Anyone had Covid-19 antibody testing yet?
Not yet but the Korean government is planning to test everyone in the entire country for antibodies, so hopefully soon.
These ELISAs are among the least reliable tests we have related to novel coronavirus, and they're our only means of (scalably) detecting these antibodies at present, with unacceptably high rates of both false positives and false negatives. I'd wait until we have something more reliable. And that's even without accounting for the fact that having a "positive" result with a reliable test does not allow us to conclude that the subject has immunity (a status distinct from having antibodies).
Thank you for saving me typing that.
Just curious, what would you wait for in terms of reliability before seeking a test? The new Roche tests (being deployed soon?) are claiming 99.8 specificity and 100% selectivity (link), is that high enough where you would pay for a test? I know the answer also depends on what % of the population we expect already had C19 also.
Also, I know it takes longer to establish whether or not antibodies provide immunity and for how long. Do we have any reason to suspect that wouldn’t be the case for C19 for at least some amount of time (weeks, months, years)?
Just a thought: We should not have to pay for the test out of pocket—it should be provided to every American as a right by health departments/our healthcare providers via government purchase. If the government wants to open up, we should not have to pay out of pocket—otherwise, we will have lots of unknown possible carriers in service industry who cannot afford a test. And yes, having everyone tested benefits even those who pay more on taxes (my family, included) to make sure our country is safe.
If it is the government’s job to be concerned with its economy running, it is also its job to be concerned with the citizens that make its economy run.
I agree, somewhat. The CHINESE government should be paying for these tests, not the US government.
I'd accept payment in Kung Pao chicken.
All out of Kung Pao chicken, but lots of bat soup left. Pick up or delivery?
I don't think you understand what a "right" is.
it should be provided to every American as a right by health departments/our healthcare providers via government purchase.
A "right" something you are born with not something that is given to you. I think you mean an "entitlement" which is a benefit provided by the government and paid for by the taxpayers.
I think the misuse of that word confuses the discussion.
Keep in mind it could take years for decision makers to choose a satisfactory test and then manufacture and distribute 350,000,000 of them. While it would be ideal to know who has antibodies, we can't afford to keep everything closed for that long. Not to mention, we still aren't sure how long the antibodies will be effective. This may become like the flu with annual mutated strains.
This is such a dumb argument. By your own definition everything is an entitlement. Name one 'right' that you have that someone bigger and stronger couldn't take from you without government protection, them giving you said right.
And the award for Largest Soy Consumer of the year goes too......
Why you lashing out at me? It's not my fault you're not smart enough to defend your opinion.
You might try using the same terminology. Entitlements are given and may never arrive. Rights are something you are born with and can be taken away. Our government doesn't give us rights, it protects them. My rights are of no cost to others while entitlements are something for which someone else pays. If I didn't say it correctly, you can always look it up.
I don't really understand how you're making this distinction. If the government never protects my rights than they never arrive either, just like your claim for entitlements. What 'rights' do you think you're born with? Prior to governments and society and such, were humans born with the same 'rights', how would one know these were 'rights' since there was no one to protect them and effectively might was the only right?
i suspect it will be quite a while before mass antibody screening. unreliable testing. availability. heavy use in healthcare community and military first. and most unfortunate...people seeing a positive as a ticket to be socially irresponsible. sad but true.
Positive test does not confer immunity. The government won't offer immunity passports or anything of the sort.
So aside from doing random surveys to determine the rate of infection amongst a population, there's no reason for the government to test individuals.
We will always have possible unknown carriers at any given time, even ones who have been previously exposed. That won't change with one round of antibody testing.
I know the answer also depends on what % of the population we expect already had C19 also.
A 99.8% specificity is quite good. And, you're right; the usefulness of the results depends upon how prevalent the antibodies are. If they're in 2% of the population, a positive test is 91% likely to be correct. If they're in 3%, that rises to 94%, and if they're in 5%, you're up to 96%.
That specificity and sensitivity is probably sufficient. The problem is that most of the tests on the market are no where near this level and unless you can get this Roche test the data is basically worthless. It's also important to note that the FDA really isn't doing much in the way of review on the antibody tests like they're doing with active infection tests so there's no real assurance of the quality of the data package Roche is using the determine specificity and sensitivity. It's very possible the data package is light or has some deficiencies given how fast companies are pushing these assays.
Is there a way of testing that's ~100% accurate that they use to verify these tests? Otherwise, how do they really know the accuracy? Or is it 1 person taking 10 tests, if 9 of 10 come back positive/negative it's 90% effective kind of thing.
