We would direct Hawkins' attention to Table 2, [Research]¹ in which it is clear that the mortality rate decreased by 8.7% and 4% among British Columbia residents aged 65 and 75-76 years respectively. In our article we also point out specific age and care groups in which the mortality rate increased and comment on possible reasons for these changes (e.g., increasing acuity among people aged 90–93 years).¹ We specifically skewed our cohorts to represent larger proportions of more elderly groups (e.g., people aged 83–85 and 90–93 years) to have sufficient numbers of people in each age group to be able to compare care patterns. Thus, because of differences in the age composition of our cohorts, the age-adjusted rates (using the combined population of the cohorts as the standard) are not comparable to overall mortality rates for the total population of elderly British Columbians or, probably, to mortality rates in any other jurisdiction. Moreover, Hawkins' citations represent mortality trends for populations and (or) time periods quite different from ours (the Australian data assess mortality trends for all age groups from 1907 to 1990; the European data cover mortality from 1960 to 1990 among people aged 60–64, 70–74 and 80–84 years; and the US data are primarily for the years 1997–1998). Therefore, for important methodological reasons, it is inappropriate to compare these mortality data with the findings of our study.

As pointed out by Noralou Roos [letter],² significant downsizing in acute care capacity and decreased mortality rates have occurred simultaneously in the jurisdictions Hawkins mentions. Thus, from our perspective, the link proposed by Hawkins between acute care downsizing and mortality is, at best, far from clear.