What we learn dissecting a media story that refuses to die.

These cute little rhesus macques will come running into the story, promise! Photograph by Joel Sartore, National Geographic Photo Ark

CNN asks on April 18:

Recovered coronavirus patients are testing positive again.

Can you get reinfected?

STAT feels compelled to answer this question on April 20:

I’ve heard reports of reinfection or “reactivated” virus. What’s going on there?

What is going on there?

You could say it’s just the media doing its job. They get asked those questions, and try to give people the best answers they have. And, in the pandemic, they are doing a pretty good job of it.

Except in some situations when there are no answers. Or any settled ones. Then the media sometimes steps in to fill the void.

There’s nothing wrong with speculation. I do it all the time.

But there’s doing it, being clear you’re doing it — and doing it in such a way the speculative narrative taps into the dark places of the reader psyche.

Which does, not coincidentally, keep readers and viewers captivated and coming back for more.

It’s a subtle business. Often it comes down to a seemingly innocuous word choice.

So I want to take an ongoing story, the one about Covid-19 reinfection, and deconstruct it as an exercise. Not because I’m so smart or virtuous, but because it has to be done.

This is a media story, not a science story. May it serve as a cautionary tale.

Before getting on with the manly job of deconstruction (Most boys like to blow things up. Where did I leave that blasting powder, anyway?) — an aside.

A science writer knows he’s in trouble when he has to look up the Carl Sagan quote about UFOs.

The quote is, “Absence of evidence is not evidence of absence.”

Sagan didn’t actually say that (I looked it up). The adage has a long history, according to Quote Investigator.

But it came to mind reading yesterday’s (20 April) Guardian:

“Right now, we have no evidence that the use of a serological test can show that an individual has immunity or is protected from reinfection.”

Ouch. But so says Dr Maria Van Kerkhove, an American infectious diseases expert who is the WHO’s technical lead on Covid-19, according to the Guardian.

As we ought to know from UFOs, “no evidence” is an invitation to open field running.

So to calm things down, a little snippet of evidence.

Cue the monkeys. Macaques, to be precise. I did promise to run them in.

On 17 April 2020, Nature Reviews Immunology reported an experiment by Bao and others: “Reinfection could not occur in SARS-CoV-2 infected rhesus macaques.”

The poor guys were infected and allowed to recover. They developed neutralizing antibodies, detectable in their blood. Then the researchers tried to infect them again. Tried hard. Didn’t work. They concluded:

Upon reinfection, viral replication was not detected in nasopharyngeal or anal swabs, and reinfected monkeys did not show any signs of COVID-19 disease recurrence. This suggests that immunity acquired following primary infection with SARS-CoV-2 may protect upon subsequent exposure to the virus.

Take that, Dr Maria Van Kerkhove. Let’s begin.

The Zombie appeared innocently enough.

It was 28 February 2020. Things the were, shall we say, chaotic in Wuhan, China. I just checked to find some other headlines from that day. Boy, here’s a good one: “Coronavirus Risk in the U.S. Is ‘Very Low,’ Trump says.”

We want to pay attention to the date. It will come up later.

The story was from Reuters and ran under the headline and subhead:

Coronavirus reappears in discharged patients, raising questions in containment fight

Experts say there are several ways discharged patients could fall ill with the virus again

When you are reading carefully, the phrase “Experts say” should raise at minimum one eyebrow.

Reuters did a pretty good job reporting the facts.

A growing number of discharged coronavirus patients in China and elsewhere are testing positive after recovering, sometimes weeks after being allowed to leave the hospital.

Uh-oh. An old friend: “growing number.”

Be warned: journalists love to spot trends.

The Germans actually have a word for people who do this to excess, which they consider worrying syndrome. The word is something like Tendenzkeitz, a tendency to see tendencies.

Back to the growing number. The number was four.

For our purposes, I’m going to call them “The Wuhan Four.”

I think Gen Covid should have its own Secaucus 7.

