Fig. 4From: Left-censored recurrent event analysis in epidemiological studies: a proposal for when the number of previous episodes is unknownCoverage according to population and maximum time at risk prior to the beginning of the cohortBack to article page