Hi Stata Forum

I have an unbalanced set of panel data with individuals having a measurement at baseline. Some may have had the measurement repeated again at a second, or third or fourth visit.
For example example, there were 15000 who had a first measurement, 5000 with a second measurement, 1000 with a third measurement.

As ID have missing measurements for the variable of interest (X) across all time-points, I have performed multiple imputation for this (missing ~10% of measurements).
Data is in long format and xtset after multiple imputation.

I want to analyse only those id who had a measurement value X<7 at the first visit, and see what their trajectory is on subsequent visits, i.e. restrict initial observations to only those with this value.

I was thinking of performing: mi estimate: xtreg X i.visit "(if X<7 on first visit)"
I am not sure how to code this last part, or how to create a variable that is conditional on a specific time point.
I am also not sure if I can apply xtgraph after multiple imputations as well...but initially, I just want to review the coefficients.

I've tried mi xeq: gen newvar=1 if X<7 & visit==1. But then I am not sure how to include only these subjects again.

Reading the cond() help file for conditional variable generation doesn't appear to solve my issue.

Thank you for any assistance.
Nitesh Nerlekar
(Using Stata 14).