OHDSI Home | Forums | Wiki | Github

Drug exposure end dates going into the future


(Manan Patel) #1

Typically, if we know our data is complete through the end of 2018 then we don’t want event dates past Dec 31, 2018.

However, in many cases we are calculating the drug_exposure_end_date from the start date + days_supply. It’s possible these can pour into the future. Should we cap the dates with Dec 31, 2018 or should we make an exception for drug_exposure_end_date and allow it to go into the future?

If there’s a THEMIS or achilles rule for this I’d love to know what it is. Thanks!


(Melanie Philofsky) #2

I’ll take a stab at this since no one else has offered an opinion :slight_smile:

You will create an OBSERVATION_PERIOD for your data with an end_date = Dec 31, 2018, the last day of complete data for the dataset.

Per the OBSERVATION_PERIOD table convention “Both the _START_DATE and the _END_DATE of the clinical event has to be between observation_period_start_date and observation_period_end_date”. So, in theory, you need the DRUG_EXPOSURE records and the DRUG_ERA records to end on or before Dec 31, 2018.


t