国人缺血性心血管病发病危险的评估方法及简易评估工具的开发研究

目的 研究开发适合我国人群疾病特点且方便临床使用的心血管病发病危险度评估方法和评估工具。方法 依据中美心肺血管疾病流行病学合作研究队列随访资料，采用Cox比例风险模型拟合最优预测模型，并校正人群危险因素长期变化趋势的影响，采用独立人群回代检验和计算ROC曲线下面积来检验模型的预测能力。进一步建立简易预测模型，并据此制定适合我国人群的心血管病综合危险度简易评估工具。结果 中美心肺血管疾病流行病学合作研究1983～1984年基线调查年龄35～59岁，剔除基线患有冠心病、脑卒中及主要危险因素资料不全者后男女共计9903人，截止到2000年平均随访15．1年，共发生冠心病事件105例、缺血性脑卒中266例、缺血性心血管病360例。基线年龄、性别、血压、血清总胆固醇、体重指数、吸烟和糖尿病与冠心病、缺血性脑卒中和缺血性心血管病(ischemic cardiovascular diseases，ICVD)事件发病有互相独立的显著关联，且联系的方向和规律一致。据此建立的分性别ICVD事件10年发病危险预测模型，经过校正人群危险因素的长期变化趋势，证明能够很好地用于1992～1994年新建立队列的ICVD发病预测，其ROC曲线下面积(AUC)男性最优模型为0．799，女性最优模型为0．844。简易模型的AUC与最优模型几乎相同。结论 初步开发的ICVD事件10年发病危险预测模型和简易评估工具具有令人满意的预测能力，也能够较好地反映国人发生心血管病的综合危险。

A study on evaluation of the risk of ischemic cardiovascular diseases in Chinese and the development of simplified tools for the evaluation

Objective To develop appropriate methods and tools for evaluation of ischemic cardiovascular diseases (ICVD) risks that is in accordance with the characteristic of cardiovascular diseases profile in China. Methods The sex-spesific best prediction models were developed using Cox proportional hazard model, based on the 17 years follow-up data from the PRC-USA Collaborative Study of Cardiovascular and Cardiopulmonary Epidemiology cohort, and adjusted for the long-term shift of the major cardiovascular risk factors. The area under ROC curve (c-statistic) were calculated and the models were applied to an independent population cohort, to test the predictive capability and accuracy. Simplified categorical models were further built for providing a base to develop the simplified evaluation tools appropriate to Chinese population. Results The PRC-USA collaborative study cohort enrolled a total of men and women age 35 to 59 in 1983-1984. 9 903 partcipants remained after exclusion of those having coronary heart disease, stroke, cancer and those with incomplete data of major risk factors, and were followed up once every 2 years for 15.1 years on average till 2000. During the follow-up, 105 CHD, 266 ischemic stroke, and 360 ischemic cardiovascular diseases occurred. Age, sex, blood pressure, serum total cholesterol, body mass index, smoking and having diabetes at the baseline were independently significantly associated to the incidences of CHD, ischemic stroke and ICVD, and the patterns of the association were similar. Upon this, sex-specific models for prediction of a 10-year risk of ICVD were developed and adjusted. The adjusted best prediction model was demonstrated well in application to the independent cohort established in 1992-1994. The c-statistic value was 0.799 for men, 0.844 for women of the best models and almost identical to that of simplified categorical models. Conclusion The prediction models and simplified tools for estimating 10-year risk of ICVD and had satisfied predictive capability and well reflected the integrated cardiovascular risk in Chinese.