1
$\begingroup$

I have a data set from a repeated measurement study comparing two groups with missing data due to lost-to-follow-up (~20%).

I know how to apply multiple imputation method for cross-sectional data. For repeated measurement (longitudinal data) the situation is a lot more complex because we need to make use of the correlation between the Y values across time-points. STATA can do this using the ICE procedure http://www.ats.ucla.edu/stat/stata/faq/mi_longitudinal.htm. But I cannot work out how to do this in SAS. I am grateful if someone can help.

$\endgroup$
1
$\begingroup$

There is an SAS macro called MMI_IMPUTE specifically designed for MI of clustered data. In addition, Mistler (2013) has written a nice tutorial that shows how to use this macro. The article is available online and should be accessible to all.

Mistler, S. A. (2013). A SAS macro for applying multiple imputation to multilevel data. In Proceedings of the SAS global forum: 2013.

$\endgroup$
0
$\begingroup$

I suggest that you don't use MMI_IMPUTE. MMI_IMPUTE uses an imputation algorithm called PAN, which was developed a while back by Joseph Schafer. Unfortunately, PAN is a bit outdated, and is less flexible than newer algorithms (e.g., the algorithm can't incorporate random effects between incomplete variables; all incomplete variables are required to be normally distributed).

I recommend that you instead use a standalone software package called Blimp. Blimp was written by Brian Keller and Craig Enders at UCLA. Unlike MMI_IMPUTE, Blimp can handle random effects between incomplete variables and can also impute some non-normal incomplete variables (e.g., binary variables). The software and associated documentation are available at http://www.appliedmissingdata.com/multilevel-imputation.html. Blimp is free, and the website contains scripts for using it from SAS.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.