A matched comparison of long term outcomes of total and unicompartmental knee replacements in different ages based on national databases: analysis of data from the National Joint Registry for England, Wales, Northern Ireland and the Isle of Man.

BACKGROUND: The two main treatment options for end stage single compartment knee arthritis are unicompartmental (UKR) or total knee replacement (TKR). We compared the long term outcomes in different age groups. METHODS: 54,215 UKRs and 54,215 TKRs from the National Joint Registry and Hospital Episode Statistics database were propensity score matched and Kaplan-Meier and regression analysis used to compare revision, reoperation, mortality and three month complications. RESULTS: UKR had higher 10 year (yr) revision rates (12% versus 5%, Hazard Ratio (HR) 2.31, p<0.001) and 10 yr re-operation rates (25% versus 21%, HR 1.12, p<0.001). UKR had lower 10 yr mortality rates (13.6% versus 15.5%, HR of 0.86, p<0.001). UKR had lower rates of medical (p<0.001) and procedure related (p<0.001) complications and deaths (HR 0.61, p=0.02). If 100 patients had a UKR instead of a TKR then over 10 yrs, if they were <55 yr old there would be 7 more re-operations and 1 less death, 55-64 yr 6 more reoperations and 2 more deaths; 65-74 yr 4 more re-operations and 2 less deaths, and ≥75 yr 2 more re-operations and 4 less deaths. CONCLUSIONS: UKR have higher revision and slightly higher re-operation rates but lower mortality rates than matched TKR. The decision to do a UKR should, in part, be based on the balance of these risks, which are influenced by patient age. In the elderly group (>75 yr) the data suggests that UKR compared to TKR have a greater absolute reduction in mortality than the increase in re-operation rate.