The comparision between two international QOL questionnaire of lung cancer:EORTC QLQ-LC43 and FACT-L / 中国癌症杂志

ABSTRACT

Background and purpose:Quality of life (QOL) questionnaires are tools to assess the quality of life of patients. How to choose a proper questionnaire is the fi rst problem of QOL assessment in clinic research. However, there is few report on it. Our current research compares the data obtained from two world-wide used QOL questionnaire of lung cancer patients EORTC QLQ-LC43 (European Organisation for the Research and Treatment of Cancer Quality of Life Questionnaire-Lung Cancer43) and FACT-L (Functional Assessment of Cancer Treatment-Lung) by various statistic methods in order to show differences between the two questionnaires and to provide some suggestions for choosing questionnaires to assess patients’ QOL in clinic research. Methods:The clinic data were obtained from 125 lung cancer patients by two different international QOL questionnaires of lung cancer patients EORTC QLQ-LC43 and FACT-L, and analyzed by various statistic methods such as Pearson correlation analysis, canonical correlation analysis and multiple regression. In addition, we also compare the general modality such as the frame, item numbers and the contents of the subscales, etc. Results:Of the two QOL there are some parts similar in length, ranking type and time construction, even in items. They all contain items in measuring the physical domain, emotion domain, function domain, society domain and lung domain. Correlations between corresponding subscales of the FACT-L and the EORTC QLQ-LC43 are all signifi cant ranging from r=0.331 for the social domain to r=0.664 for the emotional domain. Canonical correlation analysis for the two sets of subscales revealed there are four signifi cant canonical variables (canonical correlation coeffi cient r=0.87 to 0.26 and overall redundancy about 41% ). The eight EORTC QLQ-LC43 subscales are well represented by the fi ve FACT-L subscales (multiple linear regression, R 2=0.531 to 0.766) and the fi ve FACT-L subscales (except relationship with doctors) are also well represented by EORTC QLQ- LC43 subscales (R 2=0.537 to 0.823). The proportion of the corresponding subscale in the total explanation is 50% to 90%. Conclusion:These two questionnaires cover some common parts, but they measure different contents of QOL. Though have the same name, some domains cann’t be compared directly. The two questionnaires cann’t be alternative and the results of them shouldn’t be compaired directly. Both questionnaires are suited for clinic study (e.g compare the effect of two different theropies). But each has its own characters and should be chosen according to reseach goals.