Prostate cancer patients' quality of life assessments across the primary treatment trajectory: 'True' change or response shift?

• Published in: Acta oncologica 2016-07, Vol.55 (7), p.814-820
• Main Authors: Gerlich, Christian, Schuler, Michael, Jelitte, Matthias, Neuderth, Silke, Flentje, Michael, Graefen, Markus, Krüger, Alexander, Mehnert, Anja, Faller, Hermann
• Format: Article
• Language: English
• Subjects: Aged; Cognition; Emotions; Follow-Up Studies; Humans; Male; Middle Aged; Models, Biological; Prostatic Neoplasms - psychology; Prostatic Neoplasms - therapy; Quality of Life; Self Report; Socioeconomic Factors; Surveys and Questionnaires
• ISSN: 0284-186X; 1651-226X
• DOI: 10.3109/0284186x.2015.1136749

Background Self-report questionnaires are widely used to assess changes in quality of life (QoL) during the course of cancer treatment. However, comparing baseline scores to follow-up scores is only justified if patients' internal measurement standards have not changed over time, that is, no response shift occurred. We aimed to examine response shift in terms of reconceptualization, reprioritization and recalibration among prostate cancer patients. Material and methods We included 402 newly diagnosed patients (mean age 65 years) and assessed QoL at the beginning of cancer treatment and three months later. QoL was measured with the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30). We employed structural equation modeling testing measurement invariance between occasions to disentangle 'true' change and change in the measurement model (response shift). Results We found reprioritization effects for both the Physical Functioning and Role Functioning subscales of the EORTC QLQ-C30, indicating that both had gained importance for representing the latent construct of QoL at follow-up. These effects added to the worsening effect evident in the latent construct, thus rendering observed changes even more pronounced. In addition, we found recalibration effects for both the Emotional Functioning and Cognitive Functioning subscales indicating judgments becoming more lenient over time. These effects counteracted 'true' negative changes thus obscuring any substantial changes on the observed level. Conclusion Our results suggest that changes observed in some subscales of the EORTC QLQ-C30 should not be taken at face value as they may be affected by patients' changed measurement standards.