精神分裂症焦虑症状评估量表的初步编制和信效度研究

Development of anxiety evaluation scale in schizophrenia and its reliability and validity

关键词:  精神分裂症  焦虑  信度  效度

Key words:  schizophrenia  anxiety  reliability  validity

摘要:

目的编制适合中国精神分裂症患者的焦虑症状评估量表，并进行信度和效度检验。方法以29项精神分裂症患者中常见的焦虑症状作为初测量表，对上海市精神卫生中心的154例精神分裂症住院患者进行评估，间隔2周后对其中50例患者进行重测。使用项目分析及探索性因素分析初步建构精神分裂症焦虑症状评估量表，并进行验证性因子分析。使用Pearson相关系数计算各条目、各因子与总分的相关以及检验精神分裂症焦虑症状评估量表与HAMA和PANSS量表间的校标关联效度并检验重测信度。通过t检验对量表的区分效度进行检验。用组内相关法(intraclass correlations coefficient, ICC)计算评估者间一致性信度；Cronbach’s α系数计算内部一致性信度。结果精神分裂症焦虑症状评估量表包括20个条目，探索性因素分析得到表现和感知的焦虑、对周围环境的焦虑、神经症样症状、激越、躯体性焦虑5个因子，解释总方差的65.56%，验证性因子分析证实该模型（χ2/df=2.128,GFI=0.829，CFI=0.876，RMSEA=0.086）拟合度较好。各条目与总分的相关系数为0.33~0.81，各因素与量表总分的相关系数为0.54~0.91，均呈显著相关（P<0.01），具有较好的结构效度。量表总分与HAMA及其躯体性焦虑因子、精神性焦虑因子、PANSS量表焦虑抑郁因子均显著相关（P<0.01）,具有较好的校标关联效度。无焦虑症状组和焦虑症状组在量表各维度上及总分上的差异均具有统计学意义(P<0.01)，说明量表均有较好的区分效度。评定者间一致性系数为0.98(P<0.01)；Cronbach’s α系数均>0.6，具有较好的内部一致性；间隔2周的重测相关系数在0.729~0.967（P<0.01），具有较好的重测信度。结论精神分裂症焦虑症状评估量表具有较好的信度和效度，可用于中国精神分裂症患者焦虑症状的筛查和评估。

Abstract:

ObjectiveTo develop an anxiety evaluation scale in schizophrenia and to assess its reliability and validity. MethodsUsing a 29-item scale of anxiety symptoms which were commonly used in schizophrenia as the initial survey instrument, 154 schizophrenia patients in Shanghai Mental Health Center were assessed and 50 patients of them were retested after 2 weeks. By using item analysis and exploratory factor analysis, the preliminary theoretical framework of the scale was established and the confirmatory factor analysis was further conducted. Pearson correlation was used to calculate the correlation coefficient of each item, each factor score and total score; the criterion validity of the scale was tested with PANSS and HAMA and test-retest reliability was calculated. The discriminant validity of the scale was determined with t-test. The intraclass correlations coefficient (ICC) was calculated to estimate inter-rater reliability and the Cronbach’s α coefficient was employed to evaluate the internal consistency reliability. ResultsThe anxiety evaluation scale in schizophrenia included 20 items, and exploratory factor analysis revealed five factors: “expressed and perceived anxiety”, “anxiety and environment”, “neurotic symptoms”, “agitation” and “somatic anxiety” accounted for 65.56% of total variances. Confirmatory factor analysis confirmed that the model fitted well（χ2/df=2.128, GFI=0.829，CFI=0.876，RMSEA=0.086）. The correlation coefficient between each item and the total score was 0.33-0.81 and correlation coefficient between each factor score and total score of the scale was 0.54-0.91 with good structural validity (P<0.01). The total score of the scale was significantly correlated with HAMA total score, HAMA-somatic anxiety factors score, HAMA-psychiatric anxiety factors score and anxiety and depression factors scores in PANSS (P<0.01), showing a good criterion validity. Non-anxiety group and anxiety group had statistical differences in each scale dimensions and total score (P<0.01). The inter-rater reliability of the total score was 0.98 (P<0.01); Cronbach’s α coefficients were all greater than 0.6, indicating good internal consistency；the test-retest correlation coefficients at 2-week intervals was 0.729-0.967（P<0.01）, showing a good test-retest reliability. ConclusionThe anxiety symptom scale of schizophrenia has good reliability and validity, it can be used to screen and evaluate the anxiety symptoms of schizophrenia patients in China.