Can't find corresponded stay_id in edstays table.

[ChihchengHsieh]

Hello, Dear authors,

I just read through your paper.

It's really an amazing and inspiring work.

Therefore, I came here to try to play with this dataset.

However, I found some stay_id in the master_sheet can't be found in MIMIC-IV-ED

I saw you commented out this line --FROM physionet-data.mimic_core.transfers, but I still go to check the transfers table in the MIMIC-IV Core.

I do found one with the same stay_id in transfers table. However, the time is not in the right span.

The time of conducting the CXR study is 2179-08-12 07:56:53

And, the intime and outtime are 2130-08-17 12:16:00 and 2130-08-17 16:52:00, which doesn't include the time of study.

Is it because of the update of MIMIC-IV ED?

Kind Regards,

Chihcheng

[alexkarargyris]

@ChihchengHsieh Sorry for the very late response. Unfortunately Github didn't notify us. I hope you managed to resolve this. I assume that this update to MIMIC-IV-ED has created back-compatibility issues. Did you reach out to the MIMIC-IV-ED authors to investigate? I will do the same from end. I am sorry for the very late response. Apologies!

[ChihchengHsieh]

@alexkarargyris No worries. I ended up using the staying time to determine the subject_id across different versions (0.4 & 1.0), but it leads to a dataset with pretty small size. I have tried to contact the author of MIMIC-CXR, but they haven't reponded yet. If they can mange to upgrade the CXR database with 1.0 compatable subject_id, I think we can identify the stay_id again.

All the best,

Chihcheng