Key points are not available for this paper at this time.
Immortal time in observational studies can bias the results in favour of the treatment group, but it is not difficult to identify and avoid
Lévesque et al. (Fri,) studied this question.