By David Tuller, DrPH
A few days ago, I wrote a post about yet another atrocious paper from Professor Trudie Chalderâ€”this one called â€œChronic fatigue syndrome and occupational status: a retrospective longitudinal study.â€ Professor Chalder and her colleagues seem constitutionally incapable of writing anything that isnâ€™t marred by massive flaws. In this case, as I noted the other day, besides the fact that she and her colleagues misrepresented their findings by messing up their denominators in expressing percentages, they omitted any reference to the null findings on employment from the PACE trialâ€”an egregious lapse.
They also seemed confused about whether they diagnosed patients using the 2007 guidelines from the National Institute for Health and Care Excellence for CFS/ME, or using the 1991 Oxford CFS criteria. They mentioned NICE in the text but the cited reference was the 1991 paper. The differences between the case definitions are significant, so this confusion is jarring.
Given these failures, the paper itself is an unreliable and unacceptable report on the issue of employment status. The omission of the highly salient PACE data also raises serious questions about the integrity of the research team.
But the paper is full of other methodological problems. Overall, the results indicate that the interventions have little if any impact. According to the paper, of the 316 who provided both baseline and follow-up data, 88% experienced no change in employment status.
However, these figures are hard to interpret, since the authors lumped full-time employment and part-time employment together in the â€œemployedâ€ category. For all we know, many of those said to be working went from full-time to only part-time employmentâ€”but such a trend, if it exists, would be masked by the way the data are presented.
As Mark Vink pointed out on twitter, many of the outcomesâ€”including fatigue, fear avoidance, catastrophizing, and other domains–yielded null results. Others, like physical function, demonstrated only insignificant changes. The paper itself does not mention any of these poor findings in the text but only includes them in a table. Keith Geraghty has noted other issues with the paper and analyses. Certainly the final, post-treatment scores on the Chalder Fatigue Scale and the SF-36 questionnaire for physical function represent continuing severe disability.
Here are some other issues that need to be addressed.
The study seems to have had a major loss-to-follow-upâ€”of the 508 participants at the start, only 62%–or 316â€”provided follow-up data. That is a huge drop-off, and the authors tell us almost nothing about these people other than that they are possibly in worse health than those who provided follow-up data, as tends to be the case with drop-outs.
That in turn means that the baseline averages presented for all 508 participants likely represent worse health than the baseline averages of just the 316 who provided follow-up data. For unexplained reasons, the authors did not provide the baseline data for the 316 alone. That would have allowed for an assessment of whether their health actually got worse during the time period, according to various variables, or simply didnâ€™t get better.
The decision not to include these baseline data for the 316 participants who provided follow-up data is rather unusual. It suggests that the authors might have known that the pre- and post-treatment comparison of outcomes for the 316 participants who provided follow-up data would not look very attractive.
The authors make much of the â€œoptimismâ€ they see in the fact that 9 % of the 316 participants were not working at baseline but had returned to work by follow-up. But this number is offset by the 6% who stopped working between baseline and follow-up. And of course, again, we have no idea if these are full-time or part-time positions, so assessing the changes is challenging. Besides that, the authors provide no tests of significance for these resultsâ€”another bad sign.
The authors continue to suggest that â€œunhelpful beliefsâ€â€”that is, patientsâ€™ notions that they have an ongoing disease that is exacerbated with excess activity–play a role in perpetuating the syndrome. They canâ€™t stop flogging this dead horse, even after decades of flogging it.
Hereâ€™s what they write: â€œUnhelpful beliefs such as fear of activity and exercise and concerns about causing damage, combined with all or nothing behaviour and behavioural avoidance, were associated with not working and are specifically targeted in CBT and, to some extent, GET.â€ They suggest that the interventions should include more focus on employment issues.
Given that the factors presumed to be impeding recovery are already the focus of the interventions that Professor Chalder has advocated for decades, the evidence from this study is clear: These treatments are essentially worthless. Overall, they largely appear to fail to improve health and work outcomes. It is astonishing that Professor Chalder and her colleagues do not seem to realize this.
Professor Chalder and her two PACE besties, Professor Michael Sharpe and Professor Peter White, have recently published another paper purporting to make the case for CBT and GET as effective â€œevidence-basedâ€ treatments. That article is also a piece of crap. Too bad they did not incorporate these disastrous findings into that analysis. Instead, they continue to spout drivel and nonsense. When will this end?