Figure 3.
Illustration of immortal time bias in the Sullivan et al. observational cohort study of HRT in patients undergoing CABG surgery.42
The top patient received her first prescription of HRT well into follow-up, long after CABG. The time between cohort entry and the first HRT prescription is immortal and unexposed (thick line) and should have been included in the analysis with the third patient, while the subsequent exposed period should have been included in the analysis with the second patient, to avoid immortal time bias.