I think you need to put more thought into the design of your output data set. Every row (observation) in the data set will have the same variables. But not every observation will have the same quarters pass your filtering criterion. While Q3 and Q4 for SAU-Chief-Level4-Regional-Brazil qualify, it may be Q2 and Q5 for another observation, and Q4 only for another.
So what do you really want the data set to look like?
Sorry if I wasn't clearer earlier; I am new to this whole field of coding.
So basically I want to find the line items ( eg.SAU-Chief-Level4-Regional-Brazil) which has atleast one instance where there is a growth rate of more than 70% between two quarters. So ideally, if there is a >70% growth btwn Q3 and Q4 of SAU-Chief-Level4-Regional-Brazil, then the output must be something like:
SAU-Chief-Level4-Regional-Brazil 119 145
and such line items must be exported (proc export) as an excel output, where any line item with atleast one instance of >70% grwth must be filtered and exported.