Home > DARE Reviews > A meta-analysis of the efficacy of...

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet]. York (UK): Centre for Reviews and Dissemination (UK); 1995-.

Database of Abstracts of Reviews of Effects (DARE): Quality-assessed Reviews [Internet].

A meta-analysis of the efficacy of fibromyalgia treatment according to level of care

J Garcia-Campayo, J Magdalena, R Magallon, E Fernandez-Garcia, M Salas, and E Andres.

Review published: 2008.

CRD summary

This poorly reported review concluded that treating fibromyalgia in specialised care settings offered no clear advantages over treatment in primary care. These conclusions should not be considered reliable due to the pooling of different treatments and outcomes.

Authors' objectives

To compare efficacy of treatments for fibromyalgia when delivered in primary or specialised care settings.


MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL) and PsycINFO were searched from inception to June 2006 (search terms were reported). No language restrictions were applied. A randomised controlled trial (RCT) filter was used. References of review articles and included studies were checked for additional studies. Studies available only as abstracts were not considered for inclusion.

Study selection

Randomised controlled trials (RCTs) that compared any treatment (pharmacological or non-pharmacological) with placebo or active control for fibromyalgia in primary or specialised care settings were eligible for inclusion. Patients could be diagnosed with primary fibromyalgia according to recognised diagnostic criteria. It was unclear whether outcomes were prespecified. Trials that included acupuncture as an intervention were excluded.

Included trials assessed 120 different treatment interventions for primary fibromyalgia. Most treatments were pharmacological. Some involved exercise. Treatment duration varied between less than eight weeks to 24 weeks across studies. Most patients were women aged 30 to 59 years who had been diagnosed with the condition for six to 10 years. Outcomes most frequently reported were pain and global functioning. Most studies were carried out in USA and Canada after 2001.

Two independent reviewers screened all studies for inclusion. Disagreements were resolved by discussion and reaching consensus.

Assessment of study quality

Included studies were assessed independently by two reviewers who used a modified Oxford/Jadad scale that included items on randomisation, allocation concealment, blinding and reporting of patient flow. Possible scores ranged from 1 to 6. Any discrepancies in judgements were resolved by discussion or referral to a third reviewer.

Data extraction

Continuous data were extracted as mean differences between treatment groups with associated standard deviations; authors were contacted where this information was not reported. Where necessary, these data were extracted from graphs. Studies were classified as being based in primary or specialised care settings as a dichotomous variable.

The authors did not report how many reviewers performed data extraction.

Methods of synthesis

Continuous data were pooled as standardised mean differences (SMD) and associated 95% confidence intervals (CI) in a random-effects meta-analysis for overall effect and by selected outcomes. Moderating variables were explored and two studies classes as outliers were removed in a sensitivity analysis. Statistical heterogeneity was assessed using the I2 statistic. Clinical heterogeneity was stated to have been minimised by use of stringent diagnostic criteria.

Results of the review

A total of 33 RCTs that evaluated 120 treatments in 7,789 patients diagnosed with primary fibromyalgia were included. Eleven were rated as high in methodological quality (scored 5 or better). The most commonly absent items were reporting of blinding and patient flow. Nine of the 11 RCTs rated high quality were of pharmacological interventions.

Mean effect size when comparing any active treatment for primary fibromyalgia with a control indicated a significant benefit in favour of the active group (SMD 0.49, 95% CI 0.39 to 0.58, p<0.001). Mean effect size in primary care was reported to be 0.46 (95% CI 0.33 to 0.58; 15 trials) and in specialised care 0.53 (95% CI 0.38 to 0.69; 19 trials). One trial incorporated primary and specialised care arms.

When analysed according to outcome, no statistically significant differences were reported for tender points, pain, insomnia, anxiety/depression and fatigue. Global functioning effect size was reported as similar between the two settings (0.53 for primary and 0.54 for specialised).

A range of moderating variables were assessed for their effects on Fibromyalgia Impact Questionnaire scores and Global Function scores (methodological quality, treatment type, length of treatment, age, gender, country and duration of the disorder). Full results were reported in the original paper.

Analyses were repeated excluding two studies described as outliers (small sample size, short duration, more effective), which did not substantially alter the results.

Statistical heterogeneity was reported as I2 of 75% (it was unclear which analysis this referred to). A funnel plot was described, but not presented.

Authors' conclusions

Treating fibromyalgia in specialised care settings offered no clear advantages over treatment in primary care.

CRD commentary

This review addressed a broad question with searches that covered the most common databases. It appeared that no attempt was made to explore grey literature and abstracts were excluded, which suggested that the review may have been vulnerable to publication bias. It appeared that no language restrictions were applied (which would have removed language bias). An RCT filter was used, which may have reduced the sensitivity of the searches. Inclusion criteria were not clearly described in the paper; the quality criteria were detailed in full. Procedures for study selection and quality assessment were likely to have minimised reviewer error/bias, but data extraction may not have been carried out in the same way. Primary study results and pooled analyses were not clearly reported, particularly in relation to statistical heterogeneity, and it was unclear whether it was appropriate to have pooled these data (a wide range of interventions and outcomes was used). Overall this was a poorly reported review with a lack of individual study results, and the conclusions should not be considered as reliable.

Implications of the review for practice and research

Practice: The authors did not state any recommendations for practice.

Research: The authors stated that cost-effectiveness studies were required to confirm these findings.


Grant 05/90243 Spanish Fondo de Investigaciones Sanitarias.

Bibliographic details

Garcia-Campayo J, Magdalena J, Magallon R, Fernandez-Garcia E, Salas M, Andres E. A meta-analysis of the efficacy of fibromyalgia treatment according to level of care. Arthritis Research and Therapy 2008; 10:R81. [PMC free article: PMC2575627] [PubMed: 18627602]

Indexing Status

Subject indexing assigned by NLM


Adult; Female; Fibromyalgia /therapy; Humans; Male; Middle Aged; Physicians, Family; Randomized Controlled Trials as Topic; Specialization; Treatment Outcome



Database entry date


Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.

CRD has determined that this article meets the DARE scientific quality criteria for a systematic review.

Copyright © 2014 University of York.

PMID: 18627602