[en] OBJECTIVE: To assess the incidence of Total Joint Replacement (TJR) during the long-term follow-up of patients with knee osteoarthritis (OA) formerly receiving treatment with glucosamine sulphate or placebo. METHODS: Knee OA patients participating in two previous randomised, placebo-controlled, double-blind, 3-year trials of glucosamine sulphate and receiving treatment for at least 12 months, were systematically contacted to participate in a long-term follow-up retrospective assessment of the incidence of total knee replacement. RESULTS: Out of 340 patients with at least 12 months of treatment, 275 (i.e., 81%) could be retrieved and interviewed for the present evaluation: 131 formerly on placebo and 144 on glucosamine sulphate. There were no differences in baseline disease characteristics between groups or with the patients lost to follow-up. The mean duration of follow-up was approximately 5 years after trial termination and treatment discontinuation, making up a total of 2178 patient-years of observation (including treatment and follow-up). Total knee replacement had occurred in over twice as many patients from the placebo group, 19/131 (14.5%), than in those formerly receiving glucosamine sulphate, 9/144 (6.3%) (P=0.024, chi-square test), with a Relative Risk that was therefore 0.43 (95% confidence interval (CI): 0.20-0.92), i.e., a 57% decrease compared with placebo. The Kaplan Meier/Log-Rank test survival analysis confirmed a significantly decreased (P=0.026) cumulative incidence of total knee replacements in patients who had received glucosamine sulphate. A pharmacoeconomic analysis in a subgroup of subjects suggested that patients formerly on glucosamine sulphate had recurred to less symptomatic medications and use of other health resources than those from the placebo group during the last year of follow-up. CONCLUSIONS: Treatment of knee OA with glucosamine sulphate for at least 12 months and up to 3 years may prevent TJR in an average follow-up of 5 years after drug discontinuation.