Phecode_mapping -- include date.

[gpbatra]

How do we choose which date to include in phecode_mapping output table?

Admidate, operdate, etc?

Should we have a best value, or include all dates?

Should we create an ordered disease progression table, so multiple phecodes don't appear over the years if a patient is assigned a phecode early on?

Should pick one protocol and stick with it, including MPG mappings.

[carbocation]

Re: dates, for my work, I have preferentially used admidate if not null, then opdate if not null, then epistart if the others are unavailable. Not sure if others have a different experience or different preferences.

Re: ordered disease progression table, I think that would be a straightforward MIN(date) / GROUP BY if the table has all the data.

[carbocation]

Note to look at this code again once we're satisfied with the disease code, which makes some decisions about date order.

[lucidtronix]

@carbocation and I did this a while back. We jerry-rigged the phecodes into tab files and passed it to disease.linux. The disease code is siiiiickkkk! The tab files live at: gs://ml4cvd/phecode_tab_files/ and the TSVs that have been imported into bigquery are at gs://ml4cvd/phecode_bq_tsvs/