| Study characteristics |
| Methods |
1. Sequence generation: uncertain ‐ prepared by third party statistician using blocked randomisation but method of sequence generation not specified
2. Allocation concealment: low ‐ details provided in sequentially numbered opaque sealed envelopes
3. Blinding: uncertain ‐ self‐report measures; subjects not informed of existence of other groups; some outcome measures relying on assessment by blind assessor who was instructed not to enquire about treatment undertaken, and to stop participants divulging such information; 7 patients did not receive the treatment to which they were randomly assigned (for clinical and ethical reasons) demonstrating contamination and leakage
4. Incomplete outcome data: low ‐ low dropout rate, apparently balanced between groups, with reasons for dropouts specified; analysis method ensured patients with partial data were not excluded, based on assumption that data were missing at random
5. Selective outcome reporting: low ‐ authors detail deviations from trial protocol, which included non‐usage of subanaerobic exercise test and conduction of 18 month follow‐up assessments by post
6. Other sources of bias: uncertain ‐ referral methods changed half way through trial; changes in therapists during trial; interventions guided by programme rather than manual, and assessment of treatment fidelity not specified |
| Participants |
Setting: secondary care; pain management clinic
Sample population: patients managed in local GP practices with presentation consistent with CFS
Sample size: 153
Inclusion criteria: CDC criteria
Exclusion criteria: concurrent severe mental illness, planned or concurrent rehabilitation, inability to attend all sessions, ongoing physical investigations
Baseline characteristics: detailed comparison of pre‐treatment differences on demographic, psychiatric and outcome variables; twice as many males in CBT group than EAS/SMC groups |
| Interventions |
1: 'cognitive behavioural therapy'; group (8‐12 individuals); 8 sessions of 120 minutes over 16 weeks; manualisation not specified; treatment fidelity not specified; therapists were clinical psychologists, specialist physiotherapists, and senior occupational therapists (supervision not specified); therapy attempted to modify thoughts and beliefs about symptoms and illness, and behavioural responses such as rest, sleep and activity, with monitoring of activity and teaching of exercises to be paced up in small increments (with reduction if significant exacerbation of symptoms occurred)
2: 'education and support' ; sharing of experiences and learning of basic relaxation skills serving as a control for the non‐specific elements of group therapy
3: 'standard medical care'; patients managed as usual in primary care |
| Outcomes |
1. MEASURES: SF‐36 physical functioning subscale, SF‐36 mental health subscale, perceived fatigue, incremental shuttle walk test, normal walking speed, HADS anxiety, HADS depression, GHQ, Chalder fatigue scale, HU13, mood ‐ alertness, mood – hedonic tone, mood ‐ anxiety, recall – total words recalled/correct words/incorrect words, simple reaction time – reaction time/trials completed, repeated digits detection – reaction time/hit rate/false alarms, number of subjects with SF‐36 scores within range of normal population (physical functioning subscale/mental health subscale), number of patients with SF‐36 scores at least 15% greater than baseline (physical functioning subscale/mental health subscale)
2. FOLLOW‐UP TIMES: 6 months (2 months post‐treatment), 12 months |
| Notes |
Mean duration of fatigue not specified |
| Risk of bias |
| Bias |
Authors' judgement |
Support for judgement |
| Allocation concealment (selection bias) |
Low risk |
details provided in sequentially numbered opaque sealed envelopes |
