Stewart Regan W, Tuerk Peter W, Metzger Isha W, Davidson Tatiana M, Young John
Department of Psychology and Behavioral Sciences, Medical University of South Carolina.
Department of Psychology, University of Mississippi.
Psychol Serv. 2016 Feb;13(1):1-9. doi: 10.1037/ser0000069. Epub 2015 Dec 14.
Structured diagnostic interviews are widely considered to be the optimal method of assessing symptoms of posttraumatic stress; however, few clinicians report using structured assessments to guide clinical practice. One commonly cited impediment to these assessment approaches is the amount of time required for test administration and interpretation. Empirically keyed methods to reduce the administration time of structured assessments may be a viable solution to increase the use of standardized and reliable diagnostic tools. Thus, the present research conducted an initial feasibility study using a sample of treatment-seeking military veterans (N = 1,517) to develop a truncated assessment protocol based on the Clinician-Administered Posttraumatic Stress Disorder (PTSD) Scale (CAPS). Decision-tree analysis was utilized to identify a subset of predictor variables among the CAPS items that were most predictive of a diagnosis of PTSD. The algorithm-driven, atheoretical sequence of questions reduced the number of items administered by more than 75% and classified the validation sample at 92% accuracy. These results demonstrated the feasibility of developing a protocol to assess PTSD in a way that imposes little assessment burden while still providing a reliable categorization.
结构化诊断访谈被广泛认为是评估创伤后应激症状的最佳方法；然而，很少有临床医生报告使用结构化评估来指导临床实践。这些评估方法一个常见的障碍是测试实施和解读所需的时间。基于经验的减少结构化评估实施时间的方法可能是增加标准化和可靠诊断工具使用的可行解决方案。因此，本研究以寻求治疗的退伍军人样本（N = 1517）进行了初步可行性研究，以基于临床医生管理的创伤后应激障碍（PTSD）量表（CAPS）制定一个简化评估方案。利用决策树分析在CAPS项目中识别出最能预测PTSD诊断的预测变量子集。由算法驱动的、无理论基础的问题序列将实施的项目数量减少了75%以上，并以92%的准确率对验证样本进行了分类。这些结果证明了制定一种评估PTSD的方案的可行性，该方案在几乎不增加评估负担的同时仍能提供可靠的分类。
