Response shift in the assessment of anxiety, depression and perceived health in urologic cancer patients: an individual perspective.

The assessment of quality of life in cancer patients is hampered because patients may change their frames of reference during the course of the disease. The aim of this study was to test individual differences in these response shift effects. Urologic cancer patients (n= 275) were examined during the stay in the hospital (T1), 2 weeks later (T2) and 3 months later (T3). Furthermore, at T3 they were asked to retrospectively assess their situation at T1 (then-test). The difference between this retrospective assessment and the original assessment at T1 was used to determine the response shift effect (recalibration). Anxiety (Generalized Anxiety Disorder Questionnaire-2), depression (Patient Health Questionnaire-2) and health dissatisfaction (Questionnaire on Life Satisfaction) were obtained at all points in time. The effect sizes of the mean response shift effects (recalibration) ranged between 0.26 and 0.48. Nevertheless, a large proportion of the sample showed no response shift (22-38%) or even negative response shift effects (20-30%). There were significant correlations among the response shift measures of the domains (anxiety, depression and health dissatisfaction) with coefficients between 0.29 and 0.51. The results indicate that response shift should not only be assessed on the mean score level, since it is also a dimension of individual difference.