Twenty-two studies (n=1,747) were included. These studies provided 205 ES. The sample size ranged from 20 to 293.
The scores for quality ranged from 0.17 and 0.64 out of 1. The review stated that details of the scores for individual studies are available from the authors.
For all treatments, overall time periods, all outcomes and all control treatments, psychological interventions were significantly superior to control (d=0.16, p=0.02; 21 studies).
The number of comparisons used in the subgroup analyses ranged from 3 to 11.
Psychological plus multidisciplinary treatments.
Significant positive effects of psychological/multidisciplinary interventions compared with any control were found for pain intensity (d=0.27, p=0.01), HRQOL (d=0.41, p=0.05) and pain interference (d=0.23, p=0.01), but not depression. Significant positive effects for psychological/multidisciplinary interventions compared with wait-list control were found for pain intensity (d=0.50, p=0.00) and HRQOL (d=0.42, p=0.05), but not pain interference or depression. There were no significant differences between psychological/multidisciplinary interventions and active controls for pain intensity or pain interference.
At follow-up, significant positive effects were found for the effects of psychological/multidisciplinary interventions compared with active control on disability (working: d=0.36, p=0.02), but there were no significant differences between treatments for pain intensity, pain interference, home care visits or medications.
At long term follow-up, significant positive effects of psychological/multidisciplinary interventions compared with active control were found for disability (working: d=0.53, p=0.03).
Psychological or multidisciplinary treatments.
Significant positive effects post-treatment were found for any psychological treatment compared with wait-list for pain intensity (d=0.52, p=0.00), but not HRQOL or depression. Significant positive effects post-treatment were also found for multidisciplinary interventions compared with active control for pain interference (d=0.20, p=0.03), but not pain intensity. Significant positive effects were found for multidisciplinary interventions compared with active control for disability (working) at follow-up (d=0.36, p=0.02) and long-term follow-up (d=0.53, p=0.03), but not pain interference or pain intensity at follow-up.
Individual psychological treatments.
Significant positive effects post-treatment were found for CBT compared with wait-list control for pain intensity (d=0.62, p=0.00), but not HRQOL. There were no significant differences between CBT and SRT for pain intensity or depression post-treatment, or pain intensity at follow-up.
Significant positive effects post-treatment were found for SRT compared with wait-list control for pain intensity (d=0.75, p=0.00) and depression (d=0.81, p=0.02).
Analyses generally showed mild to moderate heterogeneity. The fail-safe N values ranged from 0.33 to 8.74.