Objective: To review the results of the Avon patellofemoral arthroplasty at 5 to 8 years.

Methods: The Avon patellofemoral arthroplasty was introduced in clinical practice in September 1996. We present a prospective cohort review of all patients treated in the first three years. Patients were evaluated using the Bristol knee score (BSK), the Melbourne patella score (MPS) and the Oxford knee score (OKS).

Results: 109 patellofemoral arthroplasties were performed in 92 patients between September 1996 and November 1999. The median age was 68 years (range 46 to 86 years). Nine patients (12 knees) died and two patients (two knees) were unfit for follow-up. Ten knees in 9 patients were lost to follow-up giving a follow-up rate of 89%. The mean period of follow-up was 5.6 years.

The median pain score rose from 15/40 pre-operatively to 40 points at eight years. The median (MPS) rose from 10/30 points pre-operatively to 25 points at eight years. The median (OKS) rose from 18/48 pre-operatively to 38 at eight years. 87% of knees had mild or no pain at eight years. There were no cases of failure of the prosthesis itself. All 15 revisions resulted from progression of arthrititis in the tibio-femoral joint. The five-year survival rate for all causes with 86 cases at risk was 96%.

Conclusions: The results show that this type of arthroplasty can give predictable pain relief and excellent functional improvement in patients with isolated patello-femoral arthritis. Disease progression is the main reason for revision to total knee replacement and great care is required in assessing the indications for this procedure.

Purpose: To compare the ten-year survivorship results of an established total and medial compartment knee replacement performed in a single centre over an eighteen year period.

Method: Since 1978 knee replacements have been prospectively recorded in Bristol on a database. Regular clinical and radiological review has been undertaken every two or three years up to twenty years. 408 medial St Georg Unicompartmental replacements and 531 Kinematic total knee replacements have been subject to survivorship analysis using three failure end points. One - revision or removal of the implant. Two – revision or removal and moderate or severe pain. Three – the worst case including all patients lost to follow-up.

Results: The follow-up rate was 97% in both groups. 212 patients (562 knees) died and 31 patients (35 knees) were lost to follow-up. At ten years 25 medial Sled and 20 Kinematic knee arthroplasties had been revised. There was no significant difference (p > 0.05) in the number of good and excellent results. The mean range of movement at the last follow-up was 109 degrees for the medial Sled and 100 degrees for the Kinematic (p< 0.01). 94% of the medial Sled patients obtained a range of movement equal to or greater than 90 degrees whereas only 84% of Kinematic patients obtained this range (p < 0.05).

The ten-year survivorship figures were similar for both groups. Revision as the end point was 87.5% for the medial Sled and 89.6% for the Kinematic knees. When moderate and severe pain was considered in addition to revision ten-year survivorship was 79.4% in both groups. The worst case survivorship was also 74% in both groups.

Conclusion: The ten-year survivorship results of a fixed bearing, non-congruous, Unicompartmental arthroplasty are as good as those of a total knee replacement when performed in a single centre by two consultant surgeons and a variety of trainees. The advantages of a more rapid recovery and better quality result are offset by an easier though slightly higher revision rate.