A meta-analysis of the literature on epidural corticosteroid injections for the treatment of sciatica, incorporating data from 23 trials, has shown that this treatment has no effect on long-term (≥12 months) pain or disability. Leg pain, back pain and disability levels were converted to 0–100 scales. Small but statistically significant differences were seen in short-term (>2 weeks but ≤3 months) leg pain and disability, with mean changes of −6.2 and −3.1, respectively. Nevertheless, thresholds for clinically important changes are considered to be between 10 and 30.