This presentation is part of: Poster Session - Utility Theory; Health Economics; Patient & Physician Preferences; Simulation; Technology Assessment
MEASURING CHANGE IN QUALITY OF LIFE USING PROSPECTIVE VERSUS RETROSPECTIVE MEASURES
Pythia T Nieuwkerk, MA, Marieke S Tollenaar, MA, Frans J Oort, PhD, and Mirjam AG Sprangers, PhD. Academic Medical Center / University of Amsterdam, Medical Psychology, Amsterdam, Netherlands
Purpose: The usual method to measure change in quality of life (QoL) is to compare baseline- with follow-up measurements. However, patients may change the reference value by which they evaluate their QoL over time. This may invalidate longitudinal QoL measurements. A method that aims to measure this change in reference values is the comparison of a retrospective baseline- with a conventional baseline- measurement. Another method for evaluating change in QoL is the retrospective appraisal of the perceived direction and magnitude of change, i.e., transition questions. Our objective was to measure change in QoL according to conventional baseline- and follow-up measurements, retrospective baseline- and follow-up measurements, and retrospective transition questions, and to determine which measure of change in QoL yields strongest associations with clinical measures of change. Methods: HIV infected patients (n=268) completed four MOS-HIV scales at week 0 (baseline measurement) and after 36 weeks (follow-up measurement). At week 36, they additionally completed these scales as retrospective baseline measurement, and they answered four questions about change in QoL (transition questions). Change from week 0 to week 36 in CD4-cell count, body mass index, and plasma HIV viral load served as external criterion measures of change. We compared Pearson correlation coefficients between the three measures of change in QoL and the three clinical criterion measures using an overall Chi-square test. Results: Improvement in QoL according to the retrospective baseline- follow-up measurement was significantly larger than according to conventional baseline- follow-up measurement, as patients evaluated their QoL to be worse on the retrospective- than on the prospective baseline measurement. This may indicate that patients’ reference by which they evaluate their QoL over time had changed, which could invalidate prospective assessments of change. The method for measuring change incorporating a retrospective baseline measurement yielded strongest associations with clinical measures of change. As the other retrospective method, i.e., transition questions, yielded associations similar to the prospective method, this was not likely an artefact of the retrospective method in general. Conclusion: A method for measuring change in QoL that incorporated a retrospective baseline measurement showed strongest associations with change in clinical indicators of health status, suggesting a more valid measurement of change in QoL than a conventional prospective method.
