Medical School of Ribeirão Preto, São Paulo University, Ribeirão Preto, Brazil.
National Institute For Science and Technology (INCT-TM, CNPq), Brasília, Brazil.
Psychiatry Clin Neurosci. 2019 Dec;73(12):754-760. doi: 10.1111/pcn.12931. Epub 2019 Oct 14.
The Structured Clinical Interview for the DSM is one of the most used diagnostic instruments in clinical research worldwide. The current Clinician Version of the instrument (SCID-5-CV) has not yet been assessed in respect to its psychometric qualities. We aimed to assess the clinical validity and different reliability indicators (interrater test-retest, joint interview, face-to-face vs telephone application) of the SCID-5-CV in a large sample of 180 non-prototypical and psychiatric patients based on interviews conducted by raters with different levels of clinical experience.
The SCID-5-CV was administered face-to-face and by telephone by 12 psychiatrists/psychologists who took turns as raters and observers. Clinical diagnoses were established according to DSM-5 criteria and the longitudinal, expert, all data (LEAD) procedure. We calculated the percentage of agreement, diagnostic sensitivity and specificity, and the level of agreement (kappa) for diagnostic categories and specific diagnoses.
The percentage of positive agreement between the interview and clinical diagnoses ranged between 73% and 97% and the diagnostic sensitivity/specificity were >0.70. In the joint interview, the levels of positive agreement were high (>75%) and kappa levels were >0.70 for most diagnoses. The values were less expressive, but still adequate, for interrater test-retest interviews.
The SCID-5-CV presented excellent reliability and high specificity as assessed with different methods. The clinical validity of the instrument was also confirmed, which supports its use in daily clinical practice. We highlight the adequacy of the instrument to be used via telephone and the need for careful use by professionals with little experience in psychiatric clinical practice.
《精神疾病诊断与统计手册》结构式临床访谈是全球范围内临床研究中使用最广泛的诊断工具之一。目前该工具的临床医师版（SCID-5-CV）尚未评估其心理测量学质量。我们旨在评估 SCID-5-CV 在一个由 180 名非典型和精神科患者组成的大样本中的临床有效性和不同的可靠性指标（评估者间测试-重测、联合访谈、面对面与电话应用），这些患者的访谈由具有不同临床经验的评估者进行。
12 名精神科医生/心理学家通过面对面和电话使用 SCID-5-CV 进行访谈，他们轮流担任评估者和观察员。临床诊断根据 DSM-5 标准和纵向、专家、所有数据（LEAD）程序确定。我们计算了诊断类别的阳性一致性百分比、诊断敏感性和特异性，以及诊断的一致性（kappa）水平。
访谈和临床诊断之间的阳性一致性百分比在 73%至 97%之间，诊断敏感性/特异性>0.70。在联合访谈中，阳性一致性水平较高（>75%），kappa 水平>0.70，适用于大多数诊断。评估者间测试-重测访谈的结果虽然表达力较低，但仍足够充分。
SCID-5-CV 具有良好的可靠性和高度的特异性，这是通过不同方法评估的。该工具的临床有效性也得到了证实，支持其在日常临床实践中的使用。我们强调了该工具通过电话使用的适当性，以及经验不足的精神科临床医生谨慎使用的必要性。
