The purpose of the study is to explore if a subgroup of people with osteoarthritic knees can be identified using the McKenzie System of Mechanical Diagnosis and Therapy.

In the spine this subgroup, termed derangement,has been shown to respond rapidly to specific directional exercises.

The trial will explore whether these derangements in the knee respond to specific exercises compared to a control group with no exercises and non-derangement knees given general exercises.

The Mechanical Diagnosis and Therapy (MDT) approach has been extensively used to classify and treat patients with spinal pain. Studies have shown this approach to be valid, reliable and able to predict outcome. Although the approach has been used for extremity joints there is no research on using this approach to classify individuals presenting with knee osteoarthritis.

An assessment tool that could potentially identify a subgroup of patients who would experience dramatic and rapid improvement to conservative care would be valuable.

Patients with an "osteoarthritic knee" diagnosis will be recruited after a consultation with an orthopaedic surgeon. Patients who consent to participate will be randomized into an intervention group and a control group. Baseline self reported function and pain will be collected. The control group will continue as planned on the waiting list for either a follow up orthopaedic consultation or for knee surgery. The intervention group will be assessed by one of three McKenzie credentialed therapist over 3 assessment sessions. The therapist will classify the patients as either having a "derangement" or not. Those classified as a derangement will have 2-3 follow up sessions and will be given direction specific exercises consistent with the principles of the McKenzie System over 2 weeks. Those patients who were not classified as derangements will be given 2-3 sessions of evidenced based osteoarthritis treatment consisting of strengthening exercises and advice on aerobic fitness exercise. At 2 weeks both groups will have baseline measures reassessed and will be discharged. Follow-up by telephone at 3 months and 1 year will reassess functional and pain measures.

Due to the high prevalence of knee osteoarthritis and the associated economic burden on our health care system it is important to investigate if physiotherapists are able to predict who will respond to conservative therapy. Classifying individuals with knee osteoarthritis into rapid responders or non responders to short-term physiotherapy treatment will allow more expensive medical evaluation and intervention to be directed to appropriate patients and avoid unnecessary treatment for patients likely to recover from less costly therapy.

Inclusion Criteria:

• Diagnosed with knee osteoarthritis
• Pain for longer than four months
• Knee X-ray/CT/MRI showing osteoarthritic changes
• Able to attend physiotherapy 2-3 times per week for 2 weeks
• Able to participate in exercise based therapy

Exclusion Criteria:

• Unable to provide informed consent
• Unable to understand written or spoken English
• Neurological conditions