72 comments

[ 5.1 ms ] story [ 126 ms ] thread
Wait, what? The same virus? March 2019? Are we sure that's not an off-by-one typo, or an error?

If that date stands up to scrutiny, wouldn't it call into question everything we thought we knew about the origins and history of covid-19?

Not a typo. The abstract explicitly says "long before" the first declared case.
According to the paper : All samples came out to be negative for the presence of SARS-CoV-2 genomes with the exception of March 12, 2019, in which both IP2 and IP4 target assays were positive. This striking finding indicates circulation of the virus in Barcelona long before the report of any COVID-19 case worldwide.
Yes. It would also stand in contrast to everything we know about how quickly it spreads. I am very skeptical of this claim.
It would be very unlikely, but in theory this coronavirus could have spread very weakly for months up to a year before evolving the set of mutations that allow it to spread so rapidly in humans. Indeed, it's exceedingly unlikely that the coronavirus could have had an R0 of >3 right out just after the zoonotic crossing, it's much more likely that it would be significantly lower than one and needed to infect a few people and mutate to adapt to humans.

But to such an extent, and in Barcelona? It's still quite unlikely, though in the realm of possibility.

> in theory this coronavirus could have spread very weakly for months up to a year

The evidence from the study does not support that theory:

"All samples from January 2018 to December 2019 came out to be negative for the presence of SARS-CoV-2 genomes with the exception of March 12, 2019."

This is coherent with a weak infection that subsides after a few days, and was weakly transmissible so only infected one or two people.
Is it? Would it even show up in sewage if it only weakly infected one or two people? I thought only a percentage of Coronavirus infections result in the virus ending up in the feces. And the tests are that sensitive? Do you have a source?
How much do we really know about that?

Recall the apparently 'exponential' increase in cases was in reality an exponential increase in testing - when viewed as a proportion of tests yielding a positive result the virus was spreading not especially fast, and the positive percentage went up quite slowly from a non-zero base.

This article is a preprint and has not been peer-reviewed.

Quote from article:

...”This possibility prompted us to analyze some archival WWTP samples from January 2018 to December 2019 (Figure 2). All samples came out to be negative for the presence of SARS-CoV-2 genomes with the exception of March 12, 2019, in which both IP2 and IP4 target assays were positive. This striking finding indicates circulation of the virus in Barcelona long before the report of any COVID-19 case worldwide.”

Green jelly beans cause acne, I guess. I take back anything I ever said about that one XKCD comic making a trivial point.
I don't think so unless you make a lot of assumptions. That it was a variation that is just as contagious. That it was a variation that could cause as severe symptoms. That it even came from humans. All samples from January 2018 to December 2019 came out to be negative for the presence of SARS-CoV-2 genomes with the exception of March 12, 2019. Why just one month and then nothing for so long before the outbreak? Per the study, sewage includes rainwater from open drains. A highly infected bat could have ended up dead in a sewer. Nature is full of unexpected possibilities.

I bet there are other assumptions without which it doesn't call into question what we think we know.

I don't know how they do things in Barcelona, but in the US the rainwater has its own sewer system. If it didn't the waste water treatment plants would be overrun every time it rains.
Combined sewers are common in the US.

https://en.m.wikipedia.org/wiki/Combined_sewer

My understanding is that combined sewers are being phased out whenever possible for the reason I gave.
I'm not sure how much you know about Barcelona but it's an ancient city who's history can be traced back through to the Neolithic period with very narrow streets designed before motor vehicles. It's not really compatible to any US city.
They have exactly that problem, but it didn't stop dozens of American cities being built with them.
In my original comment:

> Per the study, sewage includes rainwater from open drains.

There are even more details in the study.

Exactly. My questions were and are a polite way of expressing skepticism.
Honestly, its quite problematic.

There are vested interests with a great deal to lose if no counter narrative to the origin of the virus turns up... and they’re already seeding all kinds of conspiracy theories all over the place.

So, while I’m all for “science”, under the circumstances, the burden of “prove this isn’t a false positive” is overwhelmingly higher than the due diligence taken here.

So, yes, it would, but no, it doesn’t.

