[curious_cat_163]

> nationwide database of electronic health records (EHRs) of 116 million US patients.

If you are expert in this space: Is such a dataset available publicly? If so, are there examples of other studies that have used this? Where does one go to read more about the mechanism of this study? Thanks!

[bearjaws]

Likely this is some aggregator in the Healthcare space that uses tools that effectively fingerprint patients in each EHR.

This deduplicates patients and lets them find specific details like which medications they are on without knowing any PII.

It's very common for researchers within health systems to want to collaborate and combine populations to perform retroactive data analysis.

[jmcgough]

There is no public dataset of EHRs of this size.

[rustcleaner]

I really really really wish my records were kept tight, offline, air gapped, or otherwise not stored on a cloud system with Trust Me Bro™ HIPAA-compliant security.

(The Trust Me Bro™ security aspect is the "It's secure because people will go to jail" and "We so totes won't use this easily subpoenaed data against you" security, when it would be best if the data stayed on a RAID in my doctor's office and an offsite VPN-linked backup instead. This goes 20x for psychs.)

[whimsicalism]

Almost all health datasets are not public