Abstract: Background　Non-specific chest pain （NSCP） refers to the patients with chest pain and negative coronary angiography but without identified clear causes of physical disease after repeated examination.Identification of it mainly relies on coronary angiography due to no other effective methods.But coronary angiography is invasive and expensive.We attempted identify NSCP before coronary angiography by studying its characteristics.Objective　To identify NSCP before coronary angiography by comparing the clinical data and psychological status of ischemic chest pain （ICP） and NSCP in patients with suspected coronary heart disease and angina pectoris，and evaluating the value of multi-factor prediction model with psychological factors included in distinguishing NSCP from ICP.Methods　239 subjects （78.4%） who completed the above three scales were included in this study.There were 136 cases in NSCP group and 103 cases in ICP group.We enrolled the patients suspected with coronary heart disease and angina pectoris with paroxysmal chest pain as the chief complaint from Cardiac Center，Mongolia Autonomous People's Hospital from July 2016 to March 2017.All of them had indications for coronary angiography and underwent it for the first time.According to coronary angiography and ECG results，we divided them into NSCP group and ICP group，and compared the clinical data as well as mean scores of Somatic Symptom Scale（SSS），Patient Health Questionnaire-9（PHQ-9） and General Anxiety Disorder-7（GAD-7） between the groups.We used multivariate Logistic regression to explore the influencing factors of NSCP，and performed an ROC curve analysis to examine the value of each indicator and the combined predictive variable P in identifying NSCP.Results　 239 subjects （78.4%） who completed the above three scales were included in this study.Compared with ICP group，NSCP group had younger mean age，lower percentages of ever smokers and diabetes history，and lower mean TC，but had higher percentages of women and unemployed，and hypertension history （P<0.05）．Moreover，NSCP group showed higher mean scores of SSS，PHQ-9 and GAD-7（P<0.05）．Logistic regression analysis showed that gender，age，TG，SSS score，PHQ-9 score were the influencing factors of NSCP（OR=0.935，0.235，0.376，1.522，1.204，P<0.05）．And the AUC of the aforementioned indices in identifying NSCP was 0.527，0.625，0.564，0.873，0.637，respectively.The AUC of the combined variable P reached 0.921，with≥0.344 as the cut-off value，the sensitivity and specificity of diagnosing NSCP were 0.882 and 0.837，respectively.Conclusion　Female，younger age，decreased TG，severe somatic symptoms and depression are high risk factors for NSCP.Among them，SSS has the highest diagnostic accuracy for NSCP.But the combination of multiple indicators can yield even higher accuracy，and can detect NSCP earlier and noninvasively.

Key words: Chest pain, Non-specific chest pain, Coronary artery disease, Coronary angiography, Somatoform disorders, Self-rating scale, Multi-factors digital model

摘要： 背景　冠状动脉造影阴性的胸痛患者经反复检查仍找不出明确的躯体疾病原因时统称为非特异性胸痛（NSCP），目前尚缺乏有效识别NSCP的方法，主要依靠冠状动脉造影等来排除，但冠状动脉造影有创、价格高，本研究试图研究NSCP患者的临床特点及其心理状况，从而在冠状动脉造影前对NSCP进行识别。目的　对比分析拟诊冠心病、心绞痛患者中NSCP、缺血性胸痛（ICP）患者的临床资料及心理状况。评估包括心理因素在内的多因素联合预测模型区分ICP与NSCP的诊断价值，实现冠状动脉造影术前对NSCP的初步识别。方法　选择2016年7月—2017年3月以发作性胸痛为主诉就诊于内蒙古自治区人民医院心脏中心拟诊冠心病、心绞痛，具有冠状动脉造影指征且首次接受冠状动脉造影检查的患者为研究对象。根据冠状动脉造影与心电图检查结果分为NSCP组和ICP组。入选患者均使用躯体症状自评量表（SSS）、抑郁自评量表（PHQ-9）、广泛性焦虑量表（GAD-7）进行评估。对比两组患者的临床资料及各量表得分。对拟诊冠心病、心绞痛患者发生NSCP的影响因素进行多因素Logistic回归分析。绘制各指标识别NSCP的受试者工作特征（ROC）曲线，分析各指标及联合预测变量P对NSCP的识别价值。结果　完成以上3种量表的受试者共239例（78.4%），纳入本研究。其中NSCP组136例、ICP组103例。与ICP组相比，NSCP组年龄、有吸烟史患者比例、有糖尿病史患者比例及三酰甘油（TG）较低，女性患者比例、无业患者比例、有高血压史患者比例较高（P<0.05）。NSCP组SSS、PHQ-9和GAD-7得分高于ICP组（P<0.05）。多因素Logistic回归分析显示，年龄、性别、TG、SSS得分、PHQ-9得分是NSCP的影响因素，其OR值分别为0.935、0.235、0.376、1.522、1.204（P<0.05）。以上各指标识别NSCP的ROC曲线下面积（AUC）分别为0.527、0.625、0.564、0.873、0.637。联合测试变量P识别NSCP的AUC为0.921，以≥0.344为诊断截点诊断NSCP的灵敏度和特异度分别为0.882和0.837。结论　女性、相对年轻、TG低，躯体症状障碍及抑郁程度重是心内科拟诊冠心病、心绞痛患者发生NSCP的影响因素。单个指标中以SSS评分对NSCP的识别价值最大，多指标联合预测可进一步提高对NSCP的早期识别价值，从而实现早期无创的对NSCP患者进行初步识别。

关键词: 胸痛, 非特异性胸痛, 冠状动脉疾病, 冠状血管造影术, 躯体型障碍, 自评量表, 多因素数字模型