Ref (type)	Population	Outcome, Interventions	Results and statistical analysis	Effect size	Favours
Pain
[24] RCT	100 people with lumbar disc herniation >25% of cross-sectional area of spinal canal, who had 6 weeks of unsuccessful non-invasive treatment (physiotherapy, chiropractic treatment, rest, analgesia, or a combination) In review [23]	Difference in pain on 11-point visual analogue scale 1 to 3 months with epidural corticosteroid injections (betamethasone 10–15 mg, 1 week apart up to 3 times until successful) with discectomy (no further details reported) Absolute results reported graphically	P = 0.001 The difference between treatments was not sustained at 2 to 3 years' follow-up (results presented graphically; see further information on studies below)	Effect size not calculated	discectomy