Clinical applicability
This study failed to show that steroids had a therapeutic effect on VN in terms of DHI score, which reflects functional recovery after VN. However, a negative effect of steroids on VN recovery was found in this meta-analysis according to DHI score. It is important to recognize potential bias affecting this result. With regard to DHI, three of five studies were analyzed in our meta-analysis.6, 15, 17Shupak et al.15 only presented only post-treatment DHI scores, which were evaluated at 1, 3, 6, and 12 months after VN. Another two studies included baseline DHI scores in their analysis.6, 17 Therefore, one problem in our design was addressing analysis of DHI. One method was to analyze the amount of DHI change after treatment (from the pretreatment DHI score). The other method was to compare the final DHI score. Our meta-analysis adopted the latter method, because a return to the nondisabled DHI level is considered a significant improvement for patients who experience a severe handicap from VN. However, the limitation of this approach was that the initial status of functional impairment was not considered. Therefore, this approach might have been subject to selection bias. The study of Yoo et al.17 also might have been subject to selection bias.
An incidental but interesting finding in this study was that the therapeutic effect of steroid treatment was significant despite of its small effect size but disappeared when other treatments (such as vestibular exercise or valacyclovir) were combined with ST and NST. This finding suggests that vestibular exercise or valacyclovir have a therapeutic effect on VN. As the main purpose of this study was not to analyze the effect of vestibular exercise or valacyclovir, we are unable to draw any conclusions regarding these findings. However, our findings suggest that vestibular exercise or valacyclovir can be helpful to patients who cannot take steroids.