Patients and statisticians have used the recently released data from the PACE trial to show that cognitive behavioural therapy and graded exercise therapy did not help patients in the study to recover.
Alem Matthees, an Australian patient who obtained the data after a two-year battle over his Freedom of Information request, applied the study authors’ own pre-planned analyses that they had abandoned after the trial had finished.
Working with independent statisticians and other patients with mathematical expertise, he showed that only 7% of the cognitive behavioural therapy group, 4% of the graded exercise group, and 3% of a no-therapy comparison group recovered. Differences between the groups were within chance variation, indicating no effect of the therapies.
This contrasts starkly with the recovery rate of 22% in the therapy groups that was published by the study authors in the journal Psychological Medicine in 2013. In that paper, patients were classed as having recovered from their disability even if they became more disabled during the trial. The Lancet published a similar analysis in an earlier paper. Both journals have for years refused to correct the analyses.
Statistics professors Philip B. Stark of the University of California, Berkeley, and Bruce Levin of Columbia University co-authored the reanalysis report. Professor Levin commented that respect for The Lancet and Psychological Medicine had been “diminished worldwide” by their defence of the trial and that it would be appropriate to retract the flawed analyses.
Professor Vincent Racaniello has published the report on his Virology Blog, describing it as “an analysis that the authors never wanted you to see”. He said, “The results should put to rest once and for all any question about whether the PACE trial’s enormous mid-trial changes in assessment methods allowed the investigators to report better results than they otherwise would have had.” He called Alem Matthees’s persistence in obtaining the underlying data “heroic”.
Professor Simon Wessely summarised his overall reaction as, “OK folks, nothing to see here, move along please.”
Patient, journalist and mathematician Julie Rehmeyer has written an article drawing attention to the new analyses. She said, “Problem is, the study was bad science. And we’re now finding out exactly how bad.” She said that critics of the trial had been painted as “unhinged crusaders” but that the new analyses showed that the PACE authors’ recovery claims had gone “up in smoke”.
She reported that the authors, and the editors of The Lancet and Psychological Medicine had all declined to comment for her article, but that Professor Simon Wessely, of the UK Royal College of Psychiatrists, had defended the trial in an email exchange with her. She said that he had refused to comment on the new recovery analyses but that he summarised his overall reaction as, “OK folks, nothing to see here, move along please.”
Carly Maryhew, a patient and co-author of the analysis report, said, “Given that our re-analysis of the data has shown a shocking difference in the claimed recovery rates, Wessely’s flippant dismissal is simply ludicrous. It has been proven now that the changes to the protocol had a huge impact upon the outcomes, and it’s demonstrating to the entire academic community that science requires constant scrutiny to keep it honest.”