The PACE trial in chronic fatigue syndrome

The findings of the PACE trial seem impressive, but the discrepancy between the definitions of improvement in the protocol and paper requires an explanation. In the paper “clinically useful differences” were defined as 0·5 SD changes in fatigue or physical functioning compared with baseline. However, the criteria for improvement published in the trial protocol were much more demanding. Use of a cut-off score of 75 on the short-form 36 physical functioning subscale, as originally proposed, would halve the number of “recovered” patients. Moreover, consulting the normative data for the scale reveals that the mean score of 59 after both cognitive behaviour therapy and graded exercise improved a chronic fatigue syndrome patient’s physical functioning to the level of someone 40 years older than
himself. Is this a case of “outcome reporting bias”?

Citation:
Stouten, B., Goudsmit, E. and Riley, N. (2011) 'The PACE trial in chronic fatigue syndrome. The Lancet, 377(9780), pp. 1832-1833, doi: 10.1016/S0140-6736(11)60685-5..