Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upsurvtab: error in SE and CI of age-standardised survival #135
Comments
WetRobot
added a commit
that referenced
this issue
Aug 9, 2017
… misspecified and lead to too large confidence intervals. wrote test to confirm correctness.
|
Using survtab example, the blue lines show the old confidence intervals and black lines show the new ones. It is not a insignificant difference!
|
|
So this was closed in f19824b. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Non-standardised standard error is correct. Standard error of age-standardised survival should be calculated as follows:
st <- survtab(FUT ~ agegr,breaks = list(FUT = seq(0, 5, 1/12)))st[,SE.r.e2.as:=sqrt(sum(SE.r.e2^2*w^2)/sum(w)^2),by='surv.int']over age groups
agegrwith weightsw.Check standard errors of age-standardised survival measures other than Ederer II, too.