Erratum

Open Access
Publisher Website
Google Scholar
Abstract
We write to point out an error in the cohort study analysis program (PYRS) described1 in the IJE in 1986. Versions up to 1.21 are affected. The error produces a negative person-years contribution for individuals who are censored on the last day of the study, at the same time as reaching a time boundary (usually an age boundary such as 50th, 55th, etc birthday; occasionally the 20th, 30th etc anniversary of first exposure to risk). For a typical study design with fiveyear age groups and five- or ten-year risk intervals, this error would be expected to affect less than 1 in 1800 of those rare subjects who happen to be censored (eg who emigrate, or are lost to follow-up) on the last day of the study. It is therefore likely that most analyses done with the program will not have been affected. The error has been corrected, and we have notified all users of PYRS who are known to us with details of the necessary changes to the program: other users should contact Dr Coleman.
Keywords

This publication has 0 references indexed in Scilit: