This post at NPR’s Goats and Soda, on COVID-19 mortality in India, raises a point worth considering:
How many people have died of COVID-19 since the pandemic began?
The official global total as of this week: 4.1 million.
But everyone agrees the true toll is far greater. A study released on Tuesday looks at how much of a disparity there may be in India, one of the epicenters of the pandemic.
The analysis, from the Center for Global Development, a think tank in Washington, D.C., looks at the number of “excess deaths” that occurred in India between January 2020 and June 2021 – in other words, how many more people died during that period than during a similar period of time in 2019 or other recent years.
Drawing death data from civil registries and other sources, the report came up with three estimates for undercounts. The conclusion is that between 3.4 and 4.7 million more people died in that pandemic period than would have been predicted. That’s up to 10 times higher than the Indian government’s official death toll of 414,482.
Shorter: the mortality from COVID-19 in India alone may be greater than the presently reported global mortality. But that in turn raises a host of questions. While excess deaths represents a reasonable first order approximation it’s far from accurate. Do you include people who died as a consequence of bad policy or poorly implemented policy? Who counts? How do you count?
Right now there is no uniform generally-accepted method of tallying the number of deaths due to COVID-19 not just globally but in the United States itself. The method of counting varies from state to state.
And while I’m on the subject I’ll complain about something I’ve been whinging about for more than a year: there has been no systematic nationwide epidemiological testing for COVID-19. We don’t really know whether 5% of the population has contracted the disease of 50%. I honestly don’t know how you can formulate reasonable policy with at least making an attempt at getting better information.
“But everyone agrees the true toll is far greater. ”
In the US, at least, there is likely a major over count of covid deaths. It is repeated noted that flu and pneumonia deaths have disappeared, and that other deaths are down, too. Almost every death is counted as a covid death.
The CDC claims 10% of the population has contracted covid. That sounds about right for a major pandemic. The Asian flu killed about 116,000 people in the US out of a population of 172 million. It is surprisingly hard to find an estimate of the death rate among infected people. But if it is 1%, then 7% of the population caught it. Most estimates are substantially lower than 1%, which implies a majority of Americans caught it. That is not credible.
The PCR test is apparently unreliable, so I don’t know how you measure exposure.
I will repeat myself, we are going on to two years of exposure, and if only 10% have caught it, and less than 50% are vaccinated, there must be a large percentage of the population with some degree of immunity. Since the common cold is a mixture of coronaviruses, that is likely.
We test for flu now. Flu didnt just disappear because we are not looking for it. Also, deaths for many causes actually increased. Other deaths are not down and not everything is counted as Covid.
https://www.cdc.gov/nchs/nvss/vsrr/covid19/excess_deaths.htm
Steve
I don’t know whether we overcount COVID-19 deaths, undercount them, or both. I would not be a bit surprised if it were both.
With all the variables I’ve always thought that excess deaths, averaging for a few years would be the best measure.
Extrapolate for an aging population.
Less flu is probable, and what have you.
The fact of the matter is that we don’t know the numbers. The 6% meme of last summer was wrong. But we don’t know how to give attribution to CV or comorbidities. There is judgment. Its like saying an extremely ill cancer patient really died of pneumonia.
Interesting how Steve chose to cite comparisons to historically grounded baseline data when it comes to CV, but didn’t want to hear it when similar analyses were applied to the 2020 election. How situational. How convenient.
Everybody dies of heart failure.
Cardiac arrest Dave. Heart failure has a specific meaning.
Engineers don’t take math anymore or those skills go away when you go into finance? How are you going to make sure you are screwing other people and not yourself without good math skills? Anyway, we have really good track record for death rates and we know that they dont vary that much. The increase of 22% (form memory) we saw last year is unprecedented absent war. Elections OTOH have a history of large swings. States change voting patterns (California used to vote Republican). Almost finally, presidential elections for which we have good data are actually fairly uncommon so we are dealing with small numbers. I am guessing maybe 10-20 elections.
Really finally- Most of the numbers people dredged up were BS.
Steve
Thanks for the correction. I’ll remember it.