In the majority of patients with osteoarthritis of the knee the disease originates in the medial compartment. There are two different approaches to replacing this arthritic area. Some surgeons feel that it is always best to replace both the knee compartments with a Total Knee Replacement (TKR). Others feel it is best to replace just the damaged component of the knee with a Unicompartmental Knee Replacement (UKR). There is little agreement amongst knee surgeons. Both interventions are established and well documented procedures, yet little evidence exists to support either practice. Each intervention is considered standard care. There exists little evidence, however, to prove the clinical and cost effectiveness of either management option.

The aim of the Total or Partial Knee Arthroplasty Trial (TOPKAT) will be to assess the clinical and cost effectiveness of Total Knee Replacements versus Unicompartmental Knee Replacements in patients with medial osteoarthritis. This will be examined using an appropriate patient base and long term assessments.

The trial has a combined device/expertise based allocation depending on the local situation. Surgeons who are in equipoise and have sufficient experience to perform both TKR and UKR, randomisation and allocation can be based on "device" (UKR or TKR). The same surgeon will perform the operation for both arms of the study.

For surgeons who hold a preference for one treatment over the other, an "expertise" based randomisation will then occur. UKR surgeons will work alongside TKR surgeons. Patients recruited to the study from these sites will be randomised to one of the treatment options and treated by the appropriate surgeon. In such cases the patient is internally referred to the other surgeon's operating list.

Patients will be recruited by their consultant knee surgeons in collaboration with the local research team. TOPKAT are hoping to recruit 500 patients altogether, with 250 per arm of the trial.

The design of the study is a single layer multicentre superiority type randomised controlled trial of unilateral knee replacement patients. The randomised controlled trial design will help reduce and prevent potential bias influencing the evaluation.

Participants will be randomised to either UKR or TKR. The trial has a combined equipoise/expertise approach. It enables surgeons who are not in equipoise to deliver only one of the two operations while also allowing surgeons in equipoise to provide both operations. A surgeon who is in equipoise ('equipoise surgeon') and has sufficient experience to perform both TKR and UKR will deliver the allocated operation (UKR or TKR). The same surgeon will perform the operation for both arms of the study.

Not all surgeons are able to exhibit this equipoise. They may hold a preference for one treatment over the other often due to experience/expertise with one type of operation. Interestingly, a surgeon may also believe the patient may benefit from one particular operation even though they may not be able to perform it themselves.

Equipoise is difficult to investigate or establish. Self declaration has been used as the main approach but in order to sufficiently secure this state the following aspects are important:

• The equipoise considered must be patient- or individual-based equipoise rather than an overall or general category equipoise based on operation type. The surgeon must consider their position for each individual patient. Only if they believe that either operation will be suitable for an individual patient can the patient then be recruited.
• No surgeon will ever knowingly perform what they consider a substandard surgical procedure.

In order to complete the trial by seeking to maximise surgeon participation, an 'expertise'-based delivery of the intervention will also occur. For this approach there must be a surgeon with expertise in TKR and a surgeon with expertise in UKR in the same centre who will act together as a 'delivery unit'. Patients recruited to the study who are under the care of such a surgeon ('expertise surgeon') will be randomised to one of the two groups and treated by the appropriate surgeon. This 'expertise' approach allows for those UKR surgeons who work alongside TKR surgeons to team up and participate in the study. Subsequent surgery may be carried out by a surgeon different to that at the initial consultation. In such cases the patient is internally referred to the other surgeon's operating list. No restriction is made upon the number of delivery units within a centre. A surgeon can only be in one delivery unit, that is, they are either an 'equipoise surgeon' or an 'expertise surgeon'.

Inclusion Criteria

• Medial compartment osteoarthritis with exposed bone on both femur and tibia
• Functionally intact Anterior Cruciate Ligament (superficial damage or splitting is acceptable)
• Full thickness and good quality lateral cartilage present
• Correctable intra-articular varus deformity (suggestive of functionally intact medical cruciate ligament)
• Medically fit showing an American Society of Anesthesiologists (ASA) of 1 or 2

Exclusion Criteria

• Require revision knee replacement surgery
• Have rheumatoid arthritis or other inflammatory disorders
• Are unlikely to be able to perform required clinical assessment tasks
• Have symptomatic foot, hip or spinal pathology
• Previous knee surgery other than diagnostic arthroscopy and medial menisectomy
• Previously had septic arthritis
• Have significant damage to the patella-Femoral Joint especially on the lateral facet.