A start at summarising points made in the PACE-Trial-and-PACE-Trial-Protocol thread and the PACE Trial letters thread. Selection Criteria Oxford Criteria Extremely broad definition defines a wide heterogeneous group. Makes no mention of the many defining ME/CFS symptoms such as post exercise malaise, flu-like or neurological symptoms. Explicitly includes patients who may be suffering solely from psychological conditions. Not used or recognised outside of the UK. International Definition The authors states that an international definition was used in order to determine subgroups, there is however no published description known as the 'international definition' and it is not made at all clear exactly what criteria were used. The only reference regarding the international definition to this paper which recommends modifications to the CDC 1994 criteria, recommendations include the use of tools such as the Chader Fatigue Scale and SF-36 as primary diagnostic criteria. London ME Unpublished and largely unrecognised and untested. Again the authors gave minimal information as to exactly what criteria they used. The trial design did not allow severely affected participants to be included. A previous study the FINE Trial which did include such participants found treatment effect small and not statistically significant at one year follow-up. The PACE Trial conflated two diseases that the WHO categorises separately - neurological 'ME/PVFS' (ICD-10-G93.3) and psychiatric 'Fatigue Syndrome' (ICD-10-F.48.0) - and likely misrepresented the latter as the former. Outcome Measures: SF-36 and Chalder Fatigue Scale The PACE Trial used no truly objective outcome measures of improvement or non-improvement, any assertions made that the interventions led to objective improvements are not supported by the evidence. All that was really measured by the Chalder Fatigue Scale and SF-36 scores was change in participants beliefs. The summary of the trial results should have read 'Interventions that aim to modify participants illness beliefs were moderately successful in modifying participants illness beliefs'. An objective actigraphy test to measure movement over time was included in the initial published protocol but that test was dropped from the final protocol. Objective tests are particularly important in behaviour therapy trials for chronic fatigue syndrome as use in previous similar trials has shown no significant correlation between subjective and objective measures of therapeutic effect. The six minute walking distance (6MWD) test whilst being somewhat objective may still be subject to psychological bias such as determination and concern for triggering adverse effects. It is not known whether those who may have pushed themselves particularly hard to score highly at 52 weeks suffered any adverse events 24-48 hours or more afterwards. Chalder scale suffers from an extreme ceiling effect with participants initially averaging 28.2 out of 33 with S.D. suggesting many participants started the trial already close to the maximum with little or no scope to measure any worsening of their condition. The value of 60 on the SF-36 scale as a lower boundary of normal range is absurdly low and also below the entry criteria of 65 - meaning some participants likely met criteria for recovery before even receiving treatment. Experimental Bias The trial authors assert that perpetuation of the disorder is a result of to inappropriate and inaccurate illness beliefs why then choose exactly these beliefs as primary outcome measures? Results claimed to be significant are universally low no scores showing more than 15% improvement beyond SMC alone, this is within the scope of what might be expected from placebo/bias alone. Results may well show bias due to different trial groups having been given significantly different expectations of treatment efficacy, training manuals exhibited significantly different language, for example GET participants were led to expect treatment would likely make them 'better or very much better' whilst APT participants were led to expect 'no fundamental change in the course of the illness'. Analysis Ceiling effects and selection cut off points suggest that data cannot be normally distributed. The use of statistical means and standard deviations are therefore inappropriate and likely misleading. No frequency distribution information is given; a 15% mean improvement could indicate 100% of participants improved 15% or it could indicate a small sub-group participants improved by a large amount whilst the majority experienced no benefit. No attempt to correlate factors such as illness severity or duration with treatment outcome was published despite the ease with which this could have been investigated with the data collected. Results Even overlooking the numerous experimental flaws and failure to adequately control for bias and placebo trial results were far from impressive. The most objective measure used, the 6MWD actually showed no improvement in the group that received SMC + CBT versus the group that received SMC alone, a result listed in table 6 but otherwise omitted from any conclusion. SMC + GET did show a moderate improvement (~13%) vs SMC alone. The GET group 6MWD distance average score at 52 weeks of 379m is still far from that which might be considered within normal range. Various other studies have measured reference values including Healthy elderly subjects (50-85 yrs) 631m , healthy subjects aged 55-75yrs 659m , healthy subjects 2050yrs 593m women 638m men, and healthy Chilean women and men (20-80yrs) 576m women 644m men. (see graph below) Results do not support CBT or GET on a cost-effective basis; per session SMC typically showed more than 4 times the benefit of GET or CBT. Between 3 and 6 SMC sessions alone led to a 11.6 point increase in physical function score and a 4.5 point decrease in fatigue score, adding an additional 12 to 15 GET(CBT) sessions improved physical function score by a further 6.9(7.4) points and reduced the fatigue score by a further 3.2(3.5) points. The SF-36 and ChalderFS graphs shown in figure 2 of the paper strongly suggest diminishing returns from the treatments; it is likely that very little further benefit beyond the modest improvements seen could be gained from further sessions of any of the interventions. The participant-rated clinical global impression of change in overall health score found only 41% of the CBT group and 41% GET group experienced positive change from baseline, therefore 60% these participents experienced either minimal or negative change. 25% of the SMC group also reported positive change, this means after a year of treatment only around 15% (1 in 7) of participants improved due to CBT/GET who would not have not have done so with SMC alone. Trial co-author Michael Sharpe quotes a similar figure in an ABC radio interview saying We have a number needed to treat; I think it's about seven to get a clinically important treatment benefit with CBT and GET The moderate benefit shown from GET/CBT is consistent to that seen in other chronic illnesses such as multiple sclerosis, where these non-specific approaches help with managing symptoms and promoting coping, but form only a small part of the clinical and therapeutic armoury required to address the underlying disease processes. Potential Conflicts of Interest The UK Department of Work and Pensions (DWP) who part-funded the PACE Trial, like the permanent health insurance industry, has a potential financial vested interest if patients' disability can be portrayed as caused by or largely exacerbated by behavioural factors. Professor White and his co-Principal Investigators all are believed to have financial links with the health insurance industry.