Rehabilitation interventions for neuropathic pain: a systematic review and meta-analysis of randomized controlled trials.

The review covers only what happened immediately after treatment ended. The authors had planned to look at effects beyond six months and could not, so this says nothing about whether any benefit lasts. Most trials were small and badly reported: eleven were rated at high risk of bias and four raised some concerns, and because of the nature of the treatments nobody could be blinded to what they were receiving. The patients were older than usual for chronic pain services (median about 55 to 57 years, compared with 44 years in a large Swedish pain registry), so the findings may not fit the working-age patients you usually see. The proportion of women ranged from 6.4% to 100% across trials, so no useful conclusion can be drawn about who these treatments suit. Seven of the fifteen trials were in painful polyneuropathy and most interventions were cognitive behavioural or mindfulness based, so the pooled result leans towards psychological treatment rather than the mixed rehabilitation package a therapist might deliver. The planned sensitivity and subgroup analyses, including separating central from peripheral neuropathic pain, comparing men and women, and comparing combined with single interventions, were not done because the data were not reported well enough. Pooling many different interventions into one result means the effect of any single treatment cannot be separated from the others. Electrical and magnetic stimulation techniques were deliberately not counted as rehabilitation, so no conclusion about them can be drawn here.

The work was funded by ALF Research Grants, Region Ostergotland, Sweden (grant number RO-962091). The supplied text contains the funding statement and no declarations of competing interests from the authors.

Do not read this as proof that rehabilitation works for chronic neuropathic pain. The result comes from fifteen small trials, most were rated at high risk of bias, no trial could blind the therapist or the patient, the authors themselves graded the evidence low to very low, and everything measured was taken only right after treatment ended.