Search

Background: in Occidental languages, no widely accepted questionnaire is available which deals with health related quality of life from the specific point of view of Traditional Chinese Medicine (TCM). Some psychometric tools of this kind are available in Chinese. One of them is the Chinese Quality of Life questionnaire (ChQoL). It comprises 50 items, subdivided in 3 Domains and 13 Facets. The ChQoL was built from scratch on the basis of TCM theory. It is therefore specifically valuable for the TCM practitioner. This paper describes our translation into Italian of the ChQoL, its first application to Occidental oncological patients, and some of its psychometric properties. Methods: a translation scheme, originally inspired by the TRAPD procedure, is developed. This scheme focuses on comprehensibility and clinical usefulness more than on linguistic issues alone. The translated questionnaire is tested on a sample of 203 consecutive female patients with breast cancer. Shapiro-Wilk normality tests, Fligner- Killeen median tests, exploratory Two-step Cluster Analysis, and Tukey's test for non-additivity are applied to study the outcomes. Results: an Italian translation is proposed. It retains the TCM characteristics of the original ChQoL, it is intelligible to Occidental patients who have no previous knowledge of TCM, and it is useful for daily clinical practice. The score distribution is not Normal, and there are floor and ceiling effects. A Visual Analogue Scale is identified as a suitable choice. A 3-point Likert scale can also efficiently describe the data pattern. The original scales show non-additivity, but an Anscombe-Tukey transformation with g = 1.5 recovers additivity at the Domain level. Additivity is enhanced if different g are adopted for different Facets, except in one case. Conclusions: the translated questionnaire can be adopted both as a filing system based on TCM and as a source of outcomes for clinical trials. A Visual Analogue Scale is recommended, but a simpler 3-point Likert scale also suitably fits data. When estimating missing data, and when grouping items within Domain in order to build a summary Domain index, an Anscombe-Tukey transformation should be applied to the raw scores. [ABSTRACT FROM AUTHOR]