> vested interests with a great deal to lose if no counter narrative to the origin of the virus turns up

Guess I'm under the rock here—what are those interests? To my knowledge, Trump does the most, ahem, novel hypothesizing, which is pretty much expected routine by now.

Yes, it would be an extraordinary result, if it wasn't deeply suspicious for exactly the reasons it's extraordinary.
The date is very very unlikely to stand up to scrutiny. As others in the thread have pointed out, the data doesn't make much sense, it lacks consistency, they don't show their statistics, and it's very very likely to be a false positive finding that for some reason the authors decided to go ahead and use as the basis for a very out-there claim. Peer review isn't perfect, but it's generally good enough to at least catch something as sloppy as this paper. I'd be very very surprised if it stood up to scrutiny.
I think people should remain open to possibilities.

People started dropping like flies from AIDS in the 1980s from an otherwise completely benign virus they had for 10-12 years, which we didn't know existed and later identified as HIV. A virus which now we are still tracing further and further back, earliest human infection now being 1909. Everything that happened over the last century co-existed with a virus we didn't know was around and slowly incubating and slowly spreading.

The similarity being that we know it is possible, and its really opportunistic infections that cause issues with HIV. With SARS-CoV-2 we have identified A virus and have been trying to trace its origin and are still trying to understand its actual effects.

But maybe its not alone and isn't new.

So therefore it is a non-negligible possibility that there is an additional opportunistic infection that causes issues with SARS-CoV-2.

Or we can dismiss it as a false positive and not mention it at all.

It's still a preprint but I would be very wary of such claims.

How specific can Sars-Cov-2 tests are (to other coronaviruses)? Maybe it was a less virulent variant.

If you processed archival samples and current samples in the same facility, and on the same machines, your immediate prior should be that your archival samples were contaminated during the assay.
They run controls for this reason. So a known negative sample gets processed through the same workflow using the same machines etc...
The usual reminder for these articles:

This article is a preprint and has not been peer-reviewed. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.

It is never worth wasting your time on an unreviewed preprint outside your expertise. There is tremendous pressure to publish exciting results as an academic, and very weak tools for detecting false results, whether dishonest or just statistically unsound.
Indeed. Reminds me of that infamous study about chloroquine ran by scientists of the university of Marseille in March. Despite all the criticism regarding the very small sample and a bunch of huge methodological failures, one of the authors started a campaign claiming that hydroxychloroquine + azithromycin could cure covid-19. That has been used by many politicians in USA and Brazil as a underlying reason for their claims that social distancing measures were unnecessary.

The losses in terms of time, money and, worst, lives because of this disinformation campaign are immeasurable and I feel really bad when I think that probably none of those who intentionally spread such false claims will be ever punished.

Ugh, yeah.

On the other hand, that case is at least a legitimate dilemma -- "ignore every paper when it's published, wait a year for reproductions and meta-analysis" is the right thing to do except for the case where there's a new global pandemic and it's very valuable to be able to stop everyone dying right now if you can.

Agree, but what we've seen was the extreme opposite of a rigorous protocol.

Based on mere six cases or so, authors made an extremely bold claim. That could justify further research, though, and they were in fact made.

Within some weeks there were already more thorough studies showing the inefficacy of those drugs, but the politicization was already irreversible at that point.

I agree so much. Frequency at which the popular media is reporting headlines from unreviewed preprints is appalling. There should be much more journalistic discipline, but that cow has probably left the barn with the internet.
This is one single frozen sample from March 2019 that was positive. Note that subsequent frozen samples from later in 2019 were negative.

Without more data - I'd chalk this up to a false positive or sample contamination.

Also seems like this could end up being a case of multiple hypotheses -- if the false positive rate is very low but all the large cities are performing this test right now, someone's going to hit what will seem to them to be a very unlucky false positive.
Hard to define the null hypothesis on this one, to be honest ...
I'd have to agree given the prior probability of finding a positive result only in the March 2019 sample.

Extraordinary results need extraordinary proof.

The study tested 36 samples, from what I can tell (N is well-obfuscated). About ten of those were from before the COVID outbreak.

The paper doesn’t talk about the false positive rate of the test they used, and doesn’t compute any sort of P value.

~ 3% is a reasonable false positive rate for a COVID test, so you’d expect 0.3 of those earlier samples to test positive.

