st: Multiple imputation of repeated examinations data

[roland andersson <rolandersson@gmail.com>]

I have clinical data of patients examined one to three times with some
hours interval.

The data looks like this in the wide format
age, sex, diagnosis, time1, var1_1, var2_1, var3_1, time2, var1_2,
var2_2, var3_2, time3, var1_3, var2_3 etc

Some patients have only one examination, others up to three.

I have previously identified and used the latest examination and
imputed missing values for only that examination, but would like to
take advantage of all the data from all examinations and especially
the dynamic effect of the time. The problem is how MI or ICE can
recognise and associate the time of the examination with the results
of the examinations at that time and see the dynamic change?

I have seen this thread which is a bit similar but did not answer my
question as they dd not use the dynamic change between the
observations
http://www.stata.com/statalist/archive/2013-03/msg00217.html


Regards

Roland E Andersson
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/