Study designs of evaluations included in the review
Published RCTs evaluating a relaxation technique as active treatment, compared with at least one other active or control treatment, and reporting at least one pain outcome measure, were included. Studies with unconcealed treatment allocation, or treatment groups with less than 10 participants, were excluded. Abstracts, review articles, and evaluations of laboratory experiments were not considered for inclusion.
Specific interventions included in the review
Relaxation techniques used with either chronic cancer or chronic non-malignant pain conditions were included. The relaxation techniques were usually progressive muscle relaxation with tape recordings and regular home practice. Studies in which a relaxation technique was studied in combination with another pain intervention, such as cognitive behaviour therapy, biofeedback, hypnosis or imagery, were excluded. The treatment periods varied, as did the frequency and duration of relaxation. The control regimens included waiting-list, alternative relaxation techniques, biofeedback, education, imagery, routine care, occlusal splint and hydro-galvanic baths.
Participants included in the review
Studies recruiting patients with either chronic cancer or chronic non-malignant pain conditions were included. Evaluations of treatments for headache, migraine and acute pain conditions were excluded.
Outcomes assessed in the review
The primary outcome was the change in the severity of pain. A range of different pain outcomes were used across the included trials; the most commonly used was the McGill Pain Questionnaire. All but one study used more than one different subjective patient rating to assess pain. The other study used single investigator ratings of pain on palpation in patients with temporo-mandibular joint dysfunction. The assessments commonly involved pre- and post-treatment clinic assessments and self-report patient diaries. Psychological outcomes, such as anxiety and depression, were also considered.
How were decisions on the relevance of primary studies made?
Each study was read by both authors independently.