Jörg Dirmaier1, Silke Zaun, Uwe Koch, Timo Harfst, Holger Schulz. 1. Institute and Polyclinic for Medical Psychology, Centre for Psychosocial Medicine, University Clinic Hamburg-Eppendorf, Hamburg, Germany. dirmaier@uke.uni-hamburg.de
Abstract
OBJECTIVE: Recent years have shown an increase in the use of questionnaires measuring health-related quality of life to verify the quality of treatment in the field of oncology. An often used cancer-specific questionnaire is the "Quality of Life Core Questionnaire of the European Organization for the Research and Treatment of Cancer" (EORTC QLQ-C30). The purpose of this study is to analyze the psychometric properties of the EORTC QLQ-C30 (version 1) in order to determine the feasibility and appropriateness for its use in inpatient cancer rehabilitation in Germany with heterogeneous diagnoses. METHODS: The questionnaire was administrated to a sample of 972 cancer patients at the beginning of treatment and to 892 patients after treatment. Besides descriptive analysis, the statistical analyses include confirmatory analysis and the multitrait/multimethod approach to test the questionnaire's postulated scale structure (factorial validity) and its reliability (internal consistencies). The analysis also includes a comparison of responsiveness indices (effect size, reliable change index) to test the sensitivity of the instrument. RESULTS: The EORTC QLQ-C30 showed satisfactory levels of reliability and sensitivity, but the postulated scale structure could not be confirmed. The results illustrate that the varimax-rotated solution of a principal component analysis does not confirm the scale structure postulated by the authors. Correspondingly, the selected fit indices within the scope of the confirmatory factor analysis do not show satisfactory results either. SIGNIFICANCE OF RESULTS: We therefore consider version 1 of the EORTC QLQ-C30 to be only limitedly useful for the routine assessment of changes in the quality of life of cancer patients in inpatient rehabilitation in Germany, especially because of the instrument's length and possible redundancies. For this reason, a scoring procedure limited to a subset of items is suggested, revealing satisfactory to good psychometric indices. However, further psychometric tests are necessary, especially with regard to validity and sensitivity.
OBJECTIVE: Recent years have shown an increase in the use of questionnaires measuring health-related quality of life to verify the quality of treatment in the field of oncology. An often used cancer-specific questionnaire is the "Quality of Life Core Questionnaire of the European Organization for the Research and Treatment of Cancer" (EORTC QLQ-C30). The purpose of this study is to analyze the psychometric properties of the EORTC QLQ-C30 (version 1) in order to determine the feasibility and appropriateness for its use in inpatient cancer rehabilitation in Germany with heterogeneous diagnoses. METHODS: The questionnaire was administrated to a sample of 972 cancerpatients at the beginning of treatment and to 892 patients after treatment. Besides descriptive analysis, the statistical analyses include confirmatory analysis and the multitrait/multimethod approach to test the questionnaire's postulated scale structure (factorial validity) and its reliability (internal consistencies). The analysis also includes a comparison of responsiveness indices (effect size, reliable change index) to test the sensitivity of the instrument. RESULTS: The EORTC QLQ-C30 showed satisfactory levels of reliability and sensitivity, but the postulated scale structure could not be confirmed. The results illustrate that the varimax-rotated solution of a principal component analysis does not confirm the scale structure postulated by the authors. Correspondingly, the selected fit indices within the scope of the confirmatory factor analysis do not show satisfactory results either. SIGNIFICANCE OF RESULTS: We therefore consider version 1 of the EORTC QLQ-C30 to be only limitedly useful for the routine assessment of changes in the quality of life of cancerpatients in inpatient rehabilitation in Germany, especially because of the instrument's length and possible redundancies. For this reason, a scoring procedure limited to a subset of items is suggested, revealing satisfactory to good psychometric indices. However, further psychometric tests are necessary, especially with regard to validity and sensitivity.
Authors: Linda M Frazier; Virginia A Miller; Douglas V Horbelt; James E Delmore; Brigitte E Miller; Ellen P Averett Journal: Cancer Control Date: 2009-01 Impact factor: 3.302