Appropriateness and total knee arthroplasty: an examination of the American Academy of Orthopaedic Surgeons appropriateness rating system.

OBJECTIVE

The American Academy of Orthopaedic Surgeons (AAOS) recently published appropriateness criteria for patients with knee osteoarthritis (OA) who are being considered for total knee arthroplasty (TKA). We evaluated the extent to which predictor variables used by the AAOS contribute to final classification, rated as "appropriate," "may be appropriate" or "rarely appropriate."

METHODS

The RAND/UCLA Appropriateness method was used by AAOS to develop 864 clinical vignettes, each incorporating eight evidence-based variables associated with TKA outcome or need. Variables included function-limiting pain severity, knee OA severity, knee motion and age among others. The contribution of each variable to the overall classification was determined using multinomial regression. A classification tree method was applied to determine the combinations of variables that contributed to final classification for each vignette.

RESULTS

Multinomial regression indicated that patient age, knee motion, OA severity and location were the four most powerful predictors of final classification. Function limiting pain, knee instability and lower limb alignment contributed little to the final classification. The classification tree had an accuracy of 86.7% and the most important contributors to classification were age, knee OA severity and pattern.

CONCLUSION

Function limiting pain, the most frequent reason endorsed by patients seeking TKA does not meaningfully contribute to the newly developed AAOS appropriateness criteria. The system is highly dependent on traditional variables that surgeons consider when evaluating patients for TKA: patient age, knee OA severity, knee OA pattern and knee motion.