COMPARISON OF ARTHROPLASTY SURVIVORSHIP IN THE U.S. AND SCANDANAVIA

Abstract

Starting in the 1970s, long-term survivorship of total hip and knee arthroplasty has been under investigation for the Scandinavian population with the aid of implants registries. In the United States, no national arthroplasty registry currently exists. Nationwide inpatient discharge databases in the United States have proven useful when comparing the revision burden in the United States and Scandinavia. For this study, we compared the implant survivorship in the Medicare population with contemporaneous registry-based data from well-established and validated Scandinavian arthroplasty registries. The 5% systematic sample of Medicare claims from 1997 to 2004 were examined for primary and revision THA and TKA claims. The Medicare beneficiary ID was used to follow patients longitudinally between primary and revision surgery. De-identified data from the Norwegian and Swedish national hip and knee registry were also obtained for the same time period. During the 8-year study period, 30,583 and 62,878 elective primary total hip or knee replacements, respectively, were extracted from the Medicare data. In the same time period, 41,823 and 15,927 primary total hips or knees replacements were identified from the Norwegian registry. 82,037 primary total hips were identified from the Swedish registry. Survivorship was assessed by the Kaplan-Meier method, and Cox regression was used to evaluate the effect of patient attributes and cross-country comparisons. The K-M estimate showed that 8 years post-primary surgery, 93.6% of THA and 96.2% of TKA remained revision-free among the elderly Medicare population. By comparison, among Norwegians aged 65 and above, 96.0% of THA and 93.6% of TKA remained revision-free. In the US, men had a significantly higher risk of knee revision than women, but no significant gender difference among hip revision. In Norway, men had a significantly higher risk of hip revision, but no differences in knee revisions. In Sweden, men had a significantly higher risk of hip revision (5.4% vs. 3.3%). Older patients had a reduced risk of revision as compared with younger patients, in both the US and in Norway. The survival of THA is significantly better in Norway than in the US with a hazard ratio of 1.64 (p< 0.05). On the other hand, TKA had a better survival experience in the US than in Norway, with a hazard ratio of 0.55 (p< 0.05). This is the first study to evaluate the utility of Medicare as a source of THA and TKA survivorship data and to compare prosthesis outcomes in Medicare with those from Scandinavian arthroplasty registries. Unlike the Norwegian and Swedish registry data, the reasons for revision (e.g., femoral loosening) are not captured and thus greatly limited the value of the Medicare data as a tool to understand the need for revision, thereby helping to improve implant longevity and to reduce the associated cost and burden to the patient and care provider.