Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×

Comment Re:Along with cesium, antidepressants, pesticides, (Score 1) 138

by matthewd (#60860314) Attached to: Microplastics Revealed in the Placentas of Unborn Babies

Best thing to do for Christians is just stay out of politics entirely. Jesus told his followers his that his Kingdom was not part of the world. When he comes back, then his Kingdom will take over the planet. Until then, it doesn't make much difference what the current governments do so it doesn't really make much sense to support this or that policy or this or that politician.

Comment Re:Very bad estimations (Score 1) 334

I suspect something else it happening with this statistic.

What if, on average, it takes the average person longer to recover from Covid-19 (or to be recorded as having recovered--going x number of days without symptoms, or having taken 2 tests that are negative more than 24/48/72 hours apart, or whatever criteria are in effect in each locality) than it takes for Covid-19 to kill the average person that dies from it? What if there are not enough people collecting outcome data from people who have obviously not died from it but have recovered?

That would mean there is a perpetual lag on this death rate statistic, because deaths get much recorded sooner than recoveries. Known outcomes at the moment are weighted towards deaths. The current 17% is reflective of the information we have at the moment, not reality.

The CDC is not estimated the current known case fatality rate. The estimates they came up with are for modeling purposes. The current CFR doesn't have any bearing on modeling.

Comment Re:0.4% is a lot. (Score 1) 334

Your 6% calculation is wrong.

You can't take the number of deaths divided by the number of cases and come up with a case fatality rate. That's because you don't have outcomes for all of the cases. In fact you don't have outcomes for the majority of cases.

You need to divide the number of deaths by the total outcomes (recovered+deaths). That gives you a case fatality rate of about 17% at the moment. Clearly the lockdowns haven't worked, right? We can expect another ~200,000+ deaths easily as the current unresolved cases get resolved and lockdown restrictions are lifted, and even more to come.

Well there are obvious reasons that's not the case I'm sure you don't need that explained.

I think it is pretty clear what the CDC means by the overall case fatality rate. Maybe I'm wrong.

The CFRs for each age group that they provide all differ by an order of magnitude. Those numbers are determined by actual outcomes. For modeling purposes though, the actual current overall CFR (17% as noted above) is meaningless. If you don't believe me, consider that the CFR was running around 33% a several weeks ago. What can that number predict? Why did it drop in half in a matter of weeks?

The overall case fatality rate of 0.4% only makes sense if the individual CFR's by age bracket are weighted by actual population demographics.

If you weight them based on the age brackets of actual cases, that just gets you back to the current case fatality rate.

If you don't think that .62% of the population accounting for 42+% of deaths skews the statistics, I'm not sure anything I can say would convince you.

Comment Re:0.4% is a lot. (Score 1) 334

The assumptions behind your math are incorrect. 0.4% is the overall rate, which is a combination of the CFRs for different age ranges.

You could apply the 0.4% to the entire U.S. population, and get an expected # of deaths around 1.3 million. Or you could apply it to a subset of the U.S. population provided the age distribution of that subset matches the population as a whole.

What you can't do is just multiply the overall CFR rate times the number of cases so far and come up with the expected # of deaths and compare it to the actual number of deaths. The reason is the age distribution of cases is different than the age distribution of the population as a whole.

That's why your calculation doesn't match the actual numbers. Actual deaths are higher because the share of people getting infected that are older (and subject to a higher CFR) is higher than the share of those people of the entire population as a whole.

 

Submission + - SPAM: The CDC Is Still Botching the Coronavirus Testing Process 1

Submitted by schwit1
schwit1 writes: The disease control agency is a poster child for bureaucratic incompetence.

As The Atlantic reported yesterday, the agency has been conflating the results of two very different tests: viral tests, which determine if an individual is sick right now, and antibody tests, which are designed to reveal whether an individual has ever been exposed to the pathogen. This makes it impossible to determine the true spread of the virus at any given moment. As The Atlantic's Alexis Madrigal and Robinson Meyer write, combining the results of these two tests without providing a breakdown of how many fall into which category is, "at best, a debilitating mistake."

It isn't the first such error the CDC has made. At nearly every stage of the crisis, the agency has botched the job in small and large ways that continue to cripple our response. The agency is a poster child for bureaucratic incompetence.

Early in February, the agency misreported the results from mislabeled test samples taken from a group of individuals who were being quarantined at a San Diego military base after being evacuated from Wuhan, China, where the virus originated. The CDC initially informed the group that, following testing, everyone was negative for the virus. But after the mislabeling was discovered, it turned out that one woman had tested positive—and had been released from the hospital to return to quarantine on the base. The mixup was blamed on a miscommunication with the hospital.

A far more consequential error also occurred in February, when the agency botched the development of the first batch of test kits that states were supposed to use to begin the testing process. The CDC had already declined to use a German test backed by the World Health Organization, preferring to create its own, as is typical for the agency. This cost several weeks during the time when the virus was just beginning to spread in the U.S. And when the CDC did send out test kits to states, the majority of those kits delivered faulty results. Crucial elements of the test kits were "terribly designed," a scientist who studies viruses told Vice. Only when the testing process was turned over to the private sector was it put back on track.

In developing the faulty kits, the CDC had made basic scientific mistakes. Eventually, an agency spokesperson admitted that the agency "did not manufacture its test consistent with its own protocol." The errors were simple and avoidable, and they have consistently left independent experts aghast. "The incompetence has really exceeded what anyone would expect with the C.D.C.," one Harvard epidemiologist told The New York Times in March. "This is not a difficult problem to solve in the world of viruses."
Link to Original Source

Comment Monks watching video need to be banned too (Score 1) 90

by matthewd (#59638438) Attached to: Bruce Schneier: Banning Facial Recognition Isn't Enough

Babylon 5 exists in a universe where face recognition technology was banned. In an episode where they had to find a suspect in many hours of recorded video footage, they brought in an army of monks to review the footage and identify the suspect. Clearly the surveillance state abusing it's power.

Slashdot Top Deals

"No matter where you go, there you are..." -- Buckaroo Banzai

Close