/*
Fragment of SAS code to illustrate how to implement the inverse Chi-square
function in SAS to compute gamma-based confidence intervals for age-adjusted rates.
This example is for 95% confidence intervals.  For other confidence limits,
replace the constants 0.025 and 0.975 with the appropriate values.

Variable definitions:
   CI_lower:  Lower confidence interval limit 
   CI_upper:  Upper confidence interval limit
   maxwt:     Maximum of the weights
   rate:      Age-adjusted rate
   variance:  Variance of the age-adjusted rate

*/

   CI_lower   = (variance/(2*rate))*cinv(0.025,2*(rate**2)/variance);
         
   CI_upper   = ((variance + maxwt**2)/(2*(rate+maxwt)))*
               cinv(0.975,2*(rate+maxwt)**2/(variance+maxwt**2));