Total knee arthroplasty (TKA) is a proven and cost-effective treatment for osteoarthritis. Despite the good to excellent long-term results, some patients remain dissatisfied. Our study aimed at establishing a predictive model to aid patient selection and decision-making in TKA. Using data from our prospective arthroplasty outcome database, 113 patients were included. Pre- and postoperatively, the patients completed 107 questions in 5 questionnaires: KOOS, OKS, PCS, EQ-5D and KSS. First, outcome parameters were compared between the satisfied and dissatisfied group. Secondly, we developed a new prediction tool using regression analysis. Each outcome score was analysed with simple regression. Subsequently, the predictive weight of individual questions was evaluated applying multiple linear regression. Finally, 10 questions were retained to construct a new prediction tool.Background
Methods
