Code used to build the CHoRUS year 1 cohort OMOP tables

[briangow]

I'm adding the code I used to build the CHoRUS year 1 cohort OMOP tables to the protected chorus_year1_cohort branch.

This code has a lot of changes because it removes back ticks around the @ dataset / table calls.

More importantly, it makes changes to the SELECT statements to only include a subject_id if it is found in an associated table (of the 2,000 subjects being submitted as the year 1 cohort). It also filters hadm_id by a table that only include IDs for hospital admissions between 2011 - 2016. These are the years we have CXR and Waveform data available. See this line for how this was done to create src_admissions. This approach was taken in general.

Finally, it removes the code for adding extracted waveform features to the OMOP tables. Since these tables are not available yet they will be added at a later time by the group at Tufts.

[tompollard]

This code has a lot of changes because it removes back ticks around the @ dataset / table calls.

Is this fixing a bug in the underlying code? If so, it should really be contributed to the upstream repository, not fixed just in this fork.