I am an Emeritus Professor of Neurology at the University of Edinburgh, and have been involved in clinical trials for over 30 years. My academic career focused on evaluating ways to prevent and reduce the impacts of strokes. I retired from clinical practice in 2016, but have continued to support clinical trials for a range of different illnesses by working on trial steering committees and trial Data Monitoring Committees (DMCs). Over the years since the mid 1980s I have served on over 50 different trial Data Monitoring and Steering Committees.
How did you become the Chair of the RECOVERY Trial’s DMC?
I have worked with a number of different researchers at Oxford Population Health who were running trials, including Martin Landray, the co-Chief Principal Investigator for RECOVERY. As soon as he and Peter Horby began planning RECOVERY, Martin asked if I would be prepared to chair the study’s DMC. (I guess he knew me well enough to trust me with this rather responsible job.) I was honoured to be asked and agreed immediately.
Why do trials need a DMC and who are its members?
During a clinical trial, it is very important that the investigators do not know what the data show until the study is complete. The purpose of a DMC is therefore to monitor, in strict confidence, the trial data to ensure the safety of the participants and the integrity of the study results.
The DMC meets regularly to review the data, and at each meeting decides whether to recommend to the trial steering committee to a) continue as planned; b) continue with modifications; or c) stop recruiting new participants (either temporarily or permanently).
A DMC always includes experts both in the medical condition the trial is investigating (COVID-19 in the RECOVERY trial), and statisticians with experience in analysing clinical trial data. The DMC members should have no direct personal involvement in the study, or vested interest in the results. For smaller trials, the DMC may only have three or four members, so RECOVERY is somewhat unusual in having a DMC with nine members. The size of RECOVERY’s DMC reflects the wide range of health issues the study touches on (such as infectious disease and inflammatory responses), and importantly, the diversity of participants (including pregnant women and children).
What happens during DMC meetings?
For each treatment investigated by RECOVERY, the chief investigators calculate the target number of patients needed to answer the study question. When this number is reached, the study stops and the data are revealed to the research team. Once recruitment starts, the DMC needs to review the data regularly, so before each DMC meeting the trial’s statisticians send a report with the most up-to-date data for us to review. We then decide whether the evidence indicates that the study should be stopped before the planned number of participants have been recruited. This can happen if the treatment is very clearly beneficial, or if there are obvious concerns that the treatment is harmful.
In the RECOVERY trial, as the number of patients enrolled on the dexamethasone study increased, it began to look like a very promising treatment, but the DMC allowed recruitment to continue to its planned size to ensure the results provided proof beyond reasonable doubt of its benefits.
When it became clear that hydroxychloroquine was not an effective COVID-19 treatment and that it might even be harmful, we recommended that this arm of the study was stopped.
What happens next?
Because the data have to be kept in strictest confidence, after each meeting the DMC issues a short, simple report, which usually just recommends continuing as planned, but sometimes will recommend stopping recruitment or modifying the study in some way. It’s important that the report never gives away hints that the results are either promising or disappointing, before we can be quite certain that the signal we see represents the true effect of a treatment.
How has RECOVERY been different to other clinical trials you have been involved with?
I have never worked on a trial that has moved so quickly and recruited participants as rapidly as RECOVERY; at peak recruitment, over 500 hospital patients were joining the study each day.
RECOVERY’s DMC has met much more frequently than would be usual, sometimes as often as every two weeks (many DMCs for large trials would meet only every three-six months). RECOVERY’s adaptive design means that treatments have entered and left the study throughout the past two years. To date, twelve treatments have been assessed in RECOVERY: nine have now completed assessment and three are still under way.
RECOVERY has also been exceptional for the urgency and potential impact of its work. The DMC were all too aware that the results were globally relevant and could rapidly alter treatment worldwide. So we felt a great responsibility, with a need to balance making timely decisions with making sure we took enough time to make sure we got the decision right. With the pandemic affecting so many people in so many countries, we knew that if a treatment had even a small benefit against COVID-19, that benefit could translate into a large number of lives saved worldwide.
What do you enjoy about working on DMCs?
It is very rewarding, because you are working with interesting people who are real experts in their fields and who are all committed to ensuring the trial answers the research question reliably. There can be some tricky decisions to be made, and the discussions and interactions are a real highlight of this work. It is also a privilege to be the only people who can see the data accumulating, when even the principal investigators are not allowed to.
Are there any challenges?
It is hugely important that we don’t jump to conclusions too soon, which puts a lot of responsibility on our shoulders. For this reason, every DMC has a charter which sets out the terms of reference to guide us in deciding whether a treatment is effective or not. In the case of hydroxychloroquine, many people (including Donald Trump and other heads of state outside the USA) strongly believed that it worked, in part based on prior studies that were too small or not randomised. We therefore knew the evidence from RECOVERY had to be crystal-clear about its effects and whether it was truly helpful or harmful. Similarly, there were signs early on that dexamethasone was benefitting COVID-19 patients, but we couldn’t recommend stopping the study until we had ‘proof beyond reasonable doubt’.
How do you hope the RECOVERY Trial will impact future clinical research?
RECOVERY has demonstrated that, even in a pandemic, we need to test treatments through clinical trials rather than just giving them to patients without any evidence in the hope of some benefit. For instance, many physicians around the world were so convinced of the benefits of convalescent plasma that they didn’t see the need to do a proper, randomised controlled trial and – in the USA alone – half a million patients with COVID-19 were treated with convalescent plasma before RECOVERY and other trials proved it was ineffective. We should address uncertainty openly and be honest about what the evidence says. Where there is any doubt, patients should receive promising new treatments as part of a randomised trial, so that we can more quickly answer the question of whether they work or not.
We must expect further pandemics; RECOVERY emphasises the need to have a plan ready, so that when the next pandemic does arrive, we can act fast.
RECOVERY has been a success because many elements of the plan were already in place before the first wave of infections hit the UK and any regulatory barriers to starting the trial were overcome with top-level political and scientific support. Combined with the huge support of the NHS, this enabled the study to get off to a flying start as the first COVID-19 wave began to arrive in the UK.
I also hope that some of the very many NHS professionals who have worked on the RECOVERY Trial may be encouraged by the experience to consider a career in researching treatments of benefit to patients, including through clinical trials and that some may even wish to work on the DMCs of future trials!
The members of the RECOVERY Trial DMC are:
Screenshot from an online RECOVERY Trial DMC meeting
- Professor Peter Sandercock (Chair), University of Edinburgh, Edinburgh, UK
- Professor Janet Darbyshire, University College London, UK
- Professor Robert Fowler, Sunnybrook Hospital, Toronto, Canada
- Professor David Lalloo, Liverpool School of Tropical Medicine, Liverpool UK
- Professor David DeMets, University of Wisconsin - Madison, USA
- Professor Mohammed Munavvar, Lancashire Teaching Hospitals NHS FT, Preston, UK
- Professor Adilia Warris, University of Exeter, UK
- Dr Janet Wittes, WCG Statistics Collaborative, Washington DC, USA.
With DMC statisticians (non-voting):
- Professor Jonathan Emberson, Nuffield Department of Population Health, Oxford, UK
- Professor Natalie Staplin, Nuffield Department of Population Health, Oxford, UK.