Pham 2004.

Methods	Randomised controlled trial Double‐blind Three arms Multi‐centre (46 centres)
Participants	Outpatients with knee OA Country: France and UK N = 301 Mean age, years: 65 Female 70%; male 30%
Interventions	Diacerein 50 mg BID + 3 × 3 intra‐articular injections of saline solution versus 3 × 3 HA intra‐articular injections (NRD101) and placebo capsules versus 3 × 3 intra‐articular injections of saline solution and daily placebo capsules Duration: one year
Outcomes	Pain (0 to 100‐mm VAS) Lequesne Impairment Index Participants' global assessment (0 to 100 VAS) Percentage of painful days (0 to 100 VAS) Assessment of treatment efficacy by participant and investigator (0 to 5 scale) at the end of the study Radiography after one year (JWS measurement)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The randomised allocation schedule was centralised (Cassene Laboratories)
Allocation concealment (selection bias)	Unclear risk	Not described
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Double‐blinding of participants and personnel
Blinding of outcome assessment (detection bias) Clinical outcomes	Unclear risk	No reference about assessors' blinding to evaluate clinical outcomes
Blinding of outcome assessment (detection bias) Radiographic outcomes	Low risk	Two observers who were unaware of participants' identity, study group, signal knee and sequence of radiographs
Incomplete outcome data (attrition bias) All outcomes	High risk	Intention‐to‐treat analysis was done for efficacy outcomes. Structural evaluation was analysed only in participants with X‐rays in the baseline table. Nine participants were excluded from the analysis
Selective reporting (reporting bias)	Low risk	All data were properly reported