Gottlieb 2004a.

Study characteristics
Methods	RCT, placebo‐controlled, double‐blind Date of study: 2001 ‐ 2003 Location: 24 centres in USA
Participants	Randomised: 249 participants (mean age 44 years, 174 male) Inclusion criteria Participants with moderate‐severe psoriasis (PASI ≥ 12 or BSA ≥ 10), age ≥ 18 years Non‐response to phototherapy Non‐response to conventional systemic treatment Exclusion criteria Pregnancy, past history of malignant tumours, active infection Dropouts and withdrawals after a 30‐week study period 85/249 (34%) Reasons AE: infliximab 3 mg (7), infliximab 5 mg (3), placebo (1) Lack of efficacy: infliximab 3 mg (11), infliximab 5 mg (5), placebo (26) Other reasons: infliximab 3 mg (12), infliximab 5 mg (10), placebo (10)
Interventions	Intervention A. Infliximab (n = 99), IV, 3 mg/kg, weeks 0, 2, 6, for 10 weeks Control intervention B. Infliximab (n = 99), IV, 5 mg/kg, weeks 0, 2, 6, for 10 weeks C. Placebo (n = 51), IV, equivalent, weeks 0, 2, 6, for 10 weeks
Outcomes	Assessments at 10 weeks Primary outcomes of the trial PASI 75 Secondary outcomes of the trial PASI PGA DLQI AEs
Notes	Funding source, Quote (p 534): "Supported by Centocor Inc" Declarations of interest (p 534): "Drs Gottlieb and Menter have received research support from and served as consultants for Centocor Inc. Drs Baker, Bala, Dooley, Evans, Guzzo, and Marano, and Ms Li, are employees of Centocor Inc. "
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote (p 535): "Randomisation was carried out using adaptive treatment allocation and was stratified by the investigational site". Comment: no description of the method used to generate random sequence
Allocation concealment (selection bias)	Unclear risk	Quote (p 535): "Randomissation was carried out using adaptive treatment allocation and was stratified by the investigational site". Comment: no description of the method used to guarantee allocation concealment
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote (p 535): "Patients and investigators were unaware of treatment assignments. Double blind was achieved and maintained by using an independent pharmacist or staff member to prepare all study infusion" Comment: probably done
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote (p 535): "Patients and investigators were unaware of treatment assignments. Double blind was achieved and maintained by using an independent pharmacist or staff member to prepare all study infusion" Comment: probably done
Incomplete outcome data (attrition bias) All outcomes	Low risk	249 randomised, 249 analysed Methods for dealing with missing data: Quote (p 536): "All randomised patients were included in the efficacy analysis at week 10... Patients who discontinued... were considered to have not achieved the dichotomous end points or were assigned the baseline value for continuous end points after the event occurrence" Comment: done
Selective reporting (reporting bias)	Unclear risk	Comment: no protocol was available. The prespecified outcomes mentioned in the Methods section appeared to have been reported