Abstract
Purpose
To identify the incidence and reasons for revision of the Oxford prosthesis (OXF) in New Zealand.
Methods
Review and compare UKA and TKA data including patient-generated Oxford scores after operation.
Results
105 surgeons performed 3,624 OXF (66.5% of all UKA). UKA made up 12.8% of all knee arthroplasties. There were 216 OXF revisions and revision rate (RR) of 1.39 per 100 component-years (p100cy); for UKA this was 1.42 p100cy, and for TKA 0.54 p100cy (OXF vs TKA p< 0.0001). The indications for OXF revisions were unexplained pain (38.0%); aseptic loosening (38.0%); bearing dislocation (9.3%) and deep sepsis (4.2%). For TKA, unexplained pain (28%) was significantly lower than from OXF (p=0.005). Revision for deep sepsis was significantly lower for OXF compared with TKA (4.2% vs 13.1%, p < 0.001). The Oxford scores 6 months post-op were excellent or good in 79% of OXF vs 72% TKA patients (p< 0.0001); at 5 years after operation these were 88% for OXF and 81% for TKA (p=0.001). Twenty high-use OXF surgeons (10 or more operations/year) performed 44% of the operations (RR of 1.3 p100cy), 62 medium-use surgeons (2-9/year) performed 54% (RR of 1.3 p100cy). 23 low-use surgeons (2 or less/year) performed 2% (RR of 3.9p100cy). The differences in RR high vs low users (p< 0.001) and medium vs low groups (p< 0.001) were significant. RR for OXF high or medium users was significantly higher than the overall rate for TKA (p< 0.001).
Conclusion
RR for the OXF was 2.5 times greater than that for TKA. Deep infection rate was lower, and 6-month and 5-year function scores were significantly higher OXF vs TKA.
