search

oxford-pharmacoepi / THIN_incidence_prevalence_colorectal_cancer

0 stars 0 forks source link

make sure small sample numbers are suppressed for inc/prev counts < 5 #14

Open daniellenewby opened 3 months ago

daniellenewby commented 3 months ago

As we are doing inc/prev age standardiztion we needed results not to be suppressed so need to check that these are suppressed afterwards for inc/prev crude results and any age std results too

With THIN they suggested suppress counts less than 30 but it was a suggestion not a request so we may have to increase this min cell count to 30.

daniellenewby commented 1 month ago

for now suppress at 5 with the view we might have to change this :) use code from other birthcohort crc study where we have a smilar issue