摘要
目的应用Meta-analysis方法对国内9个有关A型行为与冠心病关系的病例对照研究进行综合定量分析。方法采用Fleiss模型和DerSimonianandNanLaird模型（D—L模型）处理数据。结果累积病例为1139人，对照为967人。采用Fleiss模型计算合并OR为2.32（95％CI1.93～2.79），D—L模型计算合并OR为2.32（95％CI1.80～2.99）。失效安全数远大于56。结论A型行为与冠心病有关，失效安全数大，可能的偏倚和混杂因素影响小。
Objective 9 case-control studies on the relation between type A Behavior and CHD in China were analysed by means of Meta-analysis. Methods Respectively, Fleiss's model and Der simonian and Nan Laird's model (D -L'S model) were applied to process data. Results The cumulative cases and controls were 1139 and 967. The pooled OR(type-A Behavior vs No type-A Behavior)were 2. 32(1. 93～2. 79) in Fleiss's model and 2. 32(1. 80～2. 99) in D-L'S model. Fail-safe Number (NfS0.01) is much large than 56. Conclusion There is statistically significant correlation between type-A Behavior and CHD. There is less bias and confounding that can affect the result of the study.
出处
《中国行为医学科学》
CSCD
1998年第4期290-292,共3页
Chinese Journal of Behavioral Medical Science
