Arthritis & Rheumatism, Volume 63,
November 2011 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Chicago, Illinois November 4-9, 2011.

Predictors of Outcomes of Total Knee Replacement Surgery.

Judge1,  Andrew, Arden2,  Nigel K., Cooper3,  Cyrus, Javaid2,  M. Kassim, Carr1,  Andrew, Field4,  Richard E., Dieppe5,  Paul A.

Oxford University, Oxford, United Kingdom
Oxford NIHR Musculoskeletal Biomedical Research Unit, University of Oxford, Oxford, UK, Oxford, United Kingdom
Southampton General Hospital, Southampton, United Kingdom
Elective Orthopaedic Cnetre, Epsom, United Kingdom
University of Exeter, Plymouth, United Kingdom


Total knee replacement (TKR) is, on average, an effective and cost-effective intervention for severe arthritis. However, a significant minority of patients are dissatisfied with the outcome and have persistent joint pain and disability.

The aim of this study was to identify pre-operative predictors of those who have poor outcomes after a TKR.


We analysed data from a large prospective cohort of patients undergoing a TKR in the UK. Pre-operative data was available on: age, gender, BMI, diagnosis, Oxford Knee Score (OKS), EQ-5D, ASA status as a proxy for general health, and socio-economic status. The primary outcome measure was the 6-month post-operative OKS. We calculated a patient-acceptable symptom state (PASS) by relating OKS outcomes to patient satisfaction with the surgery. Regression modelling was used to identify predictors of outcome.


Full data were available from 1991 patients (mean age 71.7 years, 61% female). The majority improved after surgery, but in some there was either no change or worsening of the OKS score. Using a cut-point of a post-operative OKS of 30 points or more, we identified that 71.7% had achieved an acceptable PASS.

The strongest determinant of outcome was pre-operative pain and function (less severe patients had the best outcomes, multivariate ANCOVA 1.70, 95% confidence intervals 1.43–1.96) ). Other significant predictive factors included diagnosis (those with RA did better than those with OA), socio-economic status (poor status associated with a poor outcome), anxiety and depression, age and sex. BMI was not an important predictor of outcome. Differences were observed between predictors of pain relief and functional gain. Those with OA and anxiety or depression were most likely to have post-operative pain; women, older people were most likely to have a poor functional outcome. However, less than 20% of the variance in outcome was explained by the variables in the model.


Predictors of the pain response to a TKR maybe different from those predicting functional improvement. Although significant predictors of poor pain or functional outcomes were detected in this study, the majority of the variance remained unexplained; other predictive factors need to be identified to improve our ability to select the patients who are most likely to improve.

To cite this abstract, please use the following information:
Judge, Andrew, Arden, Nigel K., Cooper, Cyrus, Javaid, M. Kassim, Carr, Andrew, Field, Richard E., et al; Predictors of Outcomes of Total Knee Replacement Surgery. [abstract]. Arthritis Rheum 2011;63 Suppl 10 :1060

Abstract Supplement

Meeting Menu