Uwe Konerding1, Jörn Moock, Thomas Kohlmann. 1. Trimberg Research Academy, University of Bamberg, c/o Kapuziner Strasse 16, 96047, Bamberg, Germany. uwe.konerding@uni-bamberg.de
Abstract
PURPOSE: EQ-5D, HUI II and SF-6D often produce very different valuations for the same health state. This paper aims at clarifying to what extent this might be caused by differences between the multi-attribute classification systems belonging to these instruments. METHODS: Subjects were 264 patients of rehabilitation clinics in Mecklenburg-Western Pomerania (44.3% female; mean age 49.1) who completed the EQ-5D, the HUI II and the SF-36 (the basis of the SF-6D). After scaling with principal component analyses for categorical data, each attribute of each classification system was regressed on the classification systems of the other two instruments, and all attributes together were subjected to ordinary principal component analysis with varimax rotation. RESULTS: Adjusted multiple R(2) for regression analyses ranged from 0.01 to 0.57. The HUI II attribute 'sensation' and the SF-6D attribute 'role limitation' are virtually unrelated to the remainder. All other attributes of all three instruments can be predicted by each other. EQ-5D and HUI II focus distinctly more on physical functioning than SF-6D. CONCLUSION: Although all three classification systems have a lot in common, they differ so much that EQ-5D, HUI II and SF-6D would produce different valuations even if these valuations were determined according to the same principle.
PURPOSE: EQ-5D, HUI II and SF-6D often produce very different valuations for the same health state. This paper aims at clarifying to what extent this might be caused by differences between the multi-attribute classification systems belonging to these instruments. METHODS: Subjects were 264 patients of rehabilitation clinics in Mecklenburg-Western Pomerania (44.3% female; mean age 49.1) who completed the EQ-5D, the HUI II and the SF-36 (the basis of the SF-6D). After scaling with principal component analyses for categorical data, each attribute of each classification system was regressed on the classification systems of the other two instruments, and all attributes together were subjected to ordinary principal component analysis with varimax rotation. RESULTS: Adjusted multiple R(2) for regression analyses ranged from 0.01 to 0.57. The HUI II attribute 'sensation' and the SF-6D attribute 'role limitation' are virtually unrelated to the remainder. All other attributes of all three instruments can be predicted by each other. EQ-5D and HUI II focus distinctly more on physical functioning than SF-6D. CONCLUSION: Although all three classification systems have a lot in common, they differ so much that EQ-5D, HUI II and SF-6D would produce different valuations even if these valuations were determined according to the same principle.
Authors: J E Ware; M Kosinski; B Gandek; N K Aaronson; G Apolone; P Bech; J Brazier; M Bullinger; S Kaasa; A Leplège; L Prieto; M Sullivan Journal: J Clin Epidemiol Date: 1998-11 Impact factor: 6.437
Authors: I-Chan Huang; Chyng-Chuang Hwang; Ming-Yen Wu; Wender Lin; Walter Leite; Albert W Wu Journal: Value Health Date: 2008 May-Jun Impact factor: 5.725
Authors: Liam M Hannan; David G T Whitehurst; Stirling Bryan; Jeremy D Road; Christine F McDonald; David J Berlowitz; Mark E Howard Journal: Qual Life Res Date: 2017-03-02 Impact factor: 4.147
Authors: David G T Whitehurst; Nicole Mittmann; Vanessa K Noonan; Marcel F Dvorak; Stirling Bryan Journal: Qual Life Res Date: 2016-04-20 Impact factor: 4.147
Authors: Sanjeewa Kularatna; Joshua Byrnes; Yih Kai Chan; Chantal F Ski; Melinda Carrington; David Thompson; Simon Stewart; Paul A Scuffham Journal: Qual Life Res Date: 2017-08-01 Impact factor: 4.147
Authors: Uwe Konerding; Sylvia G Elkhuizen; Raquel Faubel; Paul Forte; Tomi Malmström; Elpida Pavi; M F Bas Janssen Journal: Health Qual Life Outcomes Date: 2014-12-05 Impact factor: 3.186