACE trial for chronic fatigue syndrome

The PACE trial was the largest clinical trial to date into Chronic Fatigue Syndrome (CFS), also sometimes referred to as Myalgic Encephalomyelitis (ME). This randomised controlled trial involved 641 UK patients suffering from CFS and compared the effectiveness of four treatments. It found that when added to specialist medical care (SMC), cognitive behaviour therapy, and graded exercise therapy were more effective than both adaptive pacing therapy, and SMC alone.

We have received many requests over the last five years to release the patient data collected in the PACE trial, and former editor of The BMJ Richard Smith suggests we have made “a serious mistake” in not always releasing data. The data requests often cite the importance of transparency, giving other scientists the opportunity to investigate our data, and changing customs in science where data sharing is increasingly encouraged.

These are important issues and we are entirely supportive of the principles behind the sharing of scientific data. Not only that, but we have already shared data from the trial many times, when there have been sufficient safeguards in place before doing so.

Although discussed in a Nature article by Lewandowsky and Bishop earlier this year, the central issues of patient consent and the need for pre-specified analysis plans are rarely mentioned in debates about open data, and this is especially important to consider in the context of clinical trials.

Patient consent

The reason we have been unable to release data to some of those who ask for it is that trial participants did not give consent to the public release of their data when they entered the trial.

The two data requests mentioned by Richard Smith were from a member of the public, using the Freedom of Information Act, and from a researcher who has said he would make the data publicly available once received. We were concerned for the rights and welfare of trial participants and ethically bound to act in the best interest of our patients. We have also followed both our University guidance and the guidance of the UK Medical Research Council, who funded the PACE trial, in regard to data sharing.
Few people in medical research, to our knowledge, include an agreement as part of the consent that a patient’s individual data, even “anonymised data,” can be released to anyone who requests it. How many patients would take part in medical research if they knew that their data could be made publicly available afterwards? We fear that having such an agreement as part of the consent to participate in a clinical trial would put off a significant number of potential participants, particularly if the trial was of treatments of a stigmatised illness, such as CFS.

In our opinion, this issue is a threat to medical research into such conditions, which has not received the attention it deserves, although the recent section 22A amendment to the FOI Act provides some protection for current on-going research.

As moves are made into an era of greater openness in scientific research, the issue of how we balance the need for data sharing, patient confidentiality, and ongoing recruitment into clinical trials, will require careful consideration.

Sharing data with other research groups

The regulatory framework for clinical trials in the UK, which we have followed, seeks to ensure that individual participant data is only shared with other research groups who are able to abide by strict confidentiality obligations and fulfil other requirements for responsible data sharing. This framework is designed to ensure that, when they give their consent to take part in a clinical trial, participants can continue to have confidence that their information will only be used in a way that is consistent with their consent.

We therefore support the use of PACE trial data for additional research. As such, we have shared the relevant data with researchers many times, including a Cochrane Review group. This has been with an appropriate confidentiality agreement and a pre-specified statistical plan for analysis of the data.
A pre-specified analysis plan is important for ensuring that subsequent analyses and interpretations of the data are unbiased, that there is no duplication of work, and that analysis plans remain unaltered following the initial interpretation of results.

The future of clinical trials

We are conscious of the active debate around CFS and the PACE trial. However we should also respect the terms to which trial participants signed up and keep their data secure.

The issues we raise here concern not only the PACE trial but also many other studies, as concerns about patient consent apply to all medical research programmes involving people, including studies into sensitive topics and stigmatised illnesses.

If efforts are made to ensure that clinical trial data are more openly available in the future, then careful thought and discussion will be needed within the scientific community on how we can achieve this while retaining the full confidence of our trial participants.

Information for Authors