a question about surveyreg and age-adjustment

liiiiiiiib — Mon, 17 Jan 2022 07:52:18 GMT

Hi. I want to produce the duration adjusted prevalence of Diabetes complications by gender. I used this link: https://www.cdc.gov/nchs/tutorials/NHANES/NHANESAnalyses/AgeStandardization/Task1b_SAS92.htm

My questions are: can I use the proportion of each group from the study cohort (my_data) instead of "Standard Proportions for 20-year Age Groups Based on the 2000 U.S. Census Standard Population"? The same for duration, see below code. And do I have to adjust for the interaction term as in the link (sex*duration 0.1305 0.1962 0.2347 0.4386 0 0 0 0)?

Please note: sex is coded as 1 or 2; duration is coded as 1, 2, 3 or 4. Thank you.

proc surveyreg data=final nomcar; strata sdmvstra; cluster sdmvpsu; weight wtmec20yr; class dmdurcat riagendr; domain agegroup*diadmadult;
model dncat= riagendr dmdurcat riagendr*dmdurcat/noint clparm solution vadjust=none;
estimate "male       " riagendr 1 0  dmdurcat 0.1305 0.1962 0.2347 0.4386  riagendr*dmdurcat 0.1305 0.1962 0.2347 0.4386  0 0 0 0  ;
estimate "female    " riagendr 0 1 dmdurcat 0.1305 0.1962 0.2347 0.4386  riagendr*dmdurcat 0 0 0 0 0.1305 0.1962 0.2347 0.4386;
ods output estimates=sexdn;
run;

topic a question about surveyreg and age-adjustment in Statistical Procedures

a question about surveyreg and age-adjustment