|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 33 (November 20), 2005: pp. 8542-8543 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.04.0501
In Reply:MRC Tropical Epidemiology Group, London School of Hygiene & Tropical Medicine, London, United Kingdom
Department of Palliative Medicine, National Cancer Centre, Singapore
Department of Rheumatology and Immunology, Singapore General Hospital, Singapore
Department of Medical Oncology, National Cancer Centre, Singapore
Division of Clinical Trials & Epidemiological Sciences, National Cancer Centre, Singapore Coens et al rightly emphasize that the results of our recent article1 on the variability and sample size requirements for health-related quality-of-life questionnaires should be interpreted with some caution. In particular, the contents and wording of the three questionnaires, namely the Functional Assessment of Cancer Therapy-General (FACT-G), the Functional Living Index–Cancer (FLIC), and the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-C30 (EORTC QLQ-C30), are not identical and researchers need to carefully examine them before making a choice of instrument. We have put forward this cautionary note in our article, but it is helpful to emphasize and reiterate this. Coens et al suggest that "given that the QLQ-C30 does not yield a total score, the scale that probably would be most appropriate for use in analysis such as that performed by Cheung et al is the 5-item physical functioning scale." We have reservations about this view. The contents of the FACT-G and FLIC are clearly much more heterogeneous than that of the physical functioning scale (PFS) of the QLQ-C30. So a comparison between the PFS and the total scores of the FACT-G and FLIC are not useful. The FLIC only gives a total score. The FACT-G has a physical well-being and a functional well-being scale. However, careful examination of their contents would show that the concepts they measure are not comparable with that of PFS. So a scale-specific comparison is not appropriate either. We believe the most useful comparison of the three instruments should be based on their total or global scores, not only because they are often the basis for sample size calculation, but also because all of them are supposed to reflect the general level of patients’ well being. Nonetheless, it may be helpful to provide more information about the PFS. Using the same definitions, methods, and data reported in Cheung et al,1 we found that the coefficient of variation (95% CI) of PFS was 0.17 (95% CI, 0.16 to 0.18), its effect size for detecting a difference between Eastern Cooperative Oncology Group (ECOG) performance status was 1.33 (95% CI, 1.53 to 1.14), its effect size for detecting a change in ECOG performance status was 0.11 (95% CI, –0.02 to 0.23), and its correlation between pre- and post-test was 0.43 (95% CI, 0.34 to 0.52). Comparing these findings against those reported in Tables 2 to 5 in Cheung et al, it can be seen that the PFS performed better than the QLQ-C30 global scale cross-sectionally, but worse longitudinally according to these yardsticks. We are delighted to hear that the EORTC quality-of-life team is investigating whether a total score on the QLQ-C30 could be developed. We look forward to seeing the outcome. Authors’ Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. REFERENCE
1. Cheung YB, Goh C, Thumboo J, et al: Variability and sample size requirements of quality-of-life measures: A randomized study of three major questionnaires. J Clin Oncol 23:4936-4944, 2005
Related Correspondence
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
