Mortality in young offenders: retrospective cohort study

We failed to spot an error in table 1 of this paper by Carolyn Coffey and colleagues (BMJ 2003;326:1064-6). The total shown for person years of observation for the Victoria population (2 812 387) is in fact for females only. The correct total for males and females should be 6 133 674, giving a crude mortality of 3051 per 6 133 674 (0.5 per 1000). This affects the crude mortality given in the results and discussion but does not affect the standardised mortality ratios from which the paper's conclusions were drawn.