Winner 1997.

Study characteristics
Methods	Randomized, double‐blind, placebo‐controlled, 4‐period, outpatient, cross‐over, 4‐attack study of oral sumatriptan
Participants	Participants were 12‐17 yrs with > 6 month history of migraine with or without aura as defined by IHS 1988; had 1‐8 moderate or severe migraine attacks monthly during the 2 months prior to screening Randomized (N = 355); withdrawn (N = 194); intention‐to‐treat and primary efficacy analysis (N = 298)
Interventions	Each participant treated up to 4 migraine attacks of moderate or severe intensity in a cross‐over fashion; 1 attack with placebo and 3 with sumatriptan 25 mg, 50 mg, or 100 mg (same dose for all 3 attacks). Participants were randomized in a balanced fashion to 1‐12 treatment sequences.
Outcomes	Headache relief (reduction in severity from moderate or severe to none or mild) at 2 h after the first dose of study drug for the first attack and across attacks were the primary outcome measures; data from treatment of the first migraine attack are used in the systematic review Pain‐free at 2 h Presense of nausea or vomiting at 2 h Use of rescue medication Adverse events Other reported outcomes: Time to meaningful relief (within 4 h) Presense of photophobia or phonophobia Clinical disability Other time points (15, 30, 45, 60, 90, 180, and 240 min)
Headache severity scale	4‐level (none, mild, moderate, severe)
Funding source	GlaxoSmithKline
Publication	Journal and clinical trial registry
Notes	—
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: described as "randomized in a balanced manner to 1 of 12 treatment sequences"
Allocation concealment (selection bias)	Unclear risk	Comment: methods to maintain allocation concealment were not described; previous studies by sponsor had adequate allocation concealment
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: described as "double‐blind, placebo‐controlled"; methods to maintain blinding were not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	Comment: a total of 194 subjects withdrew, but were balanced across intervention groups
Selective reporting (reporting bias)	Low risk	Comment: all expected outcomes reported