中文健康问卷在中国大陆人群心理障碍筛选的适宜性研究

目的 检验源于一般健康问卷(GHQ)的中文健康问卷(台湾版)(CHQ)对中国大陆社区人群筛选心理障碍的适宜性。方法 初步研究样本为杭州市社区(n=377)，结论外推研究样本来自于全国4个城市的3707名研究对象。问卷筛选的效果分析使用受试者操作特征(ROC)曲线分析方法。结果 第一个样本的Cronbach系数CHQ-12为0．79，CHQ30为0．89，第二个样本中CHQ12为0．740因子分析CHQ-30可提取4个因子，分别为躯体症状、焦虑和担忧、社会功能低下／家庭关系不良和抑郁，CHQ-l2在两个样本中均可用1个因子来解释。ROC曲线分析显示，CHQ-l2曲线下面积为0．80(95％CI：0．70～0．89)，CHQ-30为0．72(95％CI：0．62～0．82)。CHQ-12和CHQ-30筛选的敏感度分别为76．9％和71．8％，特异度为73．8％和67．9％，Kappa值为0．44(P=0．00)和0．38(P=0．00)。在第一个样本中CHQ-12筛检的心理障碍率为18．13％(95％CI：14．16～22．10)，CHQ-30筛检的为22．80％(95％CI：18．19～27．11)；在第二个样本中CHQ-l2筛检的为21．72％(95％CI：20．39～23．05)。结论 经过适当语言调整后的CHQ特别是CHQ-12可作为精神流行病学调查和社区卫生服务识别心理障碍的筛选工具。

The application of Chinese health questionnaire for mental disorder screening in community settings in mainland China

Objective To examine the applicability of Chinese Health Questionnaire (Taiwan version)(CHQ) originated from General Health Questionnaire (GHQ) for mental disorder screening in community settings in mainland China. Methods A pilot study was conducted in Hangzhou ( n = 377). Three thousand seven hundred and seven subjects were recruited from four cities to validate the results of the pilot study. Validation of the Screening Questionnaire was analyzed, using Relative Operating Characteristic (ROC) method. Results Cronbach's alpha coefficients were calculated to be 0.79 for the 12-items and 0.89 for 30-items to CHQ version in the first sample, and 0.74 in the second sample (12-items). Four factors were extracted from the CHQ- 30, including somatic symptoms, anxiety and worry, social dysfunction,poor family relationship, and depression. CHQ- 12 could be explained by a single factor in both samples.The areas under ROC were 0.80 (95% CI : 0.70- 0.89 ) for 12 items and 0.72 (95% CI : 0.62- 0.82 ) for 30 items.The sensitivities of CHQ- 12 and CHQ- 30 were found to be 76.9 % and 71.8 %, and the specificities were 73.8 % and 67.9 % with Kappa value 0.44 ( P = 0.00 ) and 0.38 ( P = 0.00 ), respectively.The estimated rates of mental disorder were 18.13 % (95% CI : 14.16- 22.10 ) by CHQ- 12, and 22.80 % (95% CI : 18.19- 27.11) by CHQ-30 in the first sample and the rates were 21.72 % (95% CI : 20.39- 23.05 ) by CHQ- 12 in the second sample. Conclusion CHQ, especially CHQ- 12 through slight language modification, could be used for epidemiological studies and on community health care to screen for mental disorder in the mainland of China.