If a test is 80% accurate, does that mean it will test correctly for 80% of people, or 80% of tests will be correct?
If the later, couldn't someone just take 3 tests, if all 3 are positive you are 99% sure you're positive. If you test positive on 2/3, you are XX% sure, etc.
If a test is 80% accurate, does that mean it will test correctly for 80% of people, or 80% of tests will be correct?
Accuracy is published in terms of the true positive rate (sensitivity) and the true negative rate (specificity). The true positive rate is the likelihood that a positive test is correct -- a 97% sensitive test will return 97 positive results if run on a population of 100 positive individuals; the other three would get a false negative. The true negative rate is similar -- a 97% specific test will return 97 negative results if run on a population of 100 negative individuals; the other three would get a false positive.
That's why you need to know the prevalence of the condition to interpret these tests. For something that's very rare, the false positive rate can overwhelm the true positive rate.
If the later, couldn't someone just take 3 tests, if all 3 are positive you are 99% sure you're positive. If you test positive on 2/3, you are XX% sure, etc.
Error in these tests should not be expected to be independent -- if you get a false test result once, the expectation is that you'll get a false result repeatedly. (For example, you may have another antibody with a similar physical structure, and the test is detecting that instead). When re-testing, it's best to use a different type of test.
Slowly being rolled out at our hospital In Florida for a select group of employees ( tested +,exposed, 1st line,? Not clear who) then opened up in a few phases over coming weeks/months
My sister bought some tests from people (maybe a pharmacy or urgent care that has me access to it). She and her husband were antibody negative even though they and their kids were sick in early March. But I looked up the test, and China thought this test was invalid (which is where it was from). So did the US government.
Beaumont started antibody testing its employees. A friend of mine was positive for covid last month, and his antibody test said he never had it. So the consensus is that these test aren’t reliable yet. Too much variability. We need more time unfortunately to develop better tests is my thought.
I think the antibody tests by Abbott Labs and the ones used by the Stanford study https://mgoblog.com/mgoboard/cv-stanford-antibody-study-results-are-out might be more reliable? But yes, there are many antibody tests that cannot be relied upon.
There are definitely some reliable tests, but there are too many unreliable tests that unless you know exactly which test you're getting and know enough about the statistics then the test is worthless.
Pretty sure the Stanford study used the tests from China that were later prohibited by the FDA. Unless they did another one.
The only issue I have right now with the any antibody test is the accuracy. False positive and false negatives can be routine and the rush to get this out could compound inaccuracy issues.
The only issue I have right now with the any antibody test is the accuracy
I mean, that's like saying that Mary Todd Lincoln only had one issue with their night to the theatre.
Quest Diagnostics (at least in Michigan) has QuestDirect that does the antibody testing without a doctor's order. It's a blood draw and results are within 3 days. Cost is $120 and is not covered by insurance. I did it because I had some symptoms back in Feb, was curious, and didn't want to have to go through the hoops of contacting my health care provider first. It came back negative.
Do you know what type of test they use? Or the legitimacy/accuracy of the test?
I see where you can order it online, but it doesn't really explain much about what the test is besides testing for antibodies.
We are pretty sure my wife got it in early March. Urgent care she went to even mentioned it as a possibility but said they couldn't test for it at that time. They gave her Z pack and she recovered in a week or so. I had a headache and body soreness for about a week straight soon after, so I could've been one of the people that was relatively asymptomatic. Would be nice to confirm or know if it was something else.
I want to do this myself. I caught something Thanksgiving week and remained sick until mid-January. Dry, persistent cough, extreme fatigue and whatever it was didn't respond to two separate z-pack doses and I had a flue shot. My physician said at the time that whatever I had, seemingly his entire patient network was getting and he had no idea what it was. The only Covid-19 symptom that didn't manifest was a high fever, I never had one of those.
FWIW both wife & son got it, had same symptoms (wife did run a high fever) as did several co-workers. I would bet anything I will test positive for the anti-body.
I would take that bet. Unless you were actually in Wuhan in Novemer.
A lot of people saying they had symptoms this winter need to remember that the flu season this year was also very bad and obviously vastly more widespread than COVID-19. If you think you've had COVID due to symptoms you've experienced it's much more likely you had one of the other common viruses that continuously circulate in our population.
This is a good point, your symptoms need to have occurred concurrent or maybe slightly before prevalence in your area for covid19 to be likely.
It's totally obvious from excess death figures that this started in late January/early February.
Right, there are certainly excess deaths that point to the outbreak starting as early as January. However for the average person who experienced symptoms back in January, the chances that they had COVID are still extremely low.
