脑卒中发病风险Logistic回归综合评估模型研究—基于欧美西方人群循.docVIP

脑卒中发病风险Logistic回归综合评估模型研究—基于西方人群循证数据 周亮1 程迪祥2 刘岭1 张彦琦1 伍亚舟1 刘小钰1 易东1△ （1. 第三军医大学卫生统计学教研室；2. 重庆邮电大学软件学院） [通信作者] 易东，电话邮箱：yd_house@163.com [基金项目] 国家自然科学基金面上项目（编号30872184 ［摘要］目的 基于Meta分析结果，探讨疾病发病风险综合评价模型。方法 根据人群脑卒中发病危险因素队列研究和病例对照研究资料，利用Meta分析方法对危险因素的危险度进行合并，得到合并统计量OR (Odd ratio, OR)，从而建立Logistic风险评价回归模型。结果 最终纳入10个队列研究和5个病例对照研究，14个危险因素。通过数据筛选，有8个危险因素进本次模型，包括：[BMI≥25.0 kg/ m2]（Body Mass Index, BMI）、[SBP/120-140 mm Hg]（Systolic blood pressure, SBP）、[SBP≥140 mm Hg]、高血压病史、糖尿病病史、冠心病史、房颤和吸烟，其综合危险度为：1.19、1.15、1.55、2.20、1.99、1.40、1.51和2.05。结论本次研究以系统评价为基础建立西方人群脑卒中风险评价模型，有效地避免了人群样本量偏少的因素，建立的模型具有循证医学证据，为预防西方人群脑卒中的发病具有现实意义。 ［关键词］脑卒中；危险因素；西方人群；Logistic回归；风险评价模型 ［中图法分类号］ ［文献标识码］A The study of integrated assessment model by Logistic-regression in the stroke—based on the evidence-based data of Western population Zhou Liang1, Cheng Dixiang2, Liu Ling1, Zhang Yanqi1, Wu Yazhou1, Liu Xiaoyu1, Yi Dong1 (1Department of Health Statistics, College of Preventive Medicine, Third Military Medical University, Chongqing, China 400038; 2School of Software Engineering,Chongqing University of Posts and Telecommunications, 400065, China.) Supported by the National Natural Science Foundation of China (No Noand No. Corresponding author: Yi Dong, Tel: E-mail: yd_house@163.com ［Abstract］ Objective Based on the result of meta-analysis, to explore the risk integrated assessment model of disease. Methods Based on the data in the cohort study and case-control study of stroke, to combine risk degree of the factors by meta-analysis, and get the pooled OR(Odds ratio), then build the risk integrated assessment model by Logistic-regression. Results The 10 cohort studies and 5 case-control studies were included in the analyses of this study, there were 14 risk factors. After screening, the 8 factors has been admitted in the model, including [BMI≥25.0 kg/ m2], [SBP/120-140 mm Hg], [SBP≥140 mm Hg],hypertension, diabetes, CHD(corona