|
Meta-analysis of biofeedback for tension-type headache: efficacy, specificity, and treatment moderators |
Nestoriuc Y, Rief W, Martin A |
|
|
CRD summary This review found that biofeedback provided significant relief of headache symptoms compared with the pre-treatment state, but was not significantly more effective than drug therapy, physical therapy or cognitive therapy. The lack of evidence for superior efficacy over other treatments limits the reliability of the authors’ conclusion that biofeedback constituted an evidence-based treatment for tension-type headache. Authors' objectives To investigate the efficacy of biofeedback as a treatment for tension-type headache (TTH), its effects on various specific outcomes and to investigate potentially moderating effects of treatment and patient characteristics. Searching The authors searched MEDLINE, PsycINFO, Cochrane Central Register of Controlled Trials and PSYNDEX from inception to March 2007. Search terms were reported. Only published studies in English or German were sought. Additional studies were identified by a manual search (no details reported). Study selection Studies evaluating individually administered biofeedback in patients with TTH were eligible for the review. Diagnosis had to be based on either a standard classification system or a precise description of the disorder. Treatment outcome had to be measured using headache diaries, pain scales, other psychological questionnaires or physiological parameters. Headache pain was the primary outcome of the review. Studies had to have at least three months of follow-up and include at least four patients per treatment arm. Single case studies and studies of patients diagnosed with both TTH and migraine were excluded. Most included studies evaluated electromyography feedback (EMG-FB) with or without relaxation. Temperature feedback, galvanic skin response feedback and electroencephalography feedback were evaluated in one or two studies each. Comparators included no treatment, placebo and various alternative treatments. The number of biofeedback sessions ranged from 6 to 20. Mean patient age, where reported, ranged from 10.3 to 66.7 years (overall mean 35.9 years). The proportion of female patients varied from 43 per cent to 100 per cent (overall 71 per cent). Duration of TTH varied from 1.2 to 42.4 years (overall mean 13.9 years). The authors stated neither how papers were selected for the review nor how many reviewers performed the selection. Assessment of study quality Validity was assessed by two independent reviewers using a 12-point scale designed to assess internal and external validity, construct validity and validity of the statistical analysis. Data extraction For controlled trials, data on mean, standard deviation and sample size in experimental and control groups were used to calculate the effect size (Hedges' g, that is mean difference/pooled standard deviation). For studies with a pre-post design, data on mean and standard deviation before and after treatment were extracted and used with the pre-post correlation coefficient to calculate an effect size. A correction for small sample size was applied, resulting in a revised effect size (d). Data were extracted by two independent reviewers and discrepancies were resolved by discussion. Methods of synthesis Studies were pooled by meta-analysis using a random-effects model with effect sizes weighted by the inverse of their sampling variances. Effects of pre-specified moderator variables (feedback modality, EMG training site, home training, treatment duration, diagnosis, age and duration of TTH) were analysed by random-effects analysis of variance using the Q statistic. Sensitivity analyses were performed to examine the effect of using different formulas for effect-size calculation and the effect of using intention-to-treat analysis. Publication bias was assessed using a funnel plot and by calculation of a fail-safe N. Results of the review Fifty-three studies (n = 1,532) were included in the review: 32 controlled trials (24 randomised); and 21 pre-post studies. Validity scores ranged from 1 to 11 out of a possible 12 (average 5.6). Where reported, follow-up ranged from one to 60 months. Effect sizes for headache relief showed a moderate to large and statistically significant effect size of biofeedback comparing pre- and post-treatment (d = 0.73, 95% confidence interval (CI): 0.61, 0.84). Sensitivity analyses indicated that this was a conservative but reliable estimate.
Based on pooled effect sizes, biofeedback also had a large and statistically significant positive effect compared with no treatment (d = 0.81, 95% CI: 0.46, 1.16) and smaller but statistically significant benefits compared with placebo (d = 0.5, 95% CI: 0.27, 0.73) and relaxation (d = 0.2, 95% CI: 0.09, 0.32). Effect sizes for comparisons with pharmacotherapy, physical therapy and cognitive therapy did not show a significant difference between treatments. Significant pre-post effect sizes were seen for various other outcomes (details reported in the paper). Analysis of moderator variables indicated: EMG-FB with relaxation was the most effective type of biofeedback; biofeedback was more effective in children and adolescents than adults; and in adults biofeedback was more effective in patients with a longer duration of TTH. There was no evidence of significant publication bias. Authors' conclusions Biofeedback constitutes an evidence-based treatment option for tension-type headache. CRD commentary This review had clear inclusion criteria for intervention, participants and outcomes. A wide range of study designs were included. The authors searched a number of relevant databases. Language restrictions meant that relevant studies might have been missed and the review was restricted to published studies, but an assessment of publication bias did not find evidence of significant publication bias in the review. Validity was assessed using a range of relevant criteria, but only limited results were presented. Appropriate methods were used to minimise errors and bias in validity assessment and data extraction, although it was not clear whether similar methods were used in study selection. Relevant details of included studies were presented in the text and tables. Studies were combined by meta-analysis using standard methods for continuous outcomes. Differences between studies were investigated by analysis of pre-specified variables thought likely to influence the effectiveness of biofeedback. The authors' conclusions were in line with the evidence presented, but it should be noted that the strongest evidence relates to a pre- versus post-treatment comparison and not to comparison with other treatments. This limits the reliability of the authors’ conclusion that biofeedback constitutes an evidence-based treatment for TTH. Implications of the review for practice and research Practice: the authors did not state any implications for practice.
Research: the authors stated that studies comparing the efficacy of EMG-FB for episodic and chronic TTH are needed. Future studies should assess functional and behavioural outcomes, such as lost work days, health service use, general activity level and social and role functioning. The authors also suggested research to investigate whether physiological or cognitive changes were important for the success of feedback. Funding Philipps-University, Marburg, Germany Bibliographic details Nestoriuc Y, Rief W, Martin A. Meta-analysis of biofeedback for tension-type headache: efficacy, specificity, and treatment moderators. Journal of Consulting and Clinical Psychology 2008; 76(3): 379-396 Indexing Status Subject indexing assigned by NLM MeSH Behavior Therapy /methods; Biofeedback (Psychology); Chronic Disease; Electromyography; Humans; Sensitivity and Specificity; Tension-Type Headache /diagnosis /therapy AccessionNumber 12008105221 Date bibliographic record published 01/12/2008 Date abstract record published 31/03/2009 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. |
|
|
|