Clinical Review ABC of psychological medicine


BMJ 2002; 325 doi: (Published 31 August 2002) Cite this as: BMJ 2002;325:480

Question marks over evidential basis of claims

Sharpe and Wilks' review [1] contains an "evidence-based summary"
with the statement, "graded exercise and cognitive behavioural therapies
are effective in treating chronic fatigue syndrome". However, rigorous
examination of the literature indicates that this remark is not itself evidence-based,
a serious criticism since evidence-based summaries in the BMJ carry weight and
are widely quoted.

For the record, two research groups have now conducted systematic reviews of
putative treatments for chronic fatigue syndrome [2,3]. They have identified
3 eligible randomised controlled trials (RCTs) of graded exercise therapy (GET)
and 5 RCTs of cognitive behavioural therapy (CBT).

The total number of available trials is small, numbers are relatively low (6/8
trials have n<40 in the active groups), and two of the 5 CBT trials do not
show an overall significant effect (Table). No trial contains a "control"
intervention adequate to determine specific "efficacy": in only 2
trials are the treatment arms compared with an 'active', though not indistinguishable,
intervention. Yet, the chart in the clinical review [1] refers to the "efficacy"
of CBT, showing data from one trial (Prins et al, 2001) in which the comparison
groups were guided support (social worker) for 11 sessions (against 16 sessions
of CBT) and no intervention. A number of non-specific effects could have accounted
for these results, and the fact that the drop-out rate in the active arm was
40% may point in this direction, as discussed in one of the reviews [2]. Again,
the heterogeneity of the trials, the potential effect of publication or funding
bias for which there is some evidence [4], and professional doubts about the
evidence base for some behavioural therapies themselves [5] give grounds for
caution. Indeed, if a similar evidence base existed for, say, Shamanic healing
- which has no professional proponents - it would arouse little clinical interest.

Neither of the review groups has commended GET or CBT as particularly effective
for chronic fatigue syndrome patients. Whiting et al. [2] state, "all conclusions
about effectiveness should be considered together with the methodological inadequacies
of the studies. Interventions that have shown promising results include CBT
and GET"; and Mulrow et al. [3] state, "….it is unlikely that
the beneficial effects of such general treatments are specific or limited only
to patients with CFS. In other words, although these therapies may help some
people with CFS, their effectiveness does not help establish an underlying aetiology
or cause of CFS".


  1. Sharpe M, Wilks D. Fatigue. ABC of Psychological Medicine. BMJ 2002;325:480-3.
  2. Whiting P, Bagnall A-M, Sowden AJ, et al. Interventions for the treatment and
    management of chronic fatigue syndrome: a systematic review. Journal of the
    American Medical Association. 2001;286:1360-8.
  3. Mulrow CD, Ramirez G, Cornell JE, et al. Defining and Managing Chronic Fatigue
    Syndrome. Evidence Report/Technology Assessment No. 42. AHRQ Publication No.
    02-E001. Rockville (MD): Agency for Healthcare Research and Quality: October
    2001. Available from:
  4. Abbot NC, Spence VA. Research into ME/CFS in the United Kingdom: Can the
    National Research Register inform future policy? MERGE analysis No. 01-M002.
    February 2002. Available from
  5. Bolsover N. Commentary: The evidence is weaker than claimed. BMJ 2002;384:294.

Table: Randomised controlled trials of CBT and GET identified by recent
systematic reviews.
Comparison groups, diagnostic criteria, and treatment duration show considerable

Study Treatment
Diagnostic Criteria
No. of Participants
Comparison Group for CBT or GET Sessions of CBT or GET
Outcomes Investigated
Overall Effect
Deale et al, 1997 CBT Oxford 60 Relaxation techniques 13 P; Psy; Qol +
Sharpe et al, 1996 CBT Oxford 60 "Medical care" (GP follow-up) 16 P; Psy; Qol +
Risdale et al, 2001 CBT and counselling CDC 1994 45 Counselling 6 P; Psy; Qol NS
Prins et al, 2001 CBT CDC 1994 278 Guided support (social worker); No intervention 16 P; Psy; Qol +
Lloyd et al, 1993 CBT and dialysable-leukocyte extract Australian 90 Saline placebo; "Attend Clinic placebo" 6 P; Psy; Lb; Qol NS
Fulcher & White, 1997 GET Oxford 66 "Flexibility/ relaxation" 12 P; Psy; Lb; Qol +
Powell et al, 2001 GET Oxford 148 "Advice" and return to GP 4, variable P; Psy; Qol +
Wearden et al, 1998 GET and fluoxetine Oxford 136 "GET placebo" (stay within capability) 8 P; Psy; Qol +

Outcomes: P = physical; Psy = psychological; Lb = laboratory; Qol = quality
of life

Overall effect: + = a significant effect for treatment; NS = no significant
effect of treatment.

Competing interests: No competing interests

17 September 2002
Neil C. Abbot
Director of Operations
David J. Newton
MERGE (ME Research Group for Education and Support), The Gateway, North Methven St, Perth PH1 5PP