All files that is to be taken out of the 'Aging" environment is required to be reviewed. Following is the outlined process.
OUTPUT REVIEW :
--------------
5) We have embarked on reviewing output that can be taken out of the system.
To get your output reviewed, please:
a) In your assigned folder under your DUA, please create a folder naming
it: username-dua#####.output.YYYYMMDD
b) Place all files that you need reviewed and taken out in this folder.
c) Add a short (readme or description) file that gives brief description
of the files you want to take out. In particular, we are looking for N's
(of beneficiaries) to be >= 10 of any statistics you plan to take out.
The N's of each sub category should be >= 10. For example, if you want to
get # of hospital visits per zip code per month, giving total N of the
sample from which the stats per zip code is created is not suffcient.
Further, lets say an obscure county had 15 # hospital visits for a
particular month, but it could be of 5 beneficiaries only. That would
mean you would need to censor that statistic. Again, note that we are
looking at # of beneficiaries when evaluating CMS' cell suppression rule.
d) Suppress all stats that do not conform with the CMS' minimum #
beneficiaries cell suppression rule.
We also have 3 stata ado codes(thanks to Dan Feenberg)to help censor the data.
If you use stata, that help suppress these small cells. There are
scollapse.ado, stable.ado, and sreg.ado You can try them as well.
Please see:
https://data.nber.org/medicare/new_Private/output_review/output_review.html
e) send an email to outputreview@nber.org indicating the location of the
folder you created in #1 (Please do not email your output files, just give
the location of the folder). We will take it out for you after we review
it.