Problem of immortal time bias in cohort studies: example using statins for preventing progression of diabetes
Distribution of the number of citations over years.