Shapiro 2000

Methods	Multi‐centre study, USA
Participants	349 adults and adolescents randomised (4‐treatment arm study; fluticasone propionate/salmeterol: 84; fluticasone propionate: 84). Data from 13 participants excluded from analysis due to poor procedure at 1 site Inclusion criteria: ≥ 12 years of age; ATS‐defined asthma of ≥ 6 months' duration requiring pharmacotherapy for ≥ 6 months; FEV₁ between 40% and 85% predicted; ≥ 15% increase in FEV₁ 30 minutes after 2 puffs of albuterol; use of ICS 12 weeks before the study Exclusion criteria: females with positive pregnancy tests; life‐threatening asthma; hypersensitivity to sympathomimetic drugs/steroids; smoking within previous year; smoking history > 10 pack‐years; use of oral/injectable steroid therapy within 1 month of study; use of daily OCS within 6 months before the study; use of any prescription or over‐the‐counter medication that could have affected asthma or course of treatment; abnormal chest x‐ray; clinically significant abnormal 12‐lead electrocardiogram; history of concurrent disease
Interventions	• Fluticasone propionate and salmeterol 250/50 μg twice daily • Fluticasone 250 μg twice daily Third arm not used in this review
Outcomes	83% in the fluticasone propionate/salmeterol arm and 73% in the fluticasone propionate arm completed the study Paper reports: "no serious drug‐related adverse events. Two patients treated with salmeterol withdrew from the study because of adverse events; however, these adverse events were considered by the investigator to be unrelated to study drug (bilateral subcapsular cataracts and postsurgical infection)" Website SFCA3003: no fatal adverse events. No serious adverse events in fluticasone propionate/salmeterol arm; 1 in fluticasone propionate arm (asthma exacerbation)
Notes	Sponsored by GSK
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Not stated
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias) All outcomes	Low risk	Double blind
Independent Assessment of causation (detection bias) Asthma‐related events	High risk	Causation of SAEs not independently assessed
Incomplete outcome data (attrition bias) All outcomes	Low risk	131/168 (78%) completed the study
Selective reporting (reporting bias)	Low risk	Full data on GSK website