J Stausberg1, B Dahmen, S Drösler. 1. Institute for Medical Informatics, Biometry and Epidemiology, Medical Faculty, University of Duisburg-Essen, Hufelandstr. 5, 45122 Essen, Germany. stausberg@uni-essen.de
Abstract
OBJECTIVES: During the adaptation of the Australian Refined Diagnosis Related Groups for Germany mapping tables between procedure classifications were needed. The mapping between the German OPS-301 2.0 and the Australian MBS-Extended should transfer the Australian expertise by keeping a well-established terminology system. METHODS: A methodological framework for the development of mapping tables had been developed based on the model for representation of semantics provided by the European Committee of Standardization. Two approaches were used; the concept-based approach from OPS-301 2.0 to MBS-Extended and the class-based approach the other way round. A conversion had to be identified between 23,160 classes of the OPS-301 2.0 and 6,328 classes of the MBS-Extended in two asymmetrical mapping tables. RESULTS: The class-based approach leads to a low number of 6,980 conversions but misses 82.6% of the classes of the OPS-301. Because of domain incongruencies and missing domain completeness of the OPS-301 2.0 for non-operative procedures 15.7% of the MBS-Extended-classes remain without conversion. The concept-based approach leads to a slightly higher mean number of conversions per class of 1.35 in comparison to 1.31 with the class-based approach. But it was possible to find conversions for 99.5% of the OPS-301 2.0-classes. 16.3% of the DRG-relevant classes of the MBS-Extended were missed. CONCLUSIONS: The class-based approach was not useful, because the MBS-Extended is significantly broader than the OPS-301 2.0. An external validation study for the direction OPS-301 2.0 to MBS-Extended revealed a satisfactory quality. The empirical and the reference-based approach are important alternatives to the ones used in this project. There are clear criteria about the appropriate application area for the methodological approaches presented here.
OBJECTIVES: During the adaptation of the Australian Refined Diagnosis Related Groups for Germany mapping tables between procedure classifications were needed. The mapping between the German OPS-301 2.0 and the Australian MBS-Extended should transfer the Australian expertise by keeping a well-established terminology system. METHODS: A methodological framework for the development of mapping tables had been developed based on the model for representation of semantics provided by the European Committee of Standardization. Two approaches were used; the concept-based approach from OPS-301 2.0 to MBS-Extended and the class-based approach the other way round. A conversion had to be identified between 23,160 classes of the OPS-301 2.0 and 6,328 classes of the MBS-Extended in two asymmetrical mapping tables. RESULTS: The class-based approach leads to a low number of 6,980 conversions but misses 82.6% of the classes of the OPS-301. Because of domain incongruencies and missing domain completeness of the OPS-301 2.0 for non-operative procedures 15.7% of the MBS-Extended-classes remain without conversion. The concept-based approach leads to a slightly higher mean number of conversions per class of 1.35 in comparison to 1.31 with the class-based approach. But it was possible to find conversions for 99.5% of the OPS-301 2.0-classes. 16.3% of the DRG-relevant classes of the MBS-Extended were missed. CONCLUSIONS: The class-based approach was not useful, because the MBS-Extended is significantly broader than the OPS-301 2.0. An external validation study for the direction OPS-301 2.0 to MBS-Extended revealed a satisfactory quality. The empirical and the reference-based approach are important alternatives to the ones used in this project. There are clear criteria about the appropriate application area for the methodological approaches presented here.