I doubt it will survive peer review in its current form.

In fairness to the authors, the focus of the paper is testing sewage to track diseases. The most interesting result is that they stopped detecting COVID mid-pandemic due to heavy rainfall.

Edit: plugging into binomial distribution calculator says there’s over a 25% chance of getting one false positive in the pre-covid samples.

(This is not tinfoil-hatting but mere elementary statistics).

As long as you keep testing, you will get a false positive (the farther away you test, the more striking it gets). So, it is: if you keep testing, you will find something.

It might as well have been "There were SARS-Covid-2 positive samples in June 2018", but it was March 2019.

They used five assays. All of these were from the WHO in-house assays list.

In theory, their primers shouldn't have specificity issues. In practice, it's more complicated than that.

The CDC 2019-nCoV rRT-PCR assays have a particularly sketchy history. The WHO in-house assays compendium[4] explicitly states that no false positives are expected for these. But there are huge, annoying caveats to this.

(Tedium warning)

Caveat A: Information chaos. This information has become substantially harder to find on CDC's end than it was originally, having been reorganized, replaced, and removed repeatedly at this point.[1] We can expect this confusion to have forced downstream errors on the part of at least some labs attempting to use the protocols.

Caveat B: The CDC's initial primers[3] were widely criticized for having been badly selected. Example.[5]

Caveat C: The CDC's primers have changed. Without the preprint authors providing more specific information, it's impossible to say what they were testing against for this reason. The preprint authors cite the fda.gov replacement pdf[6], which in turn links out to the CDC primers and probes page[3] via [7] > More Resources on the CDC Diagnostic Panel > Research Use Only RT-PCR Primers and Probes. So all we know is that they used CDC primers, but not which revision.

[1]: CDC replaced its original instructions file, which was located at[2], with a "redirect" pdf containing a link to a replacement fda.gov page. You may find the original document by searching "Tobacco Specific Nitrosamines in Smoke". The reason for this is that the title metadata of CDC's PDFs is often wrong, since they borrow from other files while compiling them.

[2]: URL: https://cdc.gov/coronavirus/2019-ncov/downloads/rt-pcr-panel...

Contents: Real-Time RT-PCR Panel for Detection 2019-Novel Coronavirus - Instructions for Use

PDF Title: Tobacco Specific Nitrosamines in Smoke

[3]: URL: https://cdc.gov/coronavirus/2019-ncov/downloads/rt-pcr-panel...

Contents: 2019-Novel Coronavirus (2019-nCoV) Real-time rRT-PCR Panel Primers and Probes

PDF Title: Avian H7 (Eurasian Lineage) Influenza Real-time RT-PCR Panel Primer Probes

[4]: The compendium document that compiles protocols is a PDF.

URL (landing page): https://www.who.int/publications/m/item/molecular-assays-to-...

Section: "2019-nCoV rRT-PCR Diagnostic Panel Results Interpretation Guide"

[5]: https://tomeraltman.net/2020/03/03/technical-problems-COVID-...

[6]: https://www.fda.gov/media/134922/download

[7]: https://www.cdc.gov/coronavirus/2019-nCoV/lab/virus-requests...

"~ 3% is a reasonable false positive rate for a COVID test"

There is no way they are getting a 3% false positive rate. That's way too high for a 5 product RTPCR test. Not even the serology tests are that bad.

Mislabeling is also an option.
Until they show full genome sequencing of a novel strain under controlled conditions, I’d assume this is contamination or a crappy primer set. If you’re running a bunch of PCRs, especially of the same thing, there is a good chance of environmental contamination from older sample (shipping containers labs for testing have split partitions because of this). There has also been times reported where there has been primer contamination at the synthesis company because they were also building SARS-CoV-2 controls.

Also, there is a very easy way to figure out if it’s real - why didn’t they just Sanger sequence the outputs of the RT-qPCR? Such a simple experiment.

I'm sure they sneezed immediately before testing.
After some early discussion, I've yet to see much about testing sewage. Can it be used to accurately establish a relative infection rate?
This is the quote: "All samples came out to be negative for the presence of SARS-CoV-2 genomes with the exception of March 12, 2019, in which both IP2 and IP4 target assays were positive. This striking finding indicates circulation of the virus in Barcelona long before the report of any COVID-19 case worldwide.".

