Clinical trials – questions posed at a meeting this week

June 27, 2007


NB: this report also contains information on a new MRC meeting regarding ME/CFS research.

I attended a meeting at the Royal Society of Medicine on Monday 25 June which had been organised by The Lancet and The James Lind Alliance to discuss the very important issue of making research more relevant to the needs of both clinicians and patients.


During the morning panel discussion session I raised the issue of the flawed and deeply unpopular research strategy that had been produced by the Medical Research Council – a document that had no patient involvement in its production.  In response it was stated that the MRC had held a further meeting earlier in June to look again at research strategy in relation to ME/CFS.  At this point I am unable to put any more ‘meat on the bone' but the MEA will endeavour to find out more and report back in due course.


Overall, the meeting was interesting in that it brought together a number of clinicians and people involved in both funding and carrying out research who clearly believed that there were major flaws in the way that research is being initiated, performed, analysed and subsequently reported (or not reported) in the medical journals.


So it was pleasing to find that there was widespread agreement about the need to:


  • Stop repeating clinical trials that do little more than deal with questions that have already been answered.
  • Take much more account of subgrouping when designing clinical trials and analysing the results because, as several speakers pointed out, average results from clinical trials do not apply to all patients in the trial.  In fact, trial results can sometimes be driven by a relatively small group of influential (and typically high risk) participants. 
  • Make more use of risk models and scores to predict the likely effects of treatment in individual patients.
  • Make all the relevant raw research data available  possibly through a central database where it can be stored. In other words the sometimes highly selective results presented in research papers are often only the ‘tip of the iceberg', and if all the data was made available it is possible that other interpretations and conclusions might be made.   The current method of presentation – often brief statistical summaries of selected analyses – lacks transparency and precludes reanalysis.  These analyses may not even report the intentions that were stated in the protocol. Comprehensive trial data should therefore be there to share with the medical and scientific community.
  • Make sure that negative results are also published – it is scientifically unethical to fail to place this information in the public domain.
  • Remember that systematic reviews tell us about the average patient but nobody is average! 
  • Carefully report harms as well as benefits so that doctors can offer informed risk:benefit assessment to the patient.
  • Accept that miscarriage of treatment as a result of badly designed or analysed clinical trials is a miscarriage of justice as far as the patient is concerned.

All of these points can obviously be applied to research that has been, or is being, carried out in ME/CFS and I hope that they will!


Dr Charles Shepherd

The ME Association

27 June 2007

Shopping Cart