Psychometric properties of the Chinese version of the Center for Epidemiologic Studies Depression Scale-Revised in patients with cancer: A cross-sectional study

Background: Depressive symptoms are common in patients with cancer and more prevalent in Chinese patients. The Center for Epidemiologic Studies Depression Scale is one of the most widely used self-report scales to assess depressive symptoms in both community and hospitalized samples. A revised Center for Epidemiologic Studies Depression Scale has been created, but the evidence on psychometric properties is limited. Objectives: To develop the Chinese version of the scale, and to examine the cross-cultural validity, structural validity, construct validity, internal consistency, test-retest reliability, measurement error, responsiveness, and floor/ceiling effect of the scale among patients with cancer according to the recommendation in the consensus-based standards for the selection of health status measurement instruments checklist. Design: A cross-sectional survey design with 33 participants (approximately 10%) completing the followup survey for evaluating the test-retest reliability. Settings: Randomly selected eight wards of an oncology hospital in China. Participants: Of the 595 patients we approached in the randomly selected eight wards, 310 gave their informed consent and completed the survey. Methods: The Chinese version of the Center for Epidemiologic Studies Depression Scale-Revised was developed by four researchers (two translators and two reviewers) who were proficient in both English and Chinese. Participants completed the scale and the depression module of the Patient Health Questionnaire. Principal components analyses, Spearman's correlation, the Mann-Whitney U test, Cronbach's alpha, and the intraclass correlation coefficient were used. Results: The cross-culture validity was excellent based on the consensus-based standards for the selection of the health status measurement instruments checklist. A two-factor structure was determined: somatic symptoms and affective-cognitive symptoms. The sufficient construct validity was supported by that the score of the Center for Epidemiologic Studies Depression Scale-Revised was strongly correlated with the depression module of the Patient Health Questionnaire score (rho = 0.73, p < 0.001) and patients who had received chemotherapy (p = 0.002 )/radiotherapy (p = 0.035) reported higher scores of depression than those who have not. The Cronbach's alphas of the total scale and subscales ranged from 0.82-0.88. The test-retest reliability was sufficient (intraclass correlation coefficient = 0.73-0.81, all p < 0.001) for total scale and subscales. The smallest detectable change was 2.98 and the responsiveness was adequate, with no floor/ceiling effect. Conclusions: This study supports the Chinese version of the Center for Epidemiologic Studies Depression Scale-Revised as a valid and reliable measurement of depressive symptoms in patients with cancer. (C) 2019 Elsevier Ltd. All rights reserved.