And no discussion of the obvious possibility that they screwed up the testing, any discussion of whether they could repeat the test on another part of the sample or on the 2019 April-August samples they didn't test, or any sort of attempt at a Bayesian probability analysis.

Pretty ridiculous and shameful, I'd say.

(comment deleted)
The test from March 2019 wasn't even consistent for a positive result.

It found only two of five proteins searched for. Clearly a false positive.

Additionally, it was from a 'frozen archival sample' (they tested frozen samples back to 2018 and daily samples from April-May 2020).
These kind of oddball claims does seem to belong on such a website in the unlikely case other data points arrive later agreeing with it. But not fit for a broader audience.
The way they presented their oddball claims definitely don't belong . They didn't present their statistics and sample size, they got what is almost certainly a false positive, and instead of acknowledging any of that, they decided to draw a conclusion and present the idea that the virus was spreading in Barcelona in early 2019. The responsible thing to do is to retest the samples, check for sources of contamination, figure out why only two of the five primers are positive. And, check with colleagues and the literature to find out if the primers have a known false positive rate and if that could be it. And, go ahead and sequence the results of our PCR to see if you can get covid sequences out of it. And if you can't figure out how that result came out after all of that, present that with the clear disclaimer that this doesn't make sense in the context of existing data, detail what you went through to test that finding, and clearly delineate the uncertainty and call for further investigation. This is just sloppy science.
"And, go ahead and sequence the results of our PCR to see if you can get covid sequences out of it. "

Absolutely. If your hypothesis and/or data flies in the face of all (other) available data, you should leave no stone unturned to prove you are right. Actually showing that the amplified material is in fact COVID sequence would be just one very simple step in doing that. It's been a while, but the cost is about $200 and the time is about 24 hrs to purify and sequence an amplified product? It's quite telling that they didn't bother and/or didn't report the (likely negative result).

less than $200 and more than 24 hours if they're outsourcing it to any of the many sequencing providers out there. Still, less than a week total
The paper is funded by a company with a vested interest in showing the value of wastewater analysis as well, it might not help their cause if they discuss the false positive as a potential problem instead of a great new finding
I wonder if the sewage includes Barcelona's airport. Hypothetically someone with COVID passing through the airport on a connecting flight could have used the bathroom without transmitting to anyone else in the airport.
Perhaps, but this being March 2019 and knowing how easily transmissible this virus is, there is virtually zero chance it was in the wild as early as March 2019 but didn’t start spreading until nine months later.
That's a huge claim without much data to back it up. I am very sceptical.
Can there be a [PREPRINT] flag or the like added to titles for posts like these on HN? It's really important to keep in mind that preprints are not peer reviewed, and with the increase in interest in healthcare papers on this forum, it would be really helpful to be able to make that clear. It's more likely that the march 2019 sample is a false positive, not a true finding, and in peer review the statistical validity of that finding would be called into question and further validation would be required. Just posting the headline makes it seem as if this is something that has been conclusively determined without the disclaimer that medrxiv itself puts at the top of each of these publications: that preprints aren't peer reviewed and should not be used to inform clinical decisions etc.
I recently met someone who claimed the symptoms of their SARS-CoV-2 infection started on December 7, 2019, soon after her friends returned from a trip to China. She said the doctors didn't know what to make of her ("lots of bloody mucus", iirc). By February her doctors realized they should test for SARS-CoV-2, and that she tested positive. Her symptoms mostly resolved by March 2020.

The conventional narrative tells us that SARS-CoV-2 started in Wuhan in December 2019 and reached the United States in January 2020 [1]. I think it perfectly reasonable to think this particular weakling virus spread rapidly in China starting in fall 2019, became more virulent as the sun went away and people's vitamin D levels went down. I think the virus blew up in Wuhan, China and northern Italy because of the air pollution [0].

When the medical industry came up with a test and decided to treat the most obvious symptoms of the virus (oxygen saturation levels) instead of the factors that made people vulnerable to the virus, casualties got out of control [2]. Now we've realized that the virus on its own, in a season with adequate sunlight, is basically equivalent to the flue: good at killing old people, not so good for justifying the massive disruptions hoisted upon us all.

