The following studies were included:
5 randomised controlled clinical trials (RCTs);
3 observational studies with concurrent controls (case-control and cohort studies); and
13 studies of class III evidence (expert opinion, case series, case reports, and studies with historical controls).
In the largest RCT (level I evidence, 457 participants), oral prednisone in doses of 1 mg/kg per day demonstrated no statistically-significant improvement in the speed or degree of visual recovery, compared with placebo. By contrast, the intravenous group had a significantly faster visual recovery than placebo over the first 30 days (p=0.02 and p=0.0001 for the respective primary outcomes). After 6 months, however, there were no significant differences between the three treatment groups in terms of visual acuity. After 12 months of follow-up, visual acuity was better than 20/20 in 69% of the eyes studied, and 20/200 or worse in only 3%.
The remaining 4 level I studies and 3 level II studies lacked the statistical power to exclude a therapeutic benefit of steroid treatment. Results from these trials were mixed.
The 13 level III studies had serious methodological flaws, which were related to very low patient numbers, retrospective analysis, no randomisation, or lack of placebo control. Results from these studies were also mixed.