Four randomised trials (861 patients) were included in the meta-analysis. Trials scored between 56 and 84 points on the methodological assessment scale. Most common limitations were inadequate randomisation, more than 20% loss to follow-up, lack of placebo control and absence of an intention-to-treat analysis. Comparators were steroids with/without local anaesthetic or sham injection. Pain/disability was measured using a variety of different scales.
Ozone treatment was associated with greater relief from pain at six months (OR 2.66, 95% CI 1.94 to 3.63; four RCTs; Ι²=0%)
Eight observational studies were included in the review. Studies scored between 50 and 72 points on the 100-point methodological scale. None had a comparison group. Studies varied in terms of patient groups, follow-up periods and outcome measurement; these generally reported reductions in pain at follow-up and were described individually in the review.