Refine study def - Githubissues

opensafely / vaccine-effectiveness-hospital-admissions-validation

MIT License

0 stars 0 forks source link

Refine study def #1

Open LFISHER7 opened 3 years ago

LFISHER7 commented 3 years ago

Those admitted with respiratory symptoms
Hospital acquired covid
Those that die in a&e
What discharged destination codes to use (currently discharged to ward or ICU).

wjchulme commented 3 years ago

Admission routes differ by hospital/trust. For some trusts who had/have set up dedicated covid wards, it's possible that covid patients will bypass A+E completely. For others, everyone is triaged via A+E even if there is a covid ward. So may be worth looking at differences by trust.

HelenCEBM commented 3 years ago

So may be worth looking at differences by trust.

This would be good but unfortunately we don't have trust codes (even anonymised) in the dataset.

We also know that lots of people with covid have been transferred between trusts so we won't find A&E admissions corresponding to these subsequent admissions.

HelenCEBM commented 3 years ago

Hospital acquired covid

On this, it's possible that only looking at primary diagnoses could help narrow down to those who were most likely admitted with Covid (but not necessarily!). Could also look at timing of tests in relation to admission date as an approximation.