Often, the prevalence ratios of disability (for example) by single years of age show considerable fluctuations due to sampling variations. Thus, it is preferred to use average prevalence ratios for a five or ten years age interval for Sullivan method. It implies to estimate the total years of life lived in such age intervals. There are two possibilities:
in the abridged life table is
, which is greater than one year. Formulae
from 9 to 13 can be extended in that case.
derived
from the life table (i.e. 
) and values of 
calculated
by summing values of 
for single years of age x from the
unabridged life table. In the example above (see 1.3),
were calculated with the following formulae:
Because the second method is generally more accurate, it is highly recommended to use it if possible.