Differentiate missing data from NA data

[see24]

NA should be not applicable. Missing data is a problem that can be fixed by extracting the data. Ilona suggested marking missing data with 99999

[see24]

This is going to be challenging to fix on the R side of things for a few reasons.

• readxl::read_excel always sets "" to na
• When joining data from different sources R sets missing data to NA in some cases that is correct in others it is data that should be 99999

This needs to be done but could happen after submitted. Might be easiest to just do it in the final version of the db