Liebowitz 社交焦虑量表在社交焦虑障碍评定中的应用

背景在过去20年,越来越多的量表应用于评估社交焦虑障碍。Liebowitz社交焦虑量表(LiebowitzSocialAnxietyScale,LSAS)在临床及流行病学调查中应用广泛,尤其对社交焦虑障碍药物治疗疗效的评估优于其他量表3],同时该量表是唯一一个将社交焦虑障碍的回避症状及害怕症状分开评定的量表。因此我们探讨LSAS在评定社交焦虑障碍患者与正常人群的信度及效度,以及该量表在医生评定及患者自评之间可能存在的差别。方法本研究收录在2002年10月至2003年11月期间于中山大学附属第三医院精神科门诊求诊的未经治疗的社交焦虑障碍患者作为患者组,所有患者符合美国精神障碍诊断和统计手册第四版(DSM-IV)社交焦虑障碍的诊断标准。共55例患者纳入本研究,其中男性40例,女性15例。年龄为18～37岁,平均(26.2±4.8)岁,病程2～12年,平均(5.5±4.3)年,均为社交焦虑障碍的广泛亚型。收录中山大学168名社交恐怖症量表分小于24分的大学生作为正常对照组。其中男92例,女76例,年龄为20～22岁,平均(21±0.8)岁。所有研究对象均填写由作者翻译成中文的LSAS(中文版LSAS)及自编的一般情况量表。在正常对照组中随机抽取31名学生于4周后重测LSAS量表以观其重测信度。在55例患者中随机抽取33例在自己完成量表后和一名精神科医生会晤。该精神科医生就患者的情况使用LSAS量表评分。并比较自评与医评得分的可能差别。结果①患者组及正常对照组的LSAS总分与24条项目的Cronbathα信度系数分别为0.83及0.77;②LSAS总分及各因子分在31名正常对照组4周后的重测信度为0.68～0.79;③LSAS对判别社交焦虑障碍患者与在校大学生的ROC曲线下面积为0.87±0.03,以总分大于35分为界来判断有无社交焦虑障碍,其敏感度为0.77,特异度为0.81;④患者组LSAS总分明显高于正常对照组(66.6±28.3与29.1±17.3比较;P<0.001);⑤对33例患者既行自评又行医评,自评与医评得分无明显差异(64.1±30.1与61.9±23.8比较;P>0.05)。⑥无论在患者组还是在正常组,LSAS的总害怕分均高于总回避分(分别为36.8±14.1与30.0±16.0比较及16.5±9.4与12.6±9.6比较;P<0.001)。结论LSAS具有良好的内部一致性及重测信度,在评定社交焦虑障碍患者与正常人群中有很高的敏感度与特异度,并且我们建议LSAS自评总分大于或等于35分者为可能有社交焦虑障碍。LSAS在患者中的医评与自评结果无明显差异,所以临床上可使用患者自评代替医评以节约时间。我们的结果也显示出患者与正常人群的害怕症状明显均高于回避症状,因此我们认为害怕症状可能先于回避症状出现,并且可能是对社交的害怕(负性思维)导致了回避症状的出现。

The utility of Liebowitz Social Anxiety Scale in the patients with social anxiety disorder in Chinese

Background During the past two decades, a number of rating scales were developed to facilitate diagnosis and assessment of subjects with social anxiety disorder. One of the most commonly used scales for the assessment of social anxiety disorder is the Liebowitz social anxiety scale (LSAS). The LSAS is widely used in epidemiologic investigations and clinical researches,and its assessment in the pharmacotherapy efficacy for social anxiety disorder is superior to any other scale. So we designed this study to explore the validity and reliability of the LSAS in Chinese patients with social anxiety disorder and normal control, and to find the difference of the scores between the patients self -report version and clinician-administered version. Methods Fifty five patients meeting the DSM-Ⅳ diagnostic criteria for social anxiety disorder and 168 normal controls who were screened from 222 college students were rated by LSAS, social phobia scale and self-made General Information Forms. Results The Cronbath α of LSAS for the patients and the normal controls was 0.83 and 0.77, respectively. The 4-week test-retest reliability for total scores and its factors scores of LSAS in 31 normal controls were ranging from 0.68 to 0.79. The ROC area under curve value in discriminating the patients from normal controls was 0.87±0.03; the total score of 35 was considered to be the best cut-off score for LSAS, then its sensitivity was 0.77 and its specificity was 0.81; and no significant difference between the self-report version and clinician-administered version. Conclusions The LSAS is good in internal consistency and test-retest reliability, and has high sensitivity and specificity in discriminating the patients and the controls. There is no significant difference in the total score and each factor scores of LSAS between self-report version and clinician-administered version.