Frazier Lorraine, Sanner Jennifer, Yu Erica, Cron Stanley G, Moeller F Gerard
Lorraine Frazier, PhD, RN, MS, FAHA, FAAN Dean and Professor, College of Nursing, University of Arkansas for Medical Sciences, Little Rock. Jennifer Sanner, PhD, RN Assistant Professor, School of Nursing, Department of Nursing Systems, The University of Texas Health Science Center at Houston. Erica Yu, PhD, RN, ANP Assistant Professor, School of Nursing, Department of Acute & Continuing Care, The University of Texas Health Science Center at Houston. Stanley G. Cron, MSPH Research Instructor, School of Nursing, Center for Nursing Research, The University of Texas Health Science Center at Houston. F. Gerard Moeller, MD Professor of Psychiatry and Director, Center for Neurobehavioral Research on Addictions, The University of Texas Health Science Center at Houston.
J Cardiovasc Nurs. 2014 Jul;29(4):347-53. doi: 10.1097/JCN.0b013e318291ee16.
Despite the prevalence of depressive symptoms and increased risk for future cardiovascular events, depressive symptoms frequently go underrecognized in patients hospitalized for acute coronary syndrome (ACS). Identifying an effective approach to depressive symptom screening is imperative in this population.
The purpose of this cross-sectional study was to explore the agreement between Beck Depression Inventory-II (BDI-II) scores and a single screening question for depressive symptoms in 1122 patients hospitalized for ACS.
Independent-samples t tests and χ tests were used to compare the groups with BDI-II scores of 14 or higher and lower than 14. Three separate agreement analyses were conducted using categorized BDI-II scores (≥14, ≥20, and ≥29). Agreement of the BDI-II categories with the responses to the single screening question was assessed with the simple κ statistic. Sensitivity and specificity were calculated using the BDI-II categories as the criterion standards for depressive symptom screening.
The agreement analysis revealed a moderate level of agreement (κ coefficient = 0.42) between the BDI-II scores of 14 or higher and the single screening question. Of the participants who reported a BDI-II score of 14 or higher, 61.65% answered yes to the single screening question (sensitivity, 0.62). For those who had BDI-II scores of lower than 14, a total of 82% responded no to the single screening question (specificity, 0.82). When using higher BDI-II scores to define depressive symptoms (≥20 and ≥29), the level of agreement decreased, whereas sensitivity increased to 0.76 and 0.90, with a trade-off in specificity (0.79 and 0.74, respectively).
These results suggest that the single screening question for depressive symptoms correctly identifies depressive symptoms 62% of the time but inappropriately identifies depressive symptoms 18% of the time in patients hospitalized for ACS. This suggests that the single screening question for depressive symptoms may be used with caution to initially screen patients with ACS, who can then undergo a more thorough assessment for clinical depression.
尽管抑郁症状普遍存在且未来心血管事件风险增加，但在因急性冠状动脉综合征（ACS）住院的患者中，抑郁症状常常未得到充分认识。确定一种有效的抑郁症状筛查方法对于该人群至关重要。
这项横断面研究的目的是探讨在1122例因ACS住院的患者中，贝克抑郁量表第二版（BDI-II）得分与一个抑郁症状单一筛查问题之间的一致性。
使用独立样本t检验和χ检验比较BDI-II得分≥14和＜14的组。使用分类的BDI-II得分（≥14、≥20和≥29）进行三项独立的一致性分析。使用简单κ统计量评估BDI-II类别与单一筛查问题答案之间的一致性。以BDI-II类别作为抑郁症状筛查的标准，计算敏感性和特异性。
一致性分析显示，BDI-II得分≥14与单一筛查问题之间存在中等程度的一致性（κ系数=0.42）。在报告BDI-II得分≥14的参与者中，61.65%对单一筛查问题回答“是”（敏感性为0.62）。对于BDI-II得分＜14的参与者，共有82%对单一筛查问题回答“否”（特异性为0.82）。当使用更高的BDI-II得分来定义抑郁症状（≥20和≥29）时，一致性水平降低，而敏感性分别提高到0.76和0.90，但特异性有所下降（分别为0.79和0.74）。
这些结果表明，抑郁症状单一筛查问题在62%的时间里能正确识别抑郁症状，但在因ACS住院的患者中，有18%的时间会错误识别抑郁症状。这表明抑郁症状单一筛查问题在初步筛查ACS患者时可能需谨慎使用，之后这些患者可接受更全面的临床抑郁症评估。
