时间相关性偏倚在药物流行病学研究中普遍存在，常常导致低估药品的不良反应风险，控制这些偏倚是从真实世界数据中获得真实世界证据的关键，但现有研究对该问题的关注仍然不足。电子医疗数据库的发展为控制药物流行病学中时间相关性偏倚，从而更加准确地评估药品的安全性和有效性提供了契机。本研究旨在：①通过文献研究和专家咨询建立药物流行病学中识别时间相关性偏倚的标准和流程，指导研究的设计和分析；②通过蒙特卡罗模拟构建各种设计要素组合下的队列和病例对照研究，探索控制每一种时间相关性偏倚的最佳设计和分析方案组合，在保证结果偏倚最小的同时，提高精确性；③利用鄞州区域医疗数据库开展实证研究，对本研究建立的最佳设计和分析方法组合策略进行验证，探索在真实世界研究中最佳策略的适用性。本研究结果将有助于推动药物流行病学方法学指南的完善，加强从设计到分析各阶段控制时间相关性偏倚的能力，获得真实可信的证据，进而促进临床合理用药。

Time-related bias is very common in pharmacoepidemiology studies and often leads to underestimating the risk of adverse drug reactions. Controlling of these biases is the key to obtaining real-world evidence from real-world data, however, current relevant researches are far from sufficient. The development of electronic health databases provides a golden opportunity to control time-related bias in pharmacoepidemiology, thereby providing a more accurate assessment of the safety and effectiveness of drugs. Therefore, the aims of our study are: ①To establish criteria and procedures for identifying time-related bias in pharmacoepidemiology, which will be helpful for guiding design and analysis of pharmacoepidemiology studies, through literature review and expert consultation; ②To explore the best combination of design and analysis method to control each kind of time-related bias, which can improve the accuracy of results while minimizing outcome bias, through creating and compare various cohort and case-control designs by Monte Carlo simulation; ③To use the Yinzhou regional electronic medical database to verify the best combination strategy of the study design and analysis method established in this study, and to explore the applicability of the "best strategy" in real-world study. The results of our study will help to enrich the methodological guides on pharmacoepidemiology, and will further promote rational drug use in clinical practice.