Objectives: Adolescents are at high risk of non-suicidal self-injury (NSSI). Currently, there is no clinical assessment tool for adolescent NSSI behaviors measurement with global consistency. The Ottawa Self-injury Inventory (OSI) is considered as a relatively comprehensive assessment tool for NSSI, but the questionnaire is discussed with excessive content and timecostly, which may affect the reliability of the measurement results for adolescent.Thus, this study, based on OSI, aims to revise the assessment tool for adolescent with NSSI that is suitable for both clinically and scientifically, referring to the diagnostic criteria for NSSI in the 5th Diagnostic and Statistical Manual of Mental Disorder (DSM-5).
Methods: This study was led by the Second Xiangya Hospital of Central South University and collaborated with 6 mental health service institutions in China from August to December 2020. Adolescent aged from 12 to 24 years old who had self-injury behavior and met the DSM-5 diagnostic criteria for NSSI were continuously recruited in the psychiatric outpatient department or ward. After clinical diagnosis by an experienced attending psychiatrist or above, the general information and OSI were collected by questionnaires. SPSS 24.0 and AMOS structural equation model statistical softwares were used to conduct item analysis and exploratory factor analysis on the obtained data to complete the revision of the scale. Cronbach's alpha coefficient, split-half reliability, test-retest reliability, and content validity and structure validity were performed to analyze the reliability and validity and confirmatory factor analysis was carried out to test the structure validity for the revised scale.
Results: A total of 234 adolescent with NSSI were enrolled, including 33 (14.1%) males and 201 (85.9%) females with the mean age of (16.2±2.6) years old. The most common clinical diagnoses were depression disorder (57.4%), bipolar disorder (20.9%), adolescent mood disorder (17.1%), etc. Nine items (item 2, 7, 11, 13, 23, 24, 10, 17, 18) in the functional scale of OSI were deleted according to extreme grouping method, correlation analysis, and principal component analysis in exploratory factor analysis. The revised functional scale for NSSI consisted of 15 items. The reliability analysis showed that the Cronbach's alpha coefficients of NSSI thought and behavior frequency, addiction characteristics, and function scales were 0.799, 0.798, and 0.835, respectively, and the split-half coefficients were 0.714, 0.727, and 0.852, respectively. The test-retest coefficients of the latter 2 scales were 0.466 and 0.560, respectively. The correlation coefficient between sub-items and total scores in each part of the scale showed good content validity. The exploratory factor analysis showed that a component was extracted from the frequency of thoughts and behaviors of NSSI, one component was extracted from the addictive characteristics, and three components were extracted from the functional part. The three functional subscales were social influence, external emotion regulation, and internal emotion regulation. The factor load of each item was >0.400.
Conclusions: The revised Chinese version OSI targeted the adolescent patients with mental disorders has relatively ideal reliability and validity. The scale shows high stability, dependability, and a reasonable degree of fit. It is a suitable assessment tool for clinical and scientific research on adolescent with NSSI.
目的: 青少年是出现非自杀性自伤(non-suicidal self-injury，NSSI)行为的高发人群。目前国内外尚无可用于临床，评估青少年NSSI行为一致性好的工具。渥太华自伤调查表(Ottawa Self-injury Inventory，OSI)对自伤行为的评估较全面，但问卷的内容过多，测验时间较长，易影响到青少年群体测量结果的可靠性。本研究根据《精神障碍诊断与统计手册(第5版)》(DSM-5)中NSSI的诊断标准，以OSI为基础，修订适用于临床和科研、符合中国国情青少年精神障碍群体的NSSI评估工具。方法: 由中南大学湘雅二医院牵头，联合全国6家精神卫生服务机构，于2020年8月至12月开展研究。在精神科门诊或病房连续招募有自伤行为、符合DSM-5中NSSI诊断标准的12~24岁青少年精神障碍患者。由经验丰富的主治医师以上职称的精神科医生进行临床诊断后，完成一般情况调查表、OSI的填写和调查。应用SPSS 24.0统计软件和AMOS结构方程模型软件对所得数据进行条目分析和探索性因子分析，修订量表并对修订后的量表进行内部一致性、分半信度、重测信度、内容效度、结构效度等信度和效度分析，对结构效度进行验证性因子分析。结果: 共纳入234例青少年精神障碍患者，男性33例(14.1%)，女性201例(85.9%)；年龄(16.2±2.6)岁；主要的临床诊断为抑郁障碍(57.4%)、双相情感障碍(20.9%)、青少年情绪障碍(17.1%)等。根据极端分组法、相关分析及探索性因子分析中主成分分析法等，删除NSSI功能量表中的9个条目(条目2、7、11、13、23、24、10、17、18)，修订后的NSSI功能量表共由15个条目构成。信度分析结果显示NSSI想法与行为频率、成瘾特征、功能量表的内部一致性系数分别为0.799、0.798、0.835，分半信度系数分别为0.714、0.727、0.852，成瘾特征、功能量表的重测信度系数分别为0.466、0.560。各部分量表分条目得分与总分的相关系数显示内容效度良好；探索性因子分析结果显示NSSI想法与行为频率部分提取到1个成分，成瘾特征部分提取到1个成分，功能部分提取到3个成分，3个功能分量表分别为社会影响、外部情绪调节、内部情绪调节，各个条目的因子负荷量均大于0.400。结论: 修订后的针对青少年精神障碍患者的中文版OSI信度和效度较理想，量表的稳定性、可靠性较高，拟合度较为合理，适合作为研究青少年NSSI行为的临床和科研评估工具。.
Keywords: Ottawa Self-injury Inventory; adolescent; confirmatory factor analysis; non-suicidal self-injury; reliability; validity.