Hacker News Reader

Comment by dependenttypes | Hacker News Reader

I was under the impression that quarantine was an option only to "flatten the curve" (and make it easier for the hospitals to treat people) and that the same amount of people would die regardless (from the virus), is this wrong?

View on HN

> that the same amount of people would die regardless

Maybe "be infected", not "die"?

Otherwise there's little reason to flatten the curve (meant to allow for better treatment).

This is one of the assumptions underlying the Swedish approach, yes. The knob you can fiddle is how soon people die: with relaxed response, people who would have died a few months or a year from now might die of the pandemic tomorrow. (But of course, strict countermeasures also have their long-term costs, which are harder to quantify in terms of lives, although that's definitely a component of it.)

That's only assuming we never get a vaccine or an effective treatment. If we do, then the number of people to die before that point matters a lot, because we would get a sharp decline after.

In addition to your points, another reason for "flattening the curve" is to give time for hospitals and health care systems to prepare for a spike in cases, so that every patient who needs critical care can receive it. For countries like Italy that were hit at the start of the pandemic, this was a serious problem.

Where I live, in Victoria, Australia, the state government has used the lockdown time to plan for and deliver a huge increase of ICU beds [1]. Thankfully, it looks like these beds may not even be necessary, but initial modelling suggested a peak of 10,000 active cases with only 500 IUC beds available.

[1] https://www.theage.com.au/national/victoria/thousands-of-bed...

Precisely. I've simulated this numerically and shown that the curve of [deaths]-vs-[duration of widespread economic crisis] is concave, like an inverted parabola that's been truncated by eventually getting a vaccine (or effective treatment).

Perhaps counterintuitively, the points on the strategy curve that have both lowest death counts and shortest widespread economic crisis are the same. I've published these simulations here: "Surprising COVID-19 Strategy: Reduce Economic Damage and Deaths Simultaneously" https://www.youtube.com/watch?v=NuNrkG188HA

The Swedish assumption (not officially, but what I take away from communication is what they assume)

- No effective treatment or vaccine will arrive within a year.

- Measures that completely eradicate the virus aren't sustainable for the time it takes to do so or until a vaccine arrives (e.g. lockdowns),

- Without at least some immunity, "Test, Trace, Isolate" is not a viable strategy for dense regions.

- Without at least some immunity, a second wave is unavoidable, and risks being worse than the first, if it arrives early next season. It's better to have a small immunity in the low risk population now, than let risk high risk groups later.

- Even a small level of immunity has a large effectiveness in limiting spread, i.e. 10% or 20% immunity in a region can make a big difference, so if that can be achieved in the low risk population that is a good thing.

- Asymtomatic spread, presymptomatic spread, and spread from children are all possible but not significant (That is: if symptomatic spread is controlled, that is enough).

Some of these assumptions have been challeneged, some were not at all controverrsial to begin with but have changed over the last months.

It was, yes, under the early assumption that the virus spread rapidly but had a relatively low death rate. It turns out it's the opposite - it's surprisingly easy to stop the spread, but the death rates are higher. So countries that locked down under the plan that they were just flattening the curve still managed to nearly eradicate the virus despite that not being the original strategy.

> surprisingly easy to stop the spread, but the death rates are higher.

Are they? The current IFR estimates (general, not counting age divided groups) range between 0.2 and 0.9%. That might still include a significant number of individuals, but it's not very high (of course by age groups, this changes).

I still wonder whether the high death rates in some countries, including my own, were not due to the virus itself being very deadly, but because it found "niches" of vulnerable people (hospitals, nursing care homes). I don't think anyone has yet studied these dynamics in depth.

Early random sampling in Sweden indicated a IFR around 0.2%, but the more recent antibody test points to 1.1%, putting the expected final death toll up an order of magnitude

Death rate is smaller than it was initially thought to be afaik. Under 1%.

Reply on news.ycombinator.com