- 學位論文
PTSD症狀量表之信效度分析：以高雄氣爆受災民眾為例
The Reliability and Validity Analysis of PTSD Symptom Scale: A Case Study on People Affected by Kaohsiung Gas Explosion
摘要
綜觀過去研究，創傷後壓力症(Posttraumatic Stress Disorder, PTSD)是災後常見的精神疾病之一，但目前國內尚未有根據精神疾病診斷與統計手冊第五版(Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, DSM-5)之PTSD診斷準則編修量表的期刊發行，且各學者對於PTSD症狀的最佳因素結構未達成明確共識。故本研究欲發展符合DSM-5的PTSD診斷準則之症狀篩檢量表，以及對PTSD症狀的因素結構進行再驗證。 經歷2014年高雄氣爆災難事件後，對於民眾的身心健康帶來很大的影響，並產生許多創傷後壓力症狀或情緒困擾。本研究對象為1030位高雄氣爆受災民眾，於氣爆發生兩個月後填寫PTSD症狀量表和病人健康狀況問卷，並在氣爆一年後，追蹤487位受災民眾，同樣填寫PTSD症狀量表和病人健康狀況問卷。經由驗證性因素分析結果顯示，PTSD症狀量表可根據DSM-5的PTSD四大類症狀（如侵入性、逃避、認知和情緒的負面改變、警醒性與反應性的顯著改變），來建構四因素與高階一因素模型，且模型的適配度皆在可接受範圍內，並具有良好的內部一致性信度，全量表Cronbach's α係數為.95，四構念Cronbach's α係數為.87至.90，亦有良好的效標關聯效度，PTSD症狀量表及四構念，均與憂鬱(r = .27至.72, p < .01)如預期呈顯著正相關。本研究發現，雖DSM-5四因素模型的適配度不錯，但PTSD症狀的因素結構，由侵入性、逃避、認知和情緒的負面改變、煩躁不安性警醒與焦慮性警醒共五因素所構成的模型之適配度，相較於DSM-5四因素模型更佳。 整體來說，首先，PTSD症狀量表具有良好的信效度，可作為臨床上初步篩檢PTSD症狀及嚴重程度的工具。其次，本研究的因素結構檢驗，顯示PTSD症狀可由五因素結構所代表，其與DSM-5四因素模型之差異，將警醒性與反應性的顯著改變區分為煩躁不安性警醒、焦慮性警醒等兩類症狀，而透過對PTSD症狀的潛在結構有進一步之了解，有助於提供更適當的治療方法。
並列摘要
Posttraumatic Stress Disorder (PTSD), according to previous studies, is a typical mental disorder from which victims of disasters commonly suffer. Nonetheless, there are insufficient journal papers where mental scales are developed by following the diagnostic criteria in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5). Additionally, scholars have not reached a consensus about the best factor structure in the PTSD symptoms. Consequently, the present study aims at developing a screening scale which corresponds to the diagnostic criteria of PTSD symptoms in DSM-5 and focuses on reverifying the factor structure of PTSD symptoms. The 2014 Kaohsiung gas explosion had an enormous impact on residents’ health both physically and mentally, causing posttraumatic stress symptoms and emotional distress. The objects of the present study are 1030 victims who completed PSTD Symptom Scale and Patient Health Questionnaire-9 (PHQ-9) two months after the event, and 487 of the 1030 study objects were followed up with PTSD Symptom Scale and PHQ-9. The results of confirmatory factor analysis suggest that the four-factor and a high-order factor models in the PTSD Symptom Scale can be designed on the basis of the four categories of PTSD symptoms in DSM-5 (e.g., intrusion, avoidance, negative alterations in cognitions and mood, and marked alterations in arousal and reactivity), and the model achieves acceptable fit with great internal consistency reliability. The Cronbach's alpha of the scale is .95, and that of the four constructs is between .87 to .90. In addition, the scale has great criterion related validity. Furthermore, as expected, the PTSD Symptom Scale and the four constructs show the positive correlation with depression (r = .27 to .72, p < .01). On the basis of my research, though the DSM-5 four-factor model shows good fit, with the five factors of PTSD symptom structure: intrusion, avoidance, negative alterations in cognitions and mood, dysphoric arousal, and anxious arousal, the new five-factor model demonstrates better fit than DSM-5 four-factor model. Two conclusions have been made in the present study. Firstly, the present study concluded that PTSD Symptom Scale provides evidences concerning good reliability and validity, PTSD Symptom Scale can thus be applied to preliminarily examine the PTSD symptoms and their severity clinically. Secondly, based on the examination of the factor structure in the present study, PTSD symptoms can be represented by five-factor structure, which splits the one of the factor in DSM-5 four-factor structure “marked alterations in arousal and reactivity” into two factors, that is, “dysphoric arousal” and “anxious arousal”, and the present study is beneficial to provide more suitable treatments through the exploration of the PTSD latent structure.