Three trials were included in the review (728 participants from more than five centres). All three trials were reported to be high quality (Jadad score of 5).
For all trials, pregabalin had a statistically significant treatment effect with an overall decrease in pain score (WMD 1.15, 95% CI 0.81 to 1.49). All trials showed consistent statistically significant reductions in individual pain score of 50% or more (RR 4.05, 95% CI 3.01 to 5.46) and increases in the proportion of participants who rated themselves as improved on pregabalin compared to placebo (RR 1.45, 95% CI 1.26 to 1.67).
All trials illustrated statistically significant adverse events that favoured placebo in comparison to pregabalin for somnolence (RR 0.21, 95% CI 0.11 to 0.42) and incidence of dizziness (RR 0.22, 95% CI 0.12 to 0.41). There was a consistently lower proportion of participants with oedema in the control group compared to pregabalin treatment, but statistical significance was only apparent on pooling (RR 0.31, 95% CI 0.14 to 0.69).
Heterogeneity was not statistically significant in any analysis. A fail safe number of 11,027 individuals was reported; this suggested that the results were not sensitive to publication bias, but it was not clear how it was calculated.