Comment on: Randomised controlled trial of graded exercise in patients with the chronic fatigue syndrome. [BMJ. 1997]
Editor—Kathy Y Fulcher and Peter D White conclude that their findings support the use of graded aerobic exercise in the management of the chronic fatigue syndrome.1 Those readers who delve no further than the abstract and key points may welcome this “take home message.”
Several flaws in the paper, however, make accurate interpretation of the findings difficult and greatly limit its applicability. Firstly, less than two fifths of those screened for the trial actually entered it. Many of those who did not enter it were excluded on the basis of current psychiatric disorder, even though the Oxford criteria used by the authors do not specifically exclude patients with anxiety and depression.2 Given that this is already a subgroup selected by their referral to psychiatric outpatient departments, to select out those with a current psychiatric disorder makes them an unusual group indeed.
Secondly, successful randomisation should make the intervention and control groups similar. Such comparability should enable the problem of confounders, known or unknown, to be accounted for. Evidence of the comparability of cases and controls should be presented.3 Age and sex are almost universal confounders. In this study one might also add body mass index, duration of illness, and even previous athletic training as other possible confounders. The paper does not break down the age and sex of the two groups to enable comparison.
Thirdly, the main outcome measure is the self rated clinical global impression change score. This score is a validated measure of overall change.4 However, the validity of the subsequent categorisation of patients into those with a score of 1 or 2 and those with a score of 3-7, and whether this was a post hoc categorisation, is not stated. A categorisation into 1-3 (all scores representing an improvement) and 4-7 (the rest) would not have produced a significant change. Numerous other physiological outcome measures are provided, but more clinically relevant would have been consultation rates, use of drug treatment, and time off work (time off work was measured only at one year, after the crossover). In short, for those considering the options for managing this condition, especially those who commission services, the message should be that we need more information before we can tell if graded exercise will help most patients with the chronic fatigue syndrome.
You can read the rest of this comment here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2127633/pdf/9361551.pdf
Source: Sadler M. Graded exercise in chronic fatigue syndrome. Patients were selected group. BMJ. 1997 Oct 11;315(7113):947-8. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2127633/