Take for instance the hardest hit state in the country in New York. I've seen some antibody estimates that about 12% of the state has already had COVID. What would the percentage have been back in January or early February? Let's just say 3% for the sake of this argument - so that's 600,000 people with COVID in New York state in January (probably a high end estimate). And with an estimated 90% of cases being asymptomatic that leaves 60,000 people who could have been experiencing symptoms that early - or far less than 1% of the population. And with the dozens of other influenza and coronaviruses that are common in our population that makes it extremely unlikely that you or anyone you know had COVID even as late as January or February.
There is plenty of evidence that at lest small numbers of people had it in mid to late January. Late November, I’m pretty skeptical of.
yeah...because reporting, data, statistics, cause of death reports have all been SO accurate thus far. On one end we say the Chinese were lying out their asses...and the US government...well that's a whole different level of whatever. So what are the odds, University of Texas labs who like to play with Chinese virus tests...may have been part of this? Considering it's NEW...meaning newly identified, if you think MgrowOld couldn't possibly have contracted this before the timeline...you are a bigger fool than I ever gave you credit for. I know you have all this science stuff to 'back you'. But at some point, common sense gains an upper hand.
I get your point but there is a difference between skeptical and couldn't possible have had.
They are going back and finding more deaths associated with COVID-19 in January but November seems too early. Also, the dry cough comment seems out of place. Everyone I know that had it or says they had says they were coughing up a record amount of phlegm. I could have this wrong, does "hacking up a lung" mean wet or dry cough?
Please enlighten me to which part of your post you are referring to when you talk of common sense. Are you suggesting that the lab was man made and that in addition to it being released in Wuhan at some point in late 2019, it was also released in Texas in November?
i’m gonna stick with the science stuff.
science vs common sense...SCIENCE has no clue...YOU don't have a test to verify and confirm reliable data even now...swab samples are in question. If you cannot verify it NOW...how the HELL can you possibly know it didn't start before February? Oh wait...the WHO...yeah those science guys who study virus spread...they are now publishing it hit Europe well before the presumed outbreak, and possibly as early as November...
If it hit Europe...by way of travel, consider the number of Chinese who visit Disney, NYC...American colleges across our continent...you think it started in China and STAYED in China? that's some rock-solid data chief. It's no wonder this whole thing is a predictive clusterfuck if this is our mindset trying to solve the outbreak...we are fucked.
Here's common sense...it's a NEW virus we haven't seen before. It hit hardest at the peak of flu season and pneumonia season. If flu/pneumonia starts in late fall, and peak in mid-winter...I'll let you solve the rest of the question. The answer will blow your mind.
There is new evidence out of France today that people there had it in November 2019, maybe even earlier.
How about we NOT treat the bullshit coming out of communist China as gospel.
We have no fucking idea how early this virus spread out of Wuhan because China is a repressive dictatorship and they fucking lie about everything.
saw similar information RELEASED BY WHO...but hey...science you know...give me a break. Thanks for posting caup. AGREED.
don, pretty sure you had 'buckeye' fever, and definitely not the good kind. normally i'd say 'get well soon', but the chances of us winning a game the week of thanksgiving for the foreseeable future look rather bleak.
Caught something Thanksgiving week in Ohio? I'm afraid that's called the "we-lost-yet-another-one-to-OSU-blues". Real common around here.
My wife got the same antibody test from Quest here in suburban Chicago. She also tested negative, which was a shock and I am convinced is a false negative.
She got horribly sick after a 3-day business conference in Las Vegas at the end of January. She had all of the telltale signs of COVID-19. Significant respiratory distress, etc. In early February she went to her doctor and tested negative for the flu.
I am convinced she had it. I don't trust the antibody test result.
At the same time, it has historically been the case for many people to get sick after traveling, especially by plane.
I have read that the test currently available are not very reliable.
I am waiting to take the test though I believe I had COVID-19 mid-March through early April. I had a fever for 9 days - up to 104 - and aches/pains/severe fatigue for another week and half following that. I quarantined for 35 days.
The tests aren't super reliable on the individual level (if researches know the false positive/negative rate they can still be helpful from a research perspective) and we still don't know to what extent and for how long antibodies confer any level of immunity. I wouldn't waste my money on one because even if it comes back positive you shouldn't change anything about your life or following social distancing rules . If you are part of a research study great but don't go buy one.
family members out west did. two were positive. one had gone through a normal 'flu', the other was asymptomatic and had no idea they'd had it. both in their 20's-30.