精神分裂症复发(恶化)征兆问卷(家属版)的初步编制

Primary Development and Evaluation on Early Signs Inventory of Relapse/Worsen in Schizophrenia (Care Version)

目的初步编制精神分裂症复发（恶化）征兆问卷（early signs inventory of relapse／worsen in schizophrenia，ESIS）。方法通过访谈和文献复习方法，归纳总结精神分裂症复发（恶化）征兆的内容和维度，编制问卷初稿。分别对200例和100例精神分裂症患者家属进行探索性因素分析和信度分析。结果探索性因素分析抽取6个因素并对其命名，形成28题的初始问卷。各条目得分与总问卷得分间的相关系数为0．150～0．603，各条目得分与其所属因子得分间的相关系数为0．483～0．811。各因子得分间的相关系数为0．024—0．469。各因子得分与总问卷得分问相关系数为0．164～0．820。比较精神分裂症患者缓解期和复发时问卷结果，除躯体不适感因子得分两者之问差异无统计学意义外（P〉0．05），两者间问卷总分、各因子得分差异均具统计学意义（P〈0．01）。总问卷的标准Cronbach α系数为0．845，各因子的Cronbach α系数为0．590～0．779。总问卷的Spearman—Brown分半信度为0．719，各因子的Spearman—Brown分半信度为0．500～0．772。总问卷的重测相关系数为0．721，各因子的重测相关系数为0．484～0．900，相关系数有统计学意义（P〈0．01）。结论精神分裂症复发（恶化）征兆问卷信效度特性基本符合心理测量学要求，可初步用于临床指导家属评估患者复发（恶化）征兆。

Objective To develop and evaluate early signs Inventory of relapse/worsen in schizophrenia（ESIS）. Methods The primary version of ESIS edited by interviewing family members of schizophrenics and summarizing literatures, and 200 patients＇family members were measured formally and developed the formal inventory referring to the results of exploratory factor analysis. 50 stable and 50 relapse/worsen patients were investigated to evaluate the validity and reliability of the inventory. Results Exploratory factor analysis reveals that drawing-out 6 factors were termed and formed a 28-question inventory, which could explain 52. 306% of total variation. The correlations between item and total scale were 0. 150 - 0. 603 （ P 〈 0.05 ）. The correlations between item and subscale were 0. 483 - 0. 811（P 〈0.01 ）. The score differences between stable and relapse/worsen patients except subscale 1, the total scale and another 5 subseales were significant（P 〈 0.01 ）. The total Cronbach α was 0. 845 and the Cronbach α of each subscale was 0. 590 - 0. 779. The split-half reliability of ESIS was 0. 719, while the split-haft reliabilities of subscales was 0. 501 - 0. 772. The test-retest correlation of the whole scale was 0. 721 and subscales＇ test-retest correlation coefficients were 0. 484 - 0. 900 （ P 〈 0.01 ）. ConcluSion The reliability and validity of ESIS could reach the acceptable level in psychometrics, and this inventory may be used to instruct family members to identify early signs of relapse/worsen of patients.