I doubt PACE will use employment outcomes as a criterion for recovery. The results were stark. For example, the mean (SD) of lost employment days in the PACE Trial over 12 months was: CBT = 151.0 (108.2), vs, SMC = 141.7 (107.5). "There was no clear difference between treatments in terms of lost employment." Assuming normal distribution, this would suggest that about 85% of the CBT group were losing a minimum of 42 days per year of work, making the [mean -1SD] rule here as ridiculous as it was when used for "normal" fatigue and physical function which was far from recovery. However, use of the [mean -2SD] rule is impossible here due to a large SD.
Perhaps a reasonable compromise may be a threshold based on the maximum number of sick days allowed in full time employment, although this fails to account for work quality (same problem with school attendance) and type of work (participant may have downgraded their career to compensate for limitations). However, if a participant meets full criteria for "recovery" as per prior protocol (which they may have changed since then), *and* is working full-time, well their health may still not be perfect and actigraphy would have been nice, but this is still substantially better than the dubious threshold for "normal" fatigue and physical function. I would happily take that "recovery".
I would still like to know how many GET participants reached the goal of 30-45 minutes of moderate intensity exercise at least 5 times per week. Objective outcomes for fitness may also help determine who is "recovered". Just look at the buff hamster on the exercise wheel in my current avatar, *that* is what GET participants should be like after the 2.5 year followup from baseline if the primary "deconditioning" model of CFS has much relevance.
I was wondering what these work loss figures mean when benefit take up also increases. To lose work days persumably you need to be employed. But with income benefits increasing persumably as well as losing work days people were also working less hours (or not at all).