The application of lag times in cancer pharmacoepidemiology: a narrative review

With the increasing utilization of medications worldwide, coupled with the increasing availability of long-term data, there is a growing opportunity and need for robust studies evaluating drug–cancer associations. One methodology of importance in such studies is the application of lag times. In this narrative review, we discuss the main reasons for using lag times. Namely, we discuss the typically long latency period of cancer concerning both tumor promoter and initiator effects and outline why cancer latency is a key consideration when choosing a lag time. We also discuss how the use of lag times can help reduce protopathic and detection bias. Finally, we present practical advice for implementing lag periods. In general, we recommend that researchers consider the information that generated the hypothesis as well as clinical and biological knowledge to inform lag period selection. In addition, given that latency periods are usually unknown, we also advocate that researchers examine multiple lag periods in sensitivity analyses as well as duration analyses and flexible modeling approaches.