Observing the use of knee arthroplasty appropriateness tools in clinical practice: do appropriateness criteria tools predict surgeon decision-making?

Summary Objective The primary aim of this study was to evaluate the agreement between surgeons and two validated total knee arthroplasty (TKA) appropriateness tools, and secondarily to compare Australian appropriateness rates to those reported internationally. Methods A consecutive sample of patients from one public hospital arthroplasty clinic and a convenience sample from private rooms of surgeons in New South Wales, Australia (n = 11), referred for surgical opinion regarding TKA were enrolled over 1 year. Surgeons applied appropriateness tools created by Escobar et al. and the American Academy of Orthopaedic Surgeons (AAOS). Correlation between the appropriateness tools and surgeon's decisions were evaluated. Results There were 368 patients enrolled, and contrasting rates of being “appropriate” for surgery were identified between the Escobar (n = 109, 29.6%) and AAOS (n = 292, 79.3%) tools. Surgeon agreement with the Escobar tool was substantial (ĸ = 0.61, 95%CI: 0.53–0.69) compared to slight with the AAOS tool (ĸ = 0.11, 95%CI: 0.06–0.16). Of those advised against TKA (n = 179, 48.6%), the AAOS tool suggested many patients (n = 111, 62.0%) were “appropriate” compared to the Escobar tool (n = 12, 6.7%). Conclusions Surgeons rated patients seeking opinion for TKA as appropriate over half the time, however the AAOS tool had low correlation with surgeons as opposed to the Escobar tool. This was illustrated by both tools rating a majority of patients to be operated on as appropriate, but only the AAOS tool considering most patients not chosen for surgery to be appropriate. When comparing previously published appropriateness rates, appropriateness in Australia, USA, Spain and Qatar was found to be similar.