Read on Twitter
Thread on covid mortality in India: for over a month I have been wondering about and working on the issue of 'excess mortality'. In many developed countries, estimating covid mortality has now moved to excess mortality.
What this means is that researchers, the media and even in some cases govts are coming around to accepting that they will miss some COVID deaths, and so a good way to guess how many are being missed is an comparing "all-cause mortality" - total weekly deaths - with previous years
Here are the UK's official estimates: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/analysisofdeathregistrationsnotinvolvingcoronaviruscovid19englandandwales28december2019to1may2020/technicalannex
Here are South Africa's https://www.samrc.ac.za/sites/default/files/files/2020-06-09/WeeklyDeaths2June2020.pdf
And the FT is compiling excess mortality data for 20 countries here: https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441
Here are South Africa's https://www.samrc.ac.za/sites/default/files/files/2020-06-09/WeeklyDeaths2June2020.pdf
And the FT is compiling excess mortality data for 20 countries here: https://www.ft.com/content/a26fbf7e-48f8-11ea-aeb3-955839e06441
"Why doesn't India do this?" is something I get asked a lot, and some international publications have run with estimates from Mumbai and some other cities that seem to show overall mortality has fallen, and not risen. Except..
India's baseline morality data is hugely problematic. I'm no fan of saying "the data isn't great" and then doing nothing. I prefer to identify the issues with the data to build better estimates for the future.
That's what I've been doing for a long piece for @IndiaSpend. I looked at the concrete issues with India's existing mortality data and the ways to estimate excess mortality going forward https://www.indiaspend.com/how-india-could-fill-in-the-blanks-on-excess-mortality/
Saying that it's likely that we're missing some covid deaths should be entirely uncontroversial. Countries with 100% death registration accept it and try to estimate how much they're missing! We should do it too. The piece has some wise suggestions on how to do it - do read!
