Disability prevalences for Sullivan method and their variances

The determination of is the subject matter of the table 2. The different columns of that table are calculated as following:

 

 

Age Institutionalized Disability prevalences

(1) (2) (3) (4) (5) (6)

0 0.000 772 0.010 0,010 0.000012

5 0.000 741 0.024 0.024 0.000032

10 0.000 745 0.025 0,026 0.000033

65 0.010 534 0.188 0.196 0.000281

70 0.016 380 0.274 0.286 0.000507

75 0.038 310 0.385 0.370 0.000675

80 0.085 228 0.468 0.513 0.000914

85 0.211 180 0.632 0.710 0.000805

1st column:

ages.

2nd column:

the institutionalization ratios in the whole population, ,. These rates were derived from census.

3rd column:

the numbers of males in age-group i who participated to the health survey, .

4th column:

permanent disability prevalence ratios in non institutionalized population (health survey data), .

5th column:

gives , prevalence ratios of disability to use in Sullivan formulae:

6th column:

variance of , . Instead of using formula   (which must be used when the survey directly provides ) we use :

that is to say that we do not consider as a random variable, justified by the fact we use census data.

Remark: If had been determined by a survey, its variance could be non negligible.