Left-censored recurrent event analysis in epidemiological studies: a proposal for when the number of previous episodes is unknown | Publicación