Hacker News Reader

Comment by centimeter | Hacker News Reader

How can we separate out deaths of despair (suicides, overdoses, etc.) brought on by lockdowns?

I would expect deaths of despair to be slow-moving and not follow peaks of infection, as is with excess mortality:

https://ourworldindata.org/grapher/excess-mortality-raw-deat...

I certainly don't think deaths of despair will be equally distributed between troughs of infections (when there's likely to be no lockdown) and peaks (severe lockdown).

It would actually be pretty unsurprising because this would be related to the measures, which should be related to infection levels.

One potential study would be a similar comparative analysis of mortality data from Melbourne Australia.

- Longest and strictest lockdown in G20 countries.

- Quality health records, low COVID deaths (as they were aiming to eradicate COVID risk altogether).

Factors difficult to adjust for would be the relatively high (compared to other countries) levels of economic support provided by Australia to those in lockdown, easing rent and utility bill concerns, etc.

Not all lockdowns are equal, after all.

The CDC haven't released any national statistics on suicide, etc. But there is a file that contains "All Cause" mortality and "Natural Cause" mortality, so I subtracted the two and got what should be the count for non-natural causes. That would include homicide, suicide, accidents, poisonings, overdoses.

Here is the graph I got: https://imgur.com/N4lSSam

The drop-off at the end is because of reporting lag. There is more of a drop-off with these than other causes, presumably because producing a final death certificate takes more time on average (forensic autopsies, toxicology testing, etc.).

Through week 35, there were about 13,000 more non-natural cause deaths in 2020 over 2019.

By comparison, the number of excess deaths through week 35 was 258,846. That implies that 95% of the excess deaths were natural causes.

Here are some spreadsheets I've made from CDC data:

[1] Weekly counts of deaths by select causes [1] https://docs.google.com/spreadsheets/d/1G31ODc4eVgzg7etmcCV5...

[2] US Deaths by Week and Year [2] https://docs.google.com/spreadsheets/d/1qucznpabG1aUz0GSiDbi...

[3] U.S. Excess deaths by age [3] https://docs.google.com/spreadsheets/d/1rcGoWRsNxS_zJQ3pJtbW...

What do you mean? Those things will still be recorded normally.

I mean people will point to all-cause excess mortality as a justification to say "Coronavirus killed X million people" even if half of those were lockdown deaths.

But those deaths of despair are measurable and also comparable against “normal” levels, and can be easily separated from deaths where the cause is less certain but where symptoms may have aligned with potential covid infection. They track more than the body count. There will be studies of the mental health toll of the pandemic and its response. We won’t just tally those as likely disease.

You're trying to see something in the data that isn't there. Even if there was a spike in suicides, it would be a drop in the ocean compared to the excess deaths attributed to physical symptoms of COVID-19.

For what it's worth, Melbourne just went through one of the harshest lockdowns (in order to successfully eliminite the COVID-19 in the state), with no spike in suicide rates [1].

[1] https://www.abc.net.au/news/2020-08-27/no-spike-in-suicide-r...

If unemployment suicides maintain the normal relationship to unemployment rates, it’s likely that more people will die from suicides than from having covid as a primary morbidity.

Right. We'll have to wait but they will come to light. To be clear, excess could be due to:

1. COVID 2. From hospitals overwhelmed by covid cases 3. From behavior change due to fear of covid (not presenting w MI) 4. From policy changes made to combat covid (e.g. school close - diminished prosperity - lower LE)

In Norway, the number of suicides decreased from 206 March-May in 2018 to 140 this year [0]. The authors say it can be a random fluctuation, but it seems to be the lowest number since 2014 (2019 numbers not available yet)

[0] https://onlinelibrary.wiley.com/doi/10.1111/acps.13246

Corona doesn’t kill you by overdose, hanging or self-inflicted gunshots.

Right, but idiots and the media will say "the excess mortality was X, so coronavirus killed X people", even if a large fraction of that was actually due to our response to corona.

More people are dying per month from overdoses in BC, Canada right now than from COVID...

I don't know what you think this proves.

The lack of people dying from ebola doesn't justify not containing ebola.

I don't the the OP is 'trying to prove' anything, other than to demonstrate the magnitude of issues and concerns.

Furthermore - the fact that suicide-by-opiates has spiked, and is killing significantly more than those from COVID is actually fairly material.

"The lack of people dying from ebola doesn't justify not containing ebola."

It certainly does if the treatment for ebola is toxic, and more likely to kill you than ebola.

So while we don't want to entertain antivax/anticovid conspiracy ... at the same time our choices do affect outcomes in a lot of ways. 25% of Americans are having difficulty making rent right now, we can't discount these things.

How many were dying from overdoses before Corona?

Reply on news.ycombinator.com