Quantitative assessment of response shift in QOL research

Abstract

Two assumptions which characterize the measurements of quality life (QOL) studies, are questioned by the present report: First, the assumption that QOL is the same thing for all subjects, is usually related to health problems and can easily be measured across subjects. Second, when this assumption is given up by introducing subjective base-rates in an intra-subject (before-after) design, researchers still tend to assume that no response shifts occur on the QOL scales, e.g., that these scales maintain the same meaning and values for subjects over time and interventions. In our study we found that QOL had different subjective interpretations [1--4]. While for some certain issues in their family-life determined their quality of life, for others these were issues at their workplace or of their health conditions. Second, when QOL was tested against a subjective base-line (for each individual according to their choices of domains), about eleven percent of the hypertensives and normotensives showed a clear response shift of scale-calibration over the period of one year. When the scores of these subjects were excluded, the significance of certain previously reported results changed. For example, the significant difference between normotensives and hypertensives concerning the change in their subjective evaluation of QOL over the year and their initial depression became more significant, while similar changes in their evaluation of sexual impairment and control at their work-place became insignificant. These results suggest that response shifts have to be traced and quantified, before one can claim any results (or lack of results) in 'before-after' designs, concerning subjective meaningful issues like quality of life.