Casualties have gone down now that doctors aren't ventilating so aggressively. Imagine what would happen if they realized Vitamin D and zinc deficiency and chronic inflammation and concurrent bacterial infections (TB, etc) are treatable conditions that make people vulnerable to the virus.

[0] https://news.ycombinator.com/item?id=23085633

[1] https://www.cdc.gov/media/releases/2020/p0121-novel-coronavi...

[2] https://news.ycombinator.com/item?id=22993264

It seems surprising that she could maintain a positive PCR test for eight weeks. I thought the window was much smaller. Here's a paper that says the false negative rate of an RT-PCR test for SARS-CoV-2 increases from 21% on day 9 to 66% by day 21 of symptom onset (and keeps increasing after that).

https://www.acc.org/latest-in-cardiology/journal-scans/2020/...

The march 2019 finding is almost certainly a false positive. This is the more interesting and potentially more robust finding from the paper:

"Unexpectedly, analysis of archival samples revealed the increasing occurrence of SARS-CoV-2 genomes in samples from January 15 to March 4, 2020 (Figure 1, panels D and E). Of note, SARS-CoV-2 was detected in sewage 41 days (January 15) before the declaration of the first COVID-19 case (February 25), clearly evidencing the validity of wastewater surveillance to anticipate cases in the population."

This is unfortunately likely to be overshadowed by the march 2019 thing, and I'm surprised that the authors felt comfortable drawing an actual conclusion based off of the 2019 finding and presenting it as "indicating" that widespread circulation was occurring nearly a year before the pandemic hit worldwide circulation. It's not responsible science and might draw doubts on the validity of the other, more interesting conclusion with more evidence to back it up, and rightfully so.

You can call it, "economy of attention". Same could be said about the Paris case.
> I'm surprised that the authors felt comfortable drawing an actual conclusion based off of the 2019 finding and presenting it as "indicating" that widespread circulation was occurring nearly a year before the pandemic hit worldwide circulation.

“widespread” is your own embellishment.

Did a couple cursory google searches. Looks like the people funding this paper (SUEZ Spain) are selling a COVID senteniel service to track covid spread through the wastewater[0]. Nothing wrong with research funded by a for profit firm. However, in light of the, well, surprising claims this article makes, and how the article suggests that wastewater surveillance could have caught it that early and doesn't seem to be discussing the limitations of this approach (such as the apparent propensity for false positives...), it's something to keep in mind.

[0]https://www.suez.es/es-es/seccion-comercial/nuestras-referen...

Even if it were true, and I'm not even mostly convinced it is. Why would this be surprising and why would anyone immediately jump to the conclusion that it's from humans?

I mean Covid-19 didn't just pop out of the air one day, it's been present all along in animals like bats correct? How is it surpising that an animal carcass found it's way into a swage water sample?

One would have to explain how the virus appeared in both china and Spain. Did a bat carrying it make it all that way or did the virus evolve twice independently? Both seem unlikely.
This seems almost definitely a false positive, surely.
Eveything we know about how contagious this virus is would seem to indicate that this has to be a false positive. How would it even be possible for a virus like this to stay "dormant" while circulating for 6+ months before seeing the spread we have now? It seems completely counter-intuitive to me.
I had it in April 2019 in San Francisco, and have medical history to confirm the symptoms and treatment that worked in my case: steroids plus antibiotics. (Steroids stop the autoimmune reaction, antibiotics help to fight off the bacterial infection which can happen because of steroids). Back then doctors said it was a viral bronchitis "because of some strange coronavirus" which rapidly turned into pneumonia. Three different doctors all were quite puzzled.

So I've been telling ppl for a while that it's been around for longer than is generally believed. I don't think we can say it was "exactly the same" virus, because it constantly mutates. It might have been that I got less virulent and severe version. I also did self-isolate myself on day one of the symptoms, as I do with any cold or flu, which probably reduced the spread from me.

Not saying my experience automatically validates the study, rather that I expect some trustworthy study to eventually confirm that genetically similar strains have been around for almost a year longer.

If anyone's curious about the details, comment below and I'll leave info on how to get in touch.