The employment data wasn't published. The first Cochrane review of CBT highlighted this. This mysteriously disappeared from the second Cochrane review i.e. it wasn't mentioned that the employment data was missing. I wrote to the authors saying I wasn't happy with it and that I expect any update to include it. I got a reply. We shall see what happened.Re: the Cochrane review, I'm having trouble finding the 'improvement over 12 months' employment data from 'Sharpe 1993'.
It is interesting that it wasn't reported as at least one of the (Sharpe et al) authors has worked for insurance companies. They know it is important data.
----------
This details the point I'm making above:
I would like an answer to the following question.
The previous Cochrane Review [Cognitive behaviour therapy for adults with
chronic fatigue syndrome. Price JR, Couper J.
Cochrane Database Syst Rev. 2000;(2):CD001027] contained the following:
"Data reporting
Two studies (Lloyd 1993, Sharpe 1996) did not report standard
deviations. The authors of Sharpe (Sharpe 1996) have kindly provided
the missing standard deviations for this review. Missing data
from Lloyd (Lloyd 1993) are still awaited. It was, therefore,
difficult to interpret the results of Lloyd (Lloyd 1993), as the
absence of standard deviations makes the calculation of effect sizes
impossible.
There were some other deficiencies in published data, such as
Sharpe (Sharpe 1996) not reporting improvement in employment
status despite stating the intention to use this as a subsidiary
measure of functioning. These data are still awaited."
However in the updated review, [Cochrane Database Syst Rev. 2008 Jul
16;(3):CD001027.
Cognitive behaviour therapy for chronic fatigue syndrome in adults.
Price JR, Mitchell E, Tidy E, Hunot V.
Full free text at: http://tinyurl.com/PriceMitchelletal i.e.
http://mrw.interscience.wiley.com/cochrane/clsysrev/articles/CD001027/frame.html ]
despite employment status being one of the outcome measures being
investigated (e.g. page 5) and despite
giving the impression that your review investigates thoroughly
"assessment of risk of bias in included studies" and in particular "5) Selective
outcome reporting: Are reports of the study free of suggestion of selective
outcome reporting?", on page 15, there is no mention that the data was not
reported.
Of course, it could be the case that at some stage Sharpe et al could have
given the data but the data isn't included either.
As I'm sure you understand, the issue of employment status in the area of
CFS is an important one. Arguably it's the most important outcome measure
investigated.
If it was an error, will you get back to me to let me know that it has
been noted for when the review is being updated in the future.
Thanking you,