Chronic Fatigue Syndrome and Occupational Status: A Retrospective Longitudinal Study, 2021, Chalder et al

Would be interesting to have details of the hours worked by the different groups.

Were those taking up work perhaps only picking up very part-time jobs?
Also we have no detail about whether those in work and staying in work, actually reduced their hours in order to stay employed.

So the “not working” group could add only a few hours work & be counted as starting work,

but others could have a substantial reduction of their hours, and remain in the “still working” group.

It’s not a balanced measure.

 

Following on to above, how is being in work defined?

I “had a job” for a full year after being unable to “work”.
So at what point does taking sick-leave become stopping work? Presumably there is a lag-time before an individual is counted as not working?

In the other direction, starting work presumably starts on first day of work. No lag time there. Or maybe just picking up a contract to start a job counts, even if the individual hasn’t succeeded in holding down the job?

So many variables that would need to be defined, and so many that could end up skewing the data in one direction or another if not applied in a balanced manner.

 

Based on table 2, I think the 9% refers to the subgroup of 316 that provided data at baseline and follow-up. Because the numbers below add up to 316.

 

I think the main data can be summarized as follows:

This results in a chi-square value of 0.53 and a p-value = 0.465. The difference was not statistically significant, despite an adequate sample size. So, 16 therapy sessions of CBT at Kings College doesn’t make much difference for your chance to return to work.

It may actually be detrimental because 38% of the original cohort dropped out, and the authors report that ‘those who were working at baseline were less likely to have dropped out of the study’. So those working at baseline were overrepresented in the follow-up data compared to the baseline data.

 

I’ll hold my hand up and admit I did something I really shouldn’t have done: I didn't click through to original figures! (Guilty!)

 

I was looking at that and it's laughable:

So it almost means nothing. Being the UK, someone being on a 0-hour contract would qualify as being employed even if they only work 1 hour per month, as would a student that is unable to attend classes, unless they dropped out. And of course someone switching from 50 hour per week down to just a few hours volunteering would probably classify as remaining in employment. Complete joke paper.

Although there are two papers discussed here and I'm losing track of whether they are the same, or using the same data and just publishing in different places. The amount of obfuscation here is completely disqualifying for academic research, it's really critical to clean things up. It's even clearly on purpose, so that people interpret whatever they want out of it. Journals should not publish claims that require this much interpretation to even figure out what they're even talking about.

There is one piece of potentially useful data: relationships, the impact of chronic illness on relationships is absolutely devastating, it removes the very first support everyone needs from most. And it's likely a massive undercount. This whole domain has been mostly unexplored, a failure that speaks volume of how little we are considered, that no one cares about what is arguably the most important thing in life: love. What it shows is devastating and can't be explained rationally, but then none of this is rational so whatever.

 

And of course much of the findings do not have any significance tests.

 

yes, all those percentages are of the 316 who provided data at follow-up. Unbelievable how they mangled the description in the paper. These people can't even understand their own findings properly, and no one--not the authors, peer reviewers or editors--recognized this obvious problem. What a joke.

 

Trial By Error: More on that Disastrous Employment Paper from Professor Chalder and Colleagues

https://www.virology.ws/2021/12/11/...-paper-from-professor-chalder-and-colleagues/

 

https://twitter.com/user/status/1468721760921137159

https://twitter.com/user/status/1469247763439304706

 

And of course might well have made a difference to your chances of returning to the same level of work. For all we know the the overall work balance could have shifted to less demanding work, potentially much less demanding.

 

KG puts the two papers in sequence:

https://twitter.com/user/status/1469764076377186323

 

https://twitter.com/user/status/1469768773993443332


eta:
https://www.s4me.info/threads/facto...-stevelink-n-t-fear-m-hotopf-t-chalder.10645/

 

The more you folks dig into this study, the fishier it becomes.

 

Same formula, it's tried-and-tested that they can get away with it and mostly consisting of taking credit for the passage of time and leaving out bad data in "drop-outs", or even discharging patients as recovered even when they tell the clinic they are dropping out because it makes them worse:

https://twitter.com/user/status/1471392432319352842

 

Trial By Error: A Letter to Occupational Medicine From Brian Hughes & Me About Prof Chalder’s Latest Disaster

Professor Brian Hughes, a psychologist at the National University of Ireland, Galway, and I have written to Occupational Medicine outlining our major concerns and calling for the paper to be retracted. You can read our letter below or on a pre-print server here.

https://www.virology.ws/2021/12/16/...ughes-me-about-prof-chalders-latest-disaster/

 

Very useful letter by @Brian Hughes and @dave30th, it will be interesting to see if the editorial team at Occupational Medicine are as practiced at brushing off or ignoring coherent on point criticism of articles that their peer review system should have pulled as the usual suspects.

 

If anyone has a PDF of this paper, could you please send it to me! Thanks.

EDIT: Obtained.

 

"Will innumeracy cause this study to be retracted? Don’t count on it…"
https://thesciencebit.net/2021/12/1...-this-study-to-be-retracted-dont-count-on-it/

 

In hindsight, maybe a race to the bottom was a bad idea. Maybe.

Although this did achieve the original goal: making it easier to publish research. Publishing a lot of wrong research (I like that term) is a good way to publish a lot of papers. Many more papers per academic. Less knowledge, but more papers. So clearly more papers must have been the goal, it's the only thing that's actually been achieved. If it wasn't, then, oops, I guess?