Sixteen studies (n=1,617) were included.
Steroids (9 studies including 4 RCTs).
The pooled RR for good facial recovery for the RCTs was 1.16 (95% CI: 1.05, 1.29); there was no evidence of heterogeneity (p=0.59). No study showed a significant differences in the time to recovery between steroid-treated patients and controls. Three studies reported side-effects of steroid treatment. These occurred in 1 to 4% of the patients and included dyspepsia, loss of blood sugar control, recurrent duodenal ulcers, mood swings, and acute psychosis. All effects resolved when the treatment was stopped.
Acyclovir (3 studies including 1 RCT).
The RCT found a significant benefit of acyclovir combined with prednisone when compared with prednisone alone. The RR for good facial outcome was 1.22 (95% CI: 1.02, 1.45). The reported frequencies and nature of the side-effects were similar to those with steroids.
Surgery or facial nerve decompression (4 studies, 0 RCTs).
All studies were of poor quality (class IV). Only one study found a significant association between surgery and improved facial outcome. Permanent unilateral deafness was the most common serious side-effect, with one old study reporting deafness in 15% of the patients. More recent studies report much lower complication rates.