Thanks to Venk for reminding me I had planned another gratitude thread on this very question: how did we get this analysis of our #COVID19 patients @nyulangone done so quickly? @PetrilliMD @jones_prof @WhiteCoatLuke /1
1st, AMAZING TEAM dedicating 16-hour days for 2 weeks. Seriously.
But aside from that, unbelievable institutional support.
1) carte blanche from IT and finance to pull data (and repull, and repull as we worked through data/query errors - quite a complicated dataset) @PaulTestaMD
But aside from that, unbelievable institutional support.
1) carte blanche from IT and finance to pull data (and repull, and repull as we worked through data/query errors - quite a complicated dataset) @PaulTestaMD
2) heavy reliance on an exceptional data infrastructure and analyst team we have already built for rapid development of predictive analytic models, overseen by @yindalon (which itself is already doing great predictive model work - stay tuned for more on that soon)
3) warp speed review by IRB - submitted on a Friday night, approved Sunday am (after two sets revisions!) including personal review/comments by IRB director Helen Panageas
4) mini army of med students and RAs who manually reviewed all intubated patients to identify successful extubations, which we are still having trouble reliably identifying through automated query @kira_garry (rest tag yourselves here if you see this; don't know handles)
As for will we share the data: open in principle, devil in details. We are working daily with all other medical centers in NYC and beyond to put together a cross-institutional dataset open for research that could be even better; there are many other such initiatives springing up.
And for those confused by this thread, it's referring to this preprint medrxiv.org/content/10.110…
