Timing of quality of life (QoL) assessments as a source of error in oncological trials Aim of the study. To produce an empirical estimate of the nature and magnitude of the error produced by incorrect timing quality of life (QoL) measurements in patients receiving chemotherapy.
Design. In a multicentre trial, 283 patients were randomized to receive either docetaxel (T) or sequential methotrexate and 5-fluorouracil (MF). The QoL was assessed at baseline and before each treatment using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30). The study design was retrospective. Data were analysed using t-tests.
Results. Erroneous timing affected the QoL findings in both treatment arms. At baseline, there were statistically significant differences in the MF group on the nausea/vomiting scale, with ill-timed assessment showing more symptoms, and in the T group on the physical functioning scale with ill-timed assessments indicating better QoL. The mean scores of correct vs. incorrect timings over the first 14 cycles showed statistically significant differences on several scales. In the MF group, ill-timed assessments indicated significantly worse physical functioning and global QoL, and significantly more of the following symptoms: fatigue, nausea/vomiting, insomnia, appetite loss, and constipation. In the T group, ill-timed assessment showed better physical functioning, less dyspnoea and more insomnia than correctly timed assessments. The reasons for erroneous timing were not always detectable retrospectively. However, in some cases the MF group, being in standard treatment, seemed to have followed a clinical routine not involving the active participation of the study nurse responsible, whereas patients in the experimental T group were more consistently taken care of by the study nurses.
Conclusions. Incorrect timing of QoL assessments in oncological trials jeopardises both the reliability of the QoL findings within treatment and the validity of QoL outcome comparisons between treatments. This issue should be emphasized in the planning of both the study design and clinical routines.