Reply to Moons

Abstract

QOL is poor. Too often, the SF-36 is used as a health-related QOL instrument, whereas it says more about the perceived health status of the respondents. Second, the authors correctly explain in the methods section that the raw scores for each of the eight dimensions of the SF-36 have to be transformed into a score ranging from 0 to 100 [1]. However, in the results, they report SF-36 scores that are higher than 100. In the discussion section, they argued, 'the authors did not transform the raw SF-36 values, allowing age-and gendermatched comparison [1].' It is surprising that the authors use this comparison as an argument for not transforming the dimension scores. There is a plethora of articles in which the dimension scores were transformed, but still compared with matched data from normative groups. The fact that the scores were not transformed to a scale from 0 to 100 might be responsible for their findings being aberrant from the previous studies that have used the SF-36 in adults with congenital heart disease. Third, the SF-36 contains eight dimensions. These dimensions can be used to compute a physical component and mental component score. In the Loup study, also an overall QOL score was calculated. Computing such an overall SF-36 score is not advocated and seems inappropriate because constructs that are conceptually different are mixed. All these factors have potentially hampered the internal validity of the study.I join the authors in their statement that QOL research is needed to design, adapt and optimise the specific needs and complex follow-up of this special group of patients [1]. QOL studies require, however, a firm conceptual background and the use of rigorous research methods.