Harii 2008.

Study characteristics
Methods	Study design‐ multicentre, double‐blind, randomised, placebo‐controlled, two‐ dose, parallel design in glabellar lines in Japanese participants Study date‐ no information Study setting‐ outpatients
Participants	Randomised 142 participants, with mean age of 45.7± 9.1 years. Gender: 90% female, 10% male Inclusion criteria BoNTA‐näıve patients aged 20–64 years with glabellar lines of at least moderate severity at maximal contraction Exclusion criteria Any condition (such as myasthenia gravis, Lambert‐Eaton syndrome, amyotrophic lateral sclerosis, or systemic neuromuscular junction disorder) that could influence the effect of treatment (e.g. deterioration in atonia) Pregnancy History of hypersensitivity to any component of the treatment product Any condition that could impair the safety of the participant (e.g. severe heart, kidney, liver, or respiratory disease); Infection or skin disease at the injection site(s) Use of a peripheral muscle relaxant within 2 weeks of the start of the study; History of surgery at the treatment site(s) Previous aesthetic procedures within 6 months of the beginning of the study Severity of disease‐Quote:  “All subjects had either moderate (50.7%) or severe (49.3%) glabellar lines at maximal contraction.” Ethnicity‐ 100% Japanese
Interventions	Duration of study‐ 16 weeks Intervention OnabotulinumtoxinA (10 U), 2 U (0.1 mL) per site, two injections in each corrugator supercilii muscle and one injection in the procerus muscle for a total of five injection sites (N = 45), OnabotulinumtoxinA (20 U), 4 U (0.1 mL) per site, two injections in each corrugator supercilii muscle and one injection in the procerus muscle for a total of five injection sites (N = 46), Comparator  Placebo = 0.5 mL, 0.1 mL per site, two injections in each corrugator supercilii muscle and one injection in the procerus muscle for a total of five injection sites (N = 49)
Outcomes	Primary outcome Physician‐rated line severity 4 weeks after treatment at maximal contraction Secondary outcomes Physician‐assessed line severity at maximal contraction at all other posttreatment visits, line severity at rest at all visits Participant‐assessed improvement ratings at each visit Patient satisfaction ratings at weeks 4 and 16 and for the entire study period (rated at week 16) Adverse events
Notes	“This study was funded by Allergan, Inc., Irvine, California.”
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Using the random‐number‐generation function of SAS (Statistical Institute, Inc., Cary, NC), subjects were allocated to one of three treatment groups: 10‐U BoNTA, 20‐U " page 725 Comment: we considered this low risk
Allocation concealment (selection bias)	Unclear risk	Quote: "Using the random‐number‐generation function of SAS (Statistical Institute, Inc., Cary, NC), subjects were allocated to one of three treatment groups: 10‐U BoNTA, 20‐U" page 725 Comment: we considered unclear risk of bias due to the authors did not provide the methods used for maintain the allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Quote: "Vials used for treatment administration were coded to maintain the blind." page 725 Comment: we considered unclear risk of bias due to no information about visual aspect of interventions was provided
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Quote: "Vials used for treatment administration were coded to maintain the blind" page 725 Comment: we considered unclear risk of bias due to no information about visual aspect of interventions was provided
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Quote: "Six subjects in the full analysis data set discontinued: two in the 20‐U group discontinued before treatment, two in the 10‐U group moved away, one in the 20‐U group retracted consent, and one in the placebo group became pregnant." page 726 At the beginning the authors mentioned 140 participants, but they treated 139 participants and there was no explanation about the missing one. Comment: We considered unclear risk of bias because data discrepancy. We sent an e‐mail on 23 November 2015, but the electronic address was wrong and we could not find a valid e‐mail.
Selective reporting (reporting bias)	Low risk	All prespecified outcomes were reported Comment: we considered low risk of bias
Other bias	Low risk	We considered this study at low risk of other bias