Probabilities of TKR patients achieving their goals

Introduction: Prospective total knee replacement patients often enquire about the likelihood that surgery will resolve their knee pain, enable them to walk “normally”, and to resume activities important to them. However, considerable variability exists in patient outcomes, which makes it difficult for clinicians to accurately answer these questions, and for patients to make informed decisions. In this study five patient-centred outcomes were explored: pain resolution, walk without limping, perform usual work, ability to kneel, and satisfaction. Significance: The goal of this paper was not to create a fully predictive framework of outcomes after TKR, but to focus on patient-centred goals using only easily measured baseline factors to see if doing this might be achievable and potentially useful in terms of discussion of expectations in shared decision making for surgery. The added value of personalising each outcome, as opposed to providing all patients with a generic probability, was also assessed from the statistical models. Method: Data from 470 patients was used in multivariable logistic regression analyses to identify independent significant predictors for each goal. Predictors assessed were age, gender, body mass index, preoperative knee function, physical health status and mental health status. Results: The likelihood of achieving a desirable outcome varied across goals examined. Whilst 82% of patients were able to walk without a consistent limp, only 32% could kneel with ease. Furthermore, we identified a consistent pattern where patients with greater preoperative knee function and mental health, had improved odds for attaining each goal. Preoperative physical health and body mass also had some predictive utility. Conclusion: We found that when assessing the merits of undergoing total knee replacement, consideration of a patient’s pre-operative knee function and mental health allows a more accurate prediction of the benefit they may achieve.