Twenty six studies were identified including 23 RCTs (at least 2,087 participants, range 30 to 228) and three non-randomised cohort studies (1,032 participants, range 81 to 719). There were 12 high quality RCTs (score 9 or above), eight moderate quality (score 6 to 8), and three low quality (score less than 6). Two non-randomised studies were of moderate quality (score 7 to 9) and one of low quality (score greater than 7). Follow-up ranged from one month to three years.
Results for pain relief and function
Disc herniation and radiculitis (19 RCTs):
Of the eight fluoroscopy guided RCTs, all with active controls, five RCTs showed positive results for interventions versus controls, or improvements in both intervention groups at three months. Positive findings were also reported after longer follow-up (six months to one year; four RCTs). Results were unclear or not applicable for the three other small studies. Results were significant for local anaesthetic alone, local anaesthetic plus steroid, and steroid alone in the five RCTs.
Of the 11 non-fluoroscopy-guided RCTs, five were placebo-controlled. Seven showed positive results at three months, four of which had longer term follow-up where two still showed a significant positive result and for the two remaining studies results were unclear. One low quality small RCT gave non-significant results. The remaining three RCTs gave negative results after three to 12 months follow-up.
Axial or lumbar discogenic pain (one RCT with an active control and two non-randomised studies, all using fluoroscopy guidance):
Both local anaesthetic and anaesthetic plus steroid showed positive effects at three, six and 12 months follow-up in the RCT. One non-randomised study found a significant positive effect for steroid plus local anaesthetic; and the result was unclear for steroid in the other non-randomised study.
Spinal stenosis (six RCTs and one non-randomised study):
All three fluoroscopy guided RCTs, all with active controls, showed positive improvements in both groups at three and six months. The non-randomised study also found a significant positive effect for steroid at three months. Results varied in the three non-fluoroscopy guided RCTs.
In summary, for lumbar disc herniation with radiculitis the results were positive with good evidence when performed with local anaesthetic and steroids, but evidence was fair when performed with local anaesthetic only. Whereas, for lumbar disc herniation with spinal stenosis the evidence was fair for local anaesthetic plus steroids; and for axial pain without disc herniation the evidence was fair for local anaesthetic with and without steroids.
Results were not reported for most secondary outcomes.