Comment by lettergram | Hacker News Reader

Comment by lettergram | Hacker News Reader

IMO this looks to by a symptom of lack of testing. Overall the ratio of tests coming back positive is unchanged:

https://raw.githubusercontent.com/lettergram/covid19-analysi...

Even though we are doing more tests. Meaning the reduction in speed-of-spread likely has more to do with lack of testing.

That being said, I’m sure spread is being reduced, we just don’t know to what extent because we don’t have enough effective testing.

It’s also possible (probably likely) our tests are not exceptionally accurate. It’s the best we have, but given all of this is <6 months old the false positive/negative rate can be high.

Generally, Testing isn’t an effective way to measure spread. Probably we should use “hospitalizations with flu symptoms“ as the best indicator.

You can compare testing per state and deaths here:

Tests: https://raw.githubusercontent.com/lettergram/covid19-analysi...

Deaths: https://raw.githubusercontent.com/lettergram/covid19-analysi...

Overall repo: https://github.com/lettergram/covid19-analysis

Correct. Testing is only a good measurement with massive testing, even randomized testing.

Iceland's data is probably the closest to actual infections and many states probably look similar. (https://www.covid.is/data)

The reported specificity and sensitivity of RT-PCR testing is quite good, but obviously not 100%. RT-PCR is considered the "gold standard" in viral testing.

If you want to contact trace, we would need many orders of magnitude more testing.

If you want to know prevalence over time, and therefore Rt, we could ascertain it very accurately with much less testing, but it would need to be done against statistically valid and unbiased random samples of the population, not at all what is going on today.

Since the testing we are doing today doesn't really satisfy either condition, and there's also no front-line approved treatments for those that are definitively diagnosed (rather, we treat the symptoms, which are self-evident), frankly I'm not sure I understand the extreme focus on testing.

These tests have a false negative rate of 30%

My understanding is that it’s under 10% false negative and false positive typically for this type of testing.

Other tests like the rapid saliva test are totally different.

False negative is higher.

I'm not sure I understand the extreme focus on testing.

3blue1brown on YouTube created a simple model that suggests that testing and quarantine of positive cases is the most effective way to slow the spread (within the simplified model). Testing should allow us to isolate just the infected, and ease off the restrictions on others.

With statistically valid random testing, even detecting clusters / communities (as opposed to individuals) and clearing same would be conducive to reopening the economy.

Unfortunately, the path to doing so "I'm going to force you to get tested, despite you having no symptoms" is political anathema.

> "I'm going to force you to get tested, despite you having no symptoms" is political anathema

This seems less extreme and more palatable than "I'm going to force you to stay home and lose your job, despite you having no symptoms" which is very nonetheless popular right now.

No, no, there’s never going to be enough testing to test asymptomatic cases — Thats everybody.

And a test doesn’t tell you you aren’t going to be positive tomorrow. It just tells you you weren’t positive 48 hours ago!

Fauci explained all this live at the press conference. I wish people would listen. With AIDS it’s different. You can get tested, be negative, take no at-risk behaviors, and a year later that negative test is still relevant. With COVID by the time you have a negative test result back it’s no longer even relevant. Only positive tests are relevant.

The fact is that there’s no point in testing people to demonstrate that they are negative. And there will never be enough testing to test every asymptomatic person, and a significant proportion of spread is from asymptomatic carriers. Those three incontrovertible facts lead to some very basic and sobering conclusions.

So this fear of people being forced to show a negative test assumes we’ve somehow found our way into an anti-science twilight zone kabuki theater to begin with. Which frankly wouldn’t be far off from where we are already, but can we please at least acknowledge it’s based on fear and hysteria and not any kind of science?

> statistically valid random testing

Believe you missed a word in there?

Random testing for surveillance purposes would absolutely require testing asymptomatic people.

Not because you care if they're sick (though that's useful information), but because you care if the community has a change in the Sars-CoV-2 infection rate.

> With COVID by the time you have a negative test result back it’s no longer even relevant.

A family friend died on Friday. His heart went out on him. Big guy on his 60's. He got tested for COVID-19 before he died at the hospital. He and his now widow had some sinus congestion. She couldn't enter to he hospital to be with him when he died. A few days later, his widow got the negative result for her husband. She is alone, scared, dealing with this loss. can't see her kids or grandkids. She can sleep a little better at night knowing that death is not knocking at her doorstep.

Relevant? From a human perspective it is most definitely relevant. Maybe not as much from a epidemiology or number cruncher's perspective but like all expert opinions they are open to interpretation and second opinions. Statistics are easy to cherry pick for whatever agenda is being pushed behind the scenes.

Just like masks were not relevent at first. They were scarce. Suddenly they are relevant for the public to use. I believe this testing scarcity and related lack of any cohesive national mobilization is the biggest failure of our US government. From a public health perspective this dismissal of negative testing is just a way to manage the demand side for testing.

Public health and epidemiology experts have very different motivations than individuals dealing with loss and fear during this crazy time.

> I'm not sure I understand the extreme focus on testing.

Group think.

"hospitalizations with flu symptoms" is a good indicator, but you have to account for the fact that people who catch it in different areas (or with different incomes) will have different levels of access to a hospital. many people are dying (or recovering) at home.

Not just that.

NYC has seen the number of people dying at home increase from around 20 per day to 200 per day. The difference is likely a mix of COVID-19 cases, people not seeking treatment out of fear, and people just spending significantly more time at home. But, that kind of change makes accurate statistics even more difficult.

Just read an interesting article using ICU admissions as a proxy. Of course it does lag, but it may be more consistent than hospital admissions.

http://medrxiv.org/cgi/content/short/2020.04.13.20063388

It does saturate at some point, though. If you're turning patients away who are in need of ICU treatment (as some Italian hospitals were), your ICU admissions will be constant and equal to your rate of release from ICU, so it will appear that you've "flattened the curve" even if the number of sick people continues to grow exponentially.

True, the paper talks about patients needing ICU, if you can get that statistic instead.

Reply on news.ycombinator.com