诈病症状定式访谈中文版的信效度

Reliability and validity of the Chinese version of the Structured Interview of Reported Symptoms

目的:检验诈病症状定式访谈(SIRS)中文版的信度和效度,为我国司法鉴定和诈病临床评估应用提供一个有效的测评工具。方法:采用仿真模拟设计和已知团体组设计两种方法。仿真模拟设计中,选择按国际疾病与相关健康问题分类第十版(ICD-10)诊断为精神分裂症、情感障碍及其他精神障的住院患者共40名(普通精神病例组)及正常大学生80名(模拟诈病组40名,正常诚实组40名);已知团体组设计中,选择按精神障碍诊断与统计手册第4版(DSM-IV)诊断为精神分裂症、情感障碍及其他精神障的司法鉴定被试共80名(非诈病组)及可疑诈病者20名(可疑诈病组)。由1名接受过培训的精神科专业人员使用SIRS中文版进行评估,并对其中80名正常大学生同时评定明尼苏达多相人格调查表(MMPI-2)。采用相关分析和Cronbachα系数检验量表内部一致性信度。以MMPI-2中的说谎(L)、稀有回答(F)、校正/防卫(K)3个效度因子为效标。以不同组SIRS中文版的得分差异检验量表的实证效度。结果:SIRS中文版的Cronbachα系数为0.66~0.98。各分量表之间的相关系数在0.47~0.85(P<0.05)。SIRS中文版的诈病维度(不一致的症状与夸大症状)得分与MMPI-2的F分呈正相关(r=0.33~0.48,P<0.05),与L和K分呈负相关(r=-0.23^-0.52,P<0.05)。仿真模拟设计中,除SIRS中文版的评价诚实度的症状、防御性症状、过于特定症状3个因子外,模拟诈病组SIRS中文版10个因子得分均高于正常诚实组和普通精神病人组[如罕见症状因子分(10.2±3.3)vs.(0.7±1.3),(2.4±1.9);P<0.05)];已知团体组设计中,可疑诈病组SIRS中文版13个因子得分均高于非诈病组[如罕见症状(7.4±1.7)vs.(2.8±2.6),P<0.05)]。SIRS中文版在仿真模拟设计和已知团体组设计中的敏感度分别为20%和85%,特异度分别为83.75%和87.5%。结论:诈病症状定式访谈中文版具有良好的信度和效度,可为诈病的诊断提供重要依据。

Objective：To examine the reliability and validity of the Chinese version of the Structured Interview of Reported Symptoms （SIRS-CV）.Methods：In simulation design,40 patients with schizophrenia,mood disorder,and other psychiatric disorders,80 normal college students[honest group （n =40） and simulators （n =40）],and in known-group design,100 patients in a forensic clinic[suspected malingering group （n =20） and non-malingering group （n =80）] were investigated with the SIRS-CV.Eighty normal college students were also assessed with the Minnesota Multiphasic Personality Inventory-2 （MMPI-2）.The consistency of the scale was tested by using correlation analysis and Cronbach＇s α coefficient.The criterion was the three validity scales[Lie （L）,Infrequency （F）,Correction （K）] of the MMPI-2.The score of the Chinese version of SIRS difference in different groups was used to test the validate validity.Results：The Cronbach α coefficient were 0.66-0.98.There was significant positive correlation between the subscales （r =0.47-0.85,P 〈0.05）.The malingering dimension （including 8 factor） scores of SIRS-CV were positively correlated with the ＂F＂ scores of MMPI-2 （r =0.33-0.48,Ps 〈 0.05）,and negatively correlated with the ＂L＂ and ＂K＂ scores of MMPI-2 （r =-0.23--0.52,P 〈0.05）.In addition to the direct appraisal of honesty,defensive symptoms,overly specific symptoms of these three factors,the 10 factor scores of SIRS-CV in the analogue group were higher than normal group and general psychiatric patient groups[i.e.,＂rare symptoms＂,（10.2 ± 3.3） vs.（0.7 ± 1.3）,（2.4 ± 1.9）,P 〈 0.05）].The scale factor scores of SIRS-CV in the suspected malingering group scores were higher than those in the malingering group[i.e.,＂ rare symptoms＂,（7.4 ± 1.7） vs.（2.8 ±2.6）,P 〈 0.05）].The sensitivity in simulation design and known-group design were 20% and 85%,and the specificity were 83.75% and 87.5%,respectively.Conclusion：The Chinese version of the Structured Interview of Reported Symptoms is a reliable and valid measure for identification of malingering in actual work.