Reply to Commentary by Dr Peterson and Dr Groenvold

Abstract

We agree with Drs. Peterson and Groenvold that use of the overall scores of the McGill Quality of Life Questionnaire (MQOL), the Symptom Dis tress Scale (SDS), and the Enforced Social Dependency Scale (ESDS), when investigating the extent of agreement between terminally ill cancer patients and their family caregivers, cannot eluci date the whole picture of concordance between the two respondents. We recognize that collapsing information on different symptoms or different subscales into an overall score will forfeit the dis tinct characteristics of each symptom and subscale within an instrument. However, the following two reasons led Dr Tang to conduct the analyses for identifying predictors of the extent of agreement between QOL assessments made by terminally ill cancer patients and their family caregivers solely by the absolute differences in overall scores of the three selected instruments rather than analyses for individual symptoms or domains. Firstly, the primary interest of the article [1] is to understand which factors will facilitate or impede the extent of agreement between "QOL" assess ments, over and above individual symptoms or psycho-social-spiritual concerns, made by termi nally ill cancer patients and their family caregivers. There is mounting evidence that, at the end of life, cancer patients suffer from multiple physical symptoms [2] and diverse psycho-social-spiritual distress [3]. The concept of "symptom cluster" further illustrates the phenomenon that different symptoms may intertwine together [4]. Researchers have also indicated that there are strong associa tions among a loss of meaning, depression, anxiety, feeling of being a burden on others, hopelessness, and even desire for death [5-6]. One specific symptom or psycho-social-spiritual concern could not adequately represent the overall suffering that cancer patients experience at the end of life. Quality of life, with its multidimensional characteristics, addresses the aforementioned important experi ences integratively. Correspondingly, achieving the best quality of life for terminally ill cancer patients by control of pain and other symptoms and satis fying patients' psychological, social, and spiritual needs has been recognized as the main goal of end of-life care [7]. In order to identify predictors for the extent of agreement between patients' and family caregiv ers' QOL assessments, the author adopts other scholars' [8-9] conceptualization of QOL as a multidimensional construct, encompassing physi cal function, distress resulting from diseaseand/ or treatment-related symptoms, and psycho social-spiritual concerns. The author therefore chose the SDS, the ESDS, and the MQOL to measure the aforementioned three discrete dimensions of QOL, respectively. The high inter nal consistency (Cronbach's alphas of the MQOL: 0.90 and 0.92 for patients and family caregivers, respectively, of the SDS were 0.850 and 0.846; and of the ESDS were 0.948 and 0.953) and high proportion of significant item-item correlations (SDS: 93.6% and 77% for patients and family caregivers (average correlation coefficient for pa tients and family caregivers: 0.33 and 0.30), respectively; ESDS: 100% and 100% (average correlation coefficient: 0.67 and 0.70); and MQOL: 93.9% and 100% (average correlation coefficient: 0.40 and 0.47)) of the three instruments indicate that items on the SDS, the ESDS and the MQOL all measure unique dimensions, namely symptom distress, functional dependency, and psycho-so cial-spiritual concerns.