This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Coens, C. Articles by Aaronson, N. Search for Related Content PubMed PubMed Citation Articles by Coens, C. Articles by Aaronson, N. Related Articles Related Reply Social Bookmarking                            What's this?

Variability and Sample Size Requirements for Health-Related Quality-of-Life Measures: Understanding the Challenges Facing Investigators

European Organisation for Research and Treatment of Cancer Data Center, Brussels, Belgium

Henning Flechtner

Clinic for Psychiatry and Psychotherapy of Children and Adolescents, University of Cologne, Cologne, Germany

Neil Aaronson

Division of Psychosocial Research and Epidemiology, The Netherlands Cancer Institute, Amsterdam, The Netherlands

To the Editor:

We read with interest the recent article by Cheung et al¹ which addressed the important issue of variability and discriminative ability of three commonly used cancer specific health-related quality of life (HRQOL) instruments: 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). Such methodological research is important, and could well lead to better design and conduct of clinical trials where HRQOL is an important end point. Ultimately, Cheung et al concluded that "the global score of the EORTC QLQ-C30 performed less favorably than the FACT-G and FLIC in several aspects." However, we believe that some caution should be used in interpreting these findings.

The Cheung et al article focuses on the discriminative ability of the three HRQOL instruments in relation to Eastern Cooperative Oncology Group (ECOG) performance status. Toward this end, they employed the total scores of the FACT-G and the FLIC, and the global quality of life scale of the QLQ-C30. The total scores of the FACT-G and the FLIC represent a sum score of all items in those questionnaires (27 and 22 items, respectively), and include items assessing physical functioning (the HRQOL domain most closely related to the ECOG performance status measure). Conversely, the global scale of the QLQ-C30 includes only two items, one referring to overall health and the other to overall quality of life. In fact, given that the QLQ-C30 does not yield a total score, the scale that probably would be most appropriate for use in an analysis such as that performed by Cheung et al is the five-item physical functioning scale.² As they have data for this scale as well, it would be of interest to know if it performed more efficiently in relation to the ECOG scale than the global QOL scale.

As Cheung et al point out, the greater variability observed for the QLQ-C30 global QOL scale as compared to the FACT-G and FLIC total scores reflects, in large part, the simple fact that the former scale includes only two items, while the latter two scales include 10 fold that number of items. A scale composed of more items will invariably yield more precise results than a similar scale with fewer items. To date, the EORTC has not used of a total, summary score for the QLQ-C30, largely because of the inherent difficulty in interpreting such a score. Rather, it continues to advocate the use of questionnaire profiles that allow clinicians to identify those specific HRQL areas in which patients are improving or deteriorating over time. However, as noted by Cheung et al, some authors have generated such an overall summary score for the QLQ-C30.³ The EORTC is currently investigating whether such a total score, or more likely several summary scores (eg, physical and psychosocial health) can be empirically justified on the basis of structural equation models applied to large, international data sets.

Ultimately, the choice of questionnaire to be used in assessing the HRQOL of patients with cancer should be based on a careful examination of the content, the specific wording, and the psychometric properties of candidate instruments. Contrary to what is suggested by Cheung et al, the FACT-G, the FLIC, and the QLQ-C30 generally have similar psychometric properties, but differ quite substantially in specific content and question phrasing. We would continue to encourage investigators to review carefully the content of questionnaires in order to select the instrument best suited to answering the specific research questions being posed.

Authors’ Disclosures of Potential Conflicts of Interest

The authors indicated no potential conflicts of interest.

REFERENCES

1. Cheung Y-B, 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[Abstract/Free Full Text]

2. Aaronson NK, Ahmedzai S, Bergman B, et al: The European Organization for Research and Treatment of Cancer QLQ-C30: A quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 85:365-376, 1993[Abstract/Free Full Text]

3. Nordin K, Steel J, Hoffman K, et al: Alternative methods of interpreting quality of life data in advanced gastrointestinal cancer patients. Br J Cancer 85:1265-1272, 2001[CrossRef][Medline]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

Related Reply

• In Reply:
  Yin-Bun Cheung, Cynthia Goh, Julian Thumboo, Kei-Siong Khoo, and Joseph Wee
  JCO 2005 23: 8542-8543 [Full Text]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Coens, C. Articles by Aaronson, N. Search for Related Content PubMed PubMed Citation Articles by Coens, C. Articles by Aaronson, N. Related Articles Related Reply Social Bookmarking                            What's this?