Development and reliability-validity testing of Chinese urban version of perinatal depression screening scale
10.3760/cma.j.cn113903-20230824-00151
- VernacularTitle:中国城市版围产期抑郁筛查量表的编制及信效度检验
- Author:
Shan ZHANG
1
;
Xueyun GAO
;
Meina XU
;
Jing LI
;
Chengyin DOU
;
Qinghong LI
;
Dan ZHANG
;
Rufang CHEN
;
Shan WANG
;
Yang MI
;
Zhongliang ZHU
;
Hui LI
Author Information
1. 西安交通大学第一附属医院新生儿科，西安 710061
- Publication Type:Journal Article
- Keywords:
Depression, perinatal period;
Depression assessment scale;
Influencing factor analysis;
Reproducibility of results
- From:
Chinese Journal of Perinatal Medicine
2024;27(12):1042-1048
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To develop a perinatal depression screening scale for Chinese urban version and test its reliability and validity.Methods:Naturally conceived women who were≥12 weeks of gestation and had regular prenatal examinations or delivered within four weeks in Xi'an and Foshan were recruited using convenience sampling. Initial entries were created through clinical interviews and expert validation of 50 of these pregnancies. Further item screening and testing for reliability and validity using correlation analysis, critical ratio method, homogeneity test, and factor analysis were performed. The final questionnaire was formed and then used to test the subjects. Fifty women who had completed the questionnaire effectively were selected for retesting 10 to 15 days after the initial evaluation. A structural equation model was constructed using the Edinburgh Postpartum Depression Scale (EPDS) as the reference criterion to conduct the scale's exploratory and confirmatory factor analysis. The receiver operating characters curve was used to determine the cut-off point of the scale.Results:(1) 1 300 questionnaires were issued, and 1 049 valid questionnaires were received with a recovery rate of 80.7%. Among them, 601 were used for exploratory and correlation analysis, and 448 were used for confirmatory factor analysis. (2) The initial entries of the scale totaled 63. After eliminating indicators that did not meet the requirements for item analysis, the final scale consisted of 22 entries in total. The final scale consisted of six dimensions, including "abnormal mood", "loss of interest and pleasure," "sleep disorder," "self-blame and guilt," "decline in conscious thought ability," and "lack of energy". (3) The correlation coefficients between the above six dimensions and EPDS were 0.609, 0.322, 0.423, 0.522, 0.545, and 0.516, respectively. The one between the total scale and EPDS was 0.715 (all P<0.01). (4) The confirmatory factor analysis of each model and the scale showed an acceptable fit and a stable factor structure with indices Chi-square/degrees of freedom =2.504, square root of approximation error=0.058, standardized root mean square residual=0.048, comparative fit index=0.925, Tueker-Lewis index=0.911. After retaining 22 items, the Cronbach's coefficients of the six dimensions ranged from 0.708 to 0.870. (5) The test-retest reliability of the total scale, six dimensions, and 22 items were 0.936, 0.786-0.846, and 0.720-0.886, respectively. (6) The receiver operating characteristic curve of the scale was drawn with EPDS≥13 as the cut-off value, and the area under the curve (AUC) of EPDS was 0.808 (95% CI: 0.744-0.872). When the original cut-off point was 21, the Youden index was the max (0.561), with sensitivity and specificity of 0.952 and 0.609, respectively. The Kappa value was 0.691, indicating that the scale was consistent with EPDS. Conclusion:This study preliminarily established a Chinese urban version of the perinatal depression screening scale with a stable factor structure and good reliability and validity.
