Read on Twitter
This plot tracks Covid-19 deaths in Sweden taking the reporting delay into account. Out of the 80 new deaths reported today (red), only 6 died within the last 24 hours. Code here: https://github.com/adamaltmejd/corona_stats, feedback appreciated.
New version! Thanks for the comments. Quite a number of deaths added to yesterdays numbers. Somewhat worrisome but could just be that reporting speed has increased. Updated version can now be found at http://adamaltmejd.se/covid .
New day, new graph. Long weekend in Sweden so expecting lots of delay now. Don't think we are in the green just because deaths look low in the coming 4 days. (Also: taking the opportunity to learn @github actions, so now the code updates automatically too!)
Weekend lag continues. None of the 18 deaths added today actually died today. But on the other hand. It seems more and more clear the growth rate has levelled out.
April 15, second work day after Easter and we should really start to see a lot of death reports. I was expecting >200, so the current numbers are indeed promising. Changed graph to show number of days delay because legend space is running out :). Also added weekend indicators.
April 16. Easter still looks very good, but 8th of April is almost at 100 dead. Thanks for all your feedback on the graph!
Sorry I want to clarify. When I say very good I'm just referring to the growth rate in daily deaths. Absolutely horrific that this disease is killing so many people in Sweden each day.
Recount and re-definition coming next week, likely inflating the numbers. Current numbers include all who died with confirmed Covid, also outside of hospitals. Unconfirmed but suspected Covid deaths to be added. (article in Swedish: https://www.dn.se/nyheter/sverige/osakerhet-kring-svenska-dodstalen-nu-ska-antal-avlidna-faststallas/)
April 17. Curve continues to look quite flattened.
April 21st. As has been the case most Tuesdays, lots of deaths added to the numbers today. While 185 deaths is horrible, daily numbers still look stable at around 80 dead per day, maybe even trending down after a peak during easter. Too early to say though.
April 22nd. Lots of deaths are still being added to the easter weekend and the week after. What yesterday looked like it could be the start of a decline now more looks like a plateau.
Important response from Tegnell from a question by @emanuelkarlsten. FHM runs checks of infected against death registry to catch deaths outside of normal healthcare only approx once per week. Lots of response from municipalities today apparently which explains v. long delay.
April 23. Today's graph has a few updates. Mainly the projection has been tweaked to only use historical lags from the last two weeks, increasing the size of the light grey bars by quite a lot.
