Combe 2006.

Methods	Method of randomisation was computer generated Allocation concealment not described Double blinding Multicentre, parallel group study Power calculation not reported No of participants randomised = 260 No of participants analysed = 254 Authors stated that they used a modified intention to treat analysis: "all randomly assigned patients who received any test article and provided efficacy data at baseline" Measures to deal with missing data included LOCF Source of funding: Wyeth (some authors either paid consultants or employees of Wyeth)
Participants	Inclusion: At least 18 years; diagnosis of adult onset RA; disease duration ≤ 20 years; swelling in ≥ 6 joints, ≥ 6 tender joints, morning stiffness ≥ 45 minutes, ESR ≥ 28 mm/h or CRP ≥ 20 mg/L Previous stable doses of SSZ at least 4 weeks prior to the study, without signs of toxicity Exclusion: Previous treatment with etanercept or other TNF antagonist; treatment with DMARDs other than SSZ in 3 months before baseline; treatment with other biological agents or immunosuppressants within 6 months prior to the study entrance; or steroid injection in 4 weeks before study start; relevant co‐morbidities; pregnancy or lactation Stable doses of NSAIDs, analgesics or prednisone were allowed
Interventions	Etanercept 25 mg SC twice weekly + oral PBO once daily SSZ tablets (2, 2.5 or 3 g daily) + SC PBO twice weekly Etanercept 25 mg SC twice weekly + SSZ tablets (2, 2.5 or 3 g daily) Duration: 2 years
Outcomes	ACR20, ACR50, ACR70 DAS44‐ESR SJC, TJC, morning stiffness, physician and participant global assessment, pain‐VAS, general health‐VAS, ESR, CRP Functional status (HAQ) EuroQoL Adverse events
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Method of randomisation was computer generated
Allocation concealment (selection bias)	Unclear risk	Allocation concealment not described
Blinding (performance bias and detection bias) Clinical outcomes	Low risk	Stated double blind and treatments were identical
Incomplete outcome data (attrition bias) Clinical outcomes	Low risk	Analyses were based on a modified intention to treat: inclusion of all participants who received at least 1 dose of the study drug
Selective reporting (reporting bias)	Unclear risk	No evidence of a prior published protocol but wide range of outcomes assessed
Other bias	Unclear risk	Source of funding: Wyeth (some authors either paid consultants or employees of Wyeth)