Study	Reason for exclusion
Colakoglu 2011	Cross‐over trial. Ineligible comparator. In this trial, participants were allocated to two different application methods. All of them were administered BtA into perioral muscles, but only one of the arms was administered also BtA into the orbicularis oculi, while the other arm received placebo. Then, participants were crossed over. The purpose was to evaluate the necessity of Bt application into lower facial muscles in people with HFS. The authors found that in participants with mild lower facial involvement, toxin application to these muscles might not be necessary. However, these comparisons are not within the scope of our review. Also, this study did not include a placebo arm, and its design (single‐blinded cross‐over) was not contemplated in our protocol criteria.
Jitpimolmard 1998	This study was an open, non‐randomised trial that compared the effects of 2 different techniques of BtA injections. However, these comparisons were not within the scope of our review.
Mezaki 1999	This randomised trial assessed the effects of different BtA doses. However it did not include a placebo arm.
Park 1993	We found this controlled study of BtA versus placebo. In reality, this was a prospective case series of 101 participants with hemifacial spasm, which had first included, in the protocol, a randomised double‐blind phase. It was not clear whether any participants had previously received BtA treatments. Only eight participants with hemifacial spasm were enrolled in the controlled phase. There were no clear data for baseline characteristics, treatment program, or results for BtA and placebo groups regarding these eight participants. Therefore, we excluded this study.
Price 1997	A randomised trial enrolling 42 participants with hemifacial spasm analysed the effectiveness and side effects of four different BtA treatment site applications. The authors found that the brow treatment was equally effective to standard treatment with fewer side effects. However, this study did not include a placebo arm.
Sampaio 1997	This randomised controlled trial compared different BtA formulations (Botox and Dysport at a ratio of 1 to 4) in participants with blepharospasm and hemifacial spasm. No placebo group was included.
Xiao 2018	Compared BtA at different dosages, without an eligible comparator for this review.
Yoshimura 1992	Cross‐over trial. This trial enrolled 11 participants with HFS, who were randomly assigned to 4 sets of injections. Three of these injections were of BtA (formulation Botox) using 3 different doses (low, intermediate, and high dose), and one injection was of placebo (saline). Muscles were selected for injection based on clinical involvement. For each participant, the site of injections was kept constant. BtA dose was determined for each participant on the basis of the number of muscles involved, frequency and severity of the spasms. The average dose ranged from 2.5 U to 10 U per muscle. Doses of one‐half (low dose) and twice this dose (high dose) were administered on different occasions. The total dose administered to participants at any one time varied between 5 U and 90 U. BtA was diluted to a concentration of 2.5 U to 5 U per 0.1 mL. Participants were not re‐injected until any response to the previous injection (as determined by both the participants and physicians) was lost.

BtA: botulinum toxin type A

HFS: hemifacial spasm