摘要
非死亡时间是指在队列研究的随访期内研究对象不可能发生结局事件的一段时间,由非死亡时间错分或剔除引起的偏倚称为非死亡时间偏倚。在基于数据库的药物流行病学队列研究中,非死亡时间偏倚较为常见,而且可能对结果产生严重影响,以致于得出完全错误的结果和结论。因此在文献阅读和研究设计时,应识别和避免非死亡时间及其引起的偏倚。本文介绍了非死亡时间引起偏倚的原理、偏倚的识别以及避免该偏倚的研究设计和数据分析方法。
Immortal time was a period of cohort follow-up during which,because of exposure definition,the outcome under study could not occur. The immortal time bias occured when the immortal time was misclassified or excluded improperly. This bias was quite common in the pharmacoepidemiology studies based on the administrative medical database,and could have such a big impact that lead to completely incorrect results and conclusions. Thus identification and control of immortal time bias were of great importance when reading research papers or designing cohorts. In this article,we introduced the cause,identification of the immortal time bias as well as the methods of designing cohorts and analyzing data to control it.
作者
赵厚宇
李海龙
詹思延
Zhao Houyu Li Hailong Zhan Siyan(Department of Epidemiology and Biostatistics, School of Public Health, Peking University, Beijing 100191, China)
出处
《药物流行病学杂志》
CAS
2017年第3期207-212,共6页
Chinese Journal of Pharmacoepidemiology
基金
国家自然科学基金重大研究计划培育项目(编号:91646107)
