We read with interest the articles by Stevens et al. and Davis et al. regarding outcomes for pancreaticoduodenectomy (PD). We were nonplussed by the comment regarding New South Wales (NSW) exhibiting the highest post-operative mortality rate (POMR) without acknowledgement that including this long time period is likely to mask, as demonstrated in the NSW plots, marked improvement over time in their measure. Presenting the mean in the presence of such variation does not give meaningful comparisons.