Evaluation of two appropriateness criteria for total knee replacement

Insurance expansion under the Affordable Care Act will amplify a projected 6-fold increase in total knee replacement (TKR) utilization by 2030 but will not fully address TKR disparities. Promoting appropriate use of TKR would help reduce disparities and improve outcomes. There are currently no validated appropriateness criteria (AC) for TKR in the US. We evaluated the performance of 2 non-US AC in a cohort of US TKR patients.

AC1 was developed in Spain using the modified Delphi method with 624 patient scenarios. AC2 was developed in Canada using the overall Western Ontario and McMaster Universities Osteoarthritis Index score of >39 as the cutoff point for surgery. These criteria were applied to a random sample of TKR patients enrolled in our institutional registry. Preoperative clinical, radiographic, and patient-reported survey data were used in classifying patients. The rate of appropriateness was compared for the 2 AC. Inappropriate cases were further investigated to determine other mitigating factors beyond the criteria influencing the decision to operate.

In total, 508 TKR procedures were evaluated. All patients had osteoarthritic radiographic changes. On the basis of AC1, 7.7% of cases were classified as inappropriate and 11.6% uncertain. On the basis of AC2, 31.5% were classified as inappropriate. Only 4.7% of the cases were classified as inappropriate by both ACs; however, there was poor agreement between the 2 AC (κ = -0.08). Beyond the criteria, failure of nonsurgical treatment and clinically significant valgus/varus deformities influenced the decision for surgery.

There was poor agreement between 2 validated AC for TKR when tested in a US population. Culturally specific AC are needed to promote rational use of TKR.