A Scandinavian experience of register collaboration: the Nordic Arthroplasty Register Association (NARA)

Abstract

Background: The Nordic (Scandinavian) countries have had working arthroplasty registers for several years. However, the small numbers of inhabitants and the conformity within each country with respect to preferred prosthesis brands and techniques have limited register research.

Methods: A collaboration called NARA (Nordic Arthroplasty Register Association) was started in 2007, resulting in a common database for Denmark, Norway, and Sweden with regard to hip replacements in 2008 and primary knee replacements in 2009. Finland joined the project in 2010. A code set was defined for the parameters that all registers had in common, and data were re-coded, within each national register, according to the common definitions. After de-identification of the patients, the anonymous data were merged into a common database. The first study based on this common database included 280,201 hip arthroplasties and the second, 151,814 knee arthroplasties. Kaplan-Meier and Cox multiple regression analyses, with adjustment for age, sex, and diagnosis, were used to calculate prosthesis survival, with any revision as the end point. In later studies, specific reasons for revision were also used as end points.

Results: We found differences among the countries concerning patient demographics, preferred surgical approaches, fixation methods, and prosthesis brands. Prosthesis survival was best in Sweden, where cement implant fixation was used more commonly than it was in the other countries.

Conclusions: As the comparison of national results was one of the main initial aims of this collaboration, only parameters and data that all three registers could deliver were included in the database. Compared with each separate register, this combined register resulted in reduced numbers of parameters and details. In future collaborations of registers with a focus on comparing the performances of prostheses and articulations, we should probably include only the data needed specifically for the predetermined purposes, from registers that can deliver these data, rather than compiling all data from all registers that are willing to participate.