search

CSSEGISandData / COVID-19

Novel Coronavirus (COVID-19) Cases, provided by JHU CSSE
https://systems.jhu.edu/research/public-health/ncov/
29.11k stars 18.39k forks source link

Numerous states with backlogged deaths #3666

Open CSSEGISandData opened 3 years ago

CSSEGISandData commented 3 years ago

Hello all,

There are numerous states that appear to be performing audits of their COVID-19 death records and are locating previously unreported deaths. As these are added to the county and state totals on the date of discovery, they are resulting in artificial spikes in our daily deaths curves. We are working with state health departments to back distribute these to their appropriate date and county. In the interim, if a state death curve is available, we will use the difference between our current state total and the newly reported state total and place the difference in an "unassigned" entry. We will update these post with new information as well as the README.

List of states: Iowa - 250 deaths reported on January 31, 2021. Alteration in death criteria includes additional deaths from Oct 15 onwards. Delaware - 56 deaths reported on February 5. 54 of these were historical and occurred in December and January. Indiana - 1,507 historical deaths reported on February 5. The difference has been distributed to our unassigned, Indiana category. Ohio - As many as 4,000 historical deaths will be added over the next few days.

Update 02/11/2021: Delaware has been back distributed at the county level by PR #3667 Update 2/16/2021: Ohio has been back distributed at the county level by PR #3687

GiorgiLomaia commented 3 years ago

It seems like there is an issue with New Castle, Delaware case counts. It went from 591 in Feb 9 to 46237 in Feb 10. Are these back logged deaths that were added to the case counts?