Objective To introduce a method for calculating age-weighted death proporti
ons (wDP) for comparison purposes.
Materials and Methods A methodological study using secondary data from the
municipality of Sao Paulo, Brazil (1980-1994) was carried out. First, death
s are weighted in terms of years of potential life lost before the age of 1
00 years. Then, in order to eliminate distortion of comparisons among propo
rtions of years of potential life lost before the age of 100 years (PYPLL-1
00), the denominator is set to that of a standard age distribution of death
s for all causes. Conventional death proportions (DP), PYPLL-100, and wDP w
ere calculated.
Results Populations in which deaths from a particular cause occur at older
ages exhibit lower wDP than those in which deaths occur at younger ages. Th
e sum of all cause-specific wDP equals one only when the test population ha
s exactly the same age distribution of deaths for all causes as that of the
standard population.
Conclusion Age-weighted death proportions improve the information given by
conventional DP, and are strongly recommended for comparison purposes.