Insufficient Evidence for the Effect of Corticosteroid Treatment on Recovery of Vestibular Neuritis

Abstract

The authors studied the effect of corticosteroid treatment on clinical recovery and recovery of vestibular function in patients with vestibular neuritis. The comprehensive search (March 29, 2012) yielded 496 original papers, of which 5 (including 199 patients) during full-text screening satisfied our eligibility criteria. Methods assessment showed that 1 study (30 patients) provided direct evidence and carried low risk of bias. Two studies properly reported on their random and concealed allocation of treatment. In 1 study, patients were not randomly allocated to treatment. Blinding of outcomes was lacking in 2 studies, whereas outcome data were clearly incomplete for 2 studies. Given the wide variety in outcome measures and scales and follow-up duration, the meaning of the size of reported effects is not clear. Therefore, the reported effects cannot simply be compared between studies, and this precludes pooling of study results. Still, there are large differences between studies in the size of the reported absolute effects after the placebo treatment. Moreover, the difference in effects between treatments is rather small and does not always favor corticosteroids. The moderate to high risk of bias of studies precludes firm conclusions, whereas the reported short-term effects on symptom recovery and improvement of peripheral vestibular function are too small to be clinically important. No long-term effect on symptom recovery has been shown. Recommendations or decisions for corticosteroid treatment in patients with vestibular neuritis cannot be based on current best evidence and therefore should be discussed with the patient.