常用抑郁量表筛查我国青少年抑郁的效度和划界分

Screening for Adolescent Depression: Validity and Cut-off Scores for Depression Scales

目的:检验常用抑郁量表即流调用抑郁量表(CES-D)和贝克抑郁量表第2版中文版(BDI-II-C)筛查我国青少年抑郁的效度和划界分。方法:5565名中学生(年龄:14.17±1.66)完成CES-D和BDI-II-C测试,其中634人[即CES-D总分≥20分(n=322)和<20分(n=312)]进一步完成了学龄期儿童情感障碍和精神分裂症问卷(K-SADS)诊断访谈。对完成访谈的青少年以美国精神障碍诊断与统计手册第4版(DSM-IV)为标准做出抑郁障碍诊断,以此为效标,计算并比较CES-D和BDI-II-C在各自常规和最优划界分下筛查青少年抑郁症和所有抑郁障碍的效度。结果:1CES-D以总分≥29和≥24分为最优划界分,分别筛查青少年抑郁症和所有抑郁障碍的敏感度为94%-97%,特异度为74%-79%,阳性预测值为21%-30%,接受者操作特征(ROC)曲线下面积(AUC)为0.88-0.91;BDI-II-C以总分≥28和≥15分为最优划界分,分别筛查青少年抑郁症和所有抑郁障碍的敏感度为91%-95%,特异度为80%-83%,阳性预测值为23%-35%,AUC为0.88-0.91。2CES-D在常规划界分下筛查青少年抑郁症(≥24分)和所有抑郁障碍(≥20分)的特异度(55%-70%)和阳性预测值(16%-20%)均小于最优划界分;BDI-II-C在常规划界分下筛查青少年抑郁症(≥29分)和所有抑郁障碍(≥14分)结果与最优划界分相似,除筛查所有抑郁障碍的敏感度(99%)优于最优划界分(95%);3CES-D筛查抑郁症和所有抑郁障碍的特异度均小于BDI-II-C。结论:两常用抑郁量表筛查我国青少年抑郁具有较好的效标效度,建议CES-D选用最优划界分,BDI-II-C选用常规划界分。

Objective： To examine the criterion validity and cutoff scores of the Center for Epidemiologic Studies De- pression Scales（CES-D） and the Chinese version of Beck Depression inventory II（BDI-II-C） for Chinese adolescents. Methods： A total of 5565 junior-and high-school adolescents completed the CES-D and BDI-II-C. Diagnoses of major depres- sive disorder（MDD） and any other depressive disorders were made on the DSM-IV. Results： （1）At the optimal cutoffs, the CES-D scores of ≥29 and≥24, separately, for detecting adolescent MDD and any depressive disorder were with a sensitivi- ty of 94% to 97%, a specificity of 74% to 79%, a positive predict value（PPV） of 21% to 30%, and an area under receiver op- erating characteristic（ROC） curve（AUC） of 0.88 to 0.91. Similarly, the BDI-Ⅱ-C scores of≥28 and≥15, separately, for detecting adolescent MDD and any depressive disorder were with a sensitivity of 91% to 95%, a specificity of 80% to 83%, a PPV of 23% to 35%, and an AUC of 0.88 to 0.91. （2）At the commonly used cutoffs of CES-D, the specificity of 55% to 70% and PPV of 16% to 20% were lower than those at the optimal cutoffs. Meanwhile, at the commonly used cutoffs of the BDI- Ⅱ-C, all index were similar to those at the optimal cutoffs. （3）When compared to the BDI-II-C, the specificity of the CES- D was lower. Conclusion： Depression scales show good criterion validity for screening adolescent depression. Neither the BDI-Ⅱ-C nor the CES-D could be used by themselves as methods for case ascertainments, although the BDI-Ⅱ-C is some- what better than the CES-D as a screener. For detecting adolescent depression, the optimal cutoffs are suggested to be used for the CES-D and the common cutoffs to be used for BDI-II-C in China.