Total Knee Replacement (TKR)

Total knee replacement (TKR) patients are often dissatisfied with post-surgery pain and function, highlighting that that surgical and rehabilitation outcome is not optimal.

Our research is to assess and correlate patient function, biological markers, pain and psychosocial factors with the aim of developing interdisciplinary tools to predict TKR outcome and better direct intervention.

If clinicians could predict patient outcome by identifying which patients will respond best to each treatment option, the clinical and health economic benefits associated with targeted therapy would mean that patients would get the right treatment at the right time, avoiding inappropriate surgery or rehabilitation.

Complex variables

Despite a high dissatisfaction rate, TKR patients are generally treated as a homogenous group with disregard for diverse presentations and requirements. Multifactorial data is urgently required to improve diagnosis of osteoarthritis (OA).

We aim to address NICE Research Recommendations to improve OA patient care, and the OARSI OA Biomarkers Global Initiative. These conclude that patient stratification based upon a complexity of variables will enhance patient outcome. Our studies differ from existing patient-based stratification studies for OA in two important ways:

  • we exploit novel decision making and modelling tools based on Principal Component Analysis and uncertain reasoning classification developed for complex data analysis linking biology, psychosocial, biomechanical and clinical function data in individual patients, to discover which factor combinations predict poor outcomes, why, and how this can be remedied by targeting treatments.
  • we are combining detailed multi factorial data for each patient, pre and post intervention, in a longitudinal manner.

Patient modelling

We have enabled multiple protocol studies on the same subject which generates this multi-factor information to model a patient's journey from late stage OA through to recovery. Although OA is primarily a peripheral joint disease there is evidence that processing of pain in patients with knee OA is not a linear transmission of information from the joint to the central nervous system as:

  • radiographic OA and joint pathology do not correlate with self-reporting of pain
  • patients with OA experience abnormal sensory processing (eg. lower pain thresholds in areas around painful joint).

Consequently advances in structural and functional neuroimaging have shed light on some of these discrepancies.

To our knowledge we are the only research centre undertaking this type of study which will allow the development of clinical prognostic outcome tools.