Anyway, Reuters did reassure everyone that The Wuhan Four were not actually contagious:

An official at China’s National Health Commission said on Friday that such patients have not been found to be infectious.

But, later in the story, the Reuters writers, lest their readers grow too comfortable, were unable to resist a great big “on the other hand.”

Which takes us to The Source.

Most science stories can be tracked down to a source.

It’s a simple business. You can’t just make this stuff up. You have to tell your editor there’s a journal article or something weighty and you are going to interpret it for the readers.

Sometimes the source is in plain sight.Sometimes, finding it is like hiking uphill to a little bubbling spring up in the mountains. This one was a little hard to find.

It proved to be something published (online) in the Journal of the American Medical Association (JAMA) on 27 February, the day before the Reuters story appeared.

I say “something” because the Reuters authors called it a “study,” which sounds hefty, but is a dubious word choice. It was actually something called a Research Letter.

Research Letters sometimes come in reporting anomalous observations that the writers think others in their field should know about. This one was written by 7 Wuhan doctors — 5 MD, 2 MS, and 2 also PhD.

The anomaly was that four of their patients, all of whom happened to be medical workers themselves, had testing PCR positive again in weeks following their official hospital discharge, which had required the requisite two negative PCR tests > 24 hours apart.

Now, not everything was known about the progression of Covid-19 back in February. The early wishful thinking was that it was 14 days and done, from onset of symptoms to clinical recovery.

To ruin the suspense, that is no longer the thinking. WHO changed its guidelines to 3–6 weeks in some cases, 8 in extreme ones. I note a recent (6 April) paper in the BMJ plotting viral load on a scale that goes out to 60 days (8.6 weeks).

So now we need go back and take a look the date of the letter. One of the Wuhan Four, we know, was discharged 15 February. That’s less than 2 weeks.

Don’t get me wrong. This is how science and medicine are supposed to work. When the evidence changes, you change your mind.

The problem is that a lot of people — including the reading public and Chinese officials — aren’t very comfortable with messy things that fall outside neat boxes. Were the Wuhan Four among the Recovered, or what? What stamp do we put on their passports? Or their cellphones? All of this will come up again with Immunity.

[Sidebar: PCR on the downhill slope.]

Medium doesn’t do sidebars.

Sidebars are used for technical explainers that will interrupt the flow of a story and threaten, quite frankly, to bore the average reader to death.

So at any time, feel free to scroll down to where it says [End of Sidebar].

PCR. Right. It’s been an experience to see politicians glibly talking about PCR tests on TV. A few weeks ago they couldn’t even pronounce Polymerase.

(Super-hard Bonus Round Jeopardy Question! The answer: “Buttercup Bakery.” †)

The PCR amplifies DNA, RNA and other nucleic acids. And it’s a great amplifier. Sometimes too good.

For patients coming in who might be getting Covid-19, the more sensitive the amplifier, the better. Most false negatives (maybe 20%) for the PCR tests have to do with bad swabbing — not the PCR part. But if the virus is there, PCR will find it. It’s a very sensitive test.

That’s great on what I call the uphill slope of the infection curve.

On the downhill slope, during recovery, it causes problems. In particular, PCR will amplify anything resembling the target it can get its little molecular grippers on: dead virus, fragments of nucleic acid and debris, antibody neutralized virus.

So post-infection, on the downhill slope, false positives are a problem.

(This problem of over-zealous PCR comes up all the time in discussions of sample contamination.There was a period of time when all the Neanderthal DNA in Svante Pääbo’s lab was coming from his lab assistants.)

The problem is you need to draw a line somewhere, but where? The numerical estimates of viral load computed from PCR tests in actual patients do tend downward, but with fluctuations that make the price of West Texas Crude look stable.

Here are charts for nine patients taken PCR tests taken to 28 days out, where the horizontal Pass/Fail is the line of detection. Note the ones (e.g., bottom left) where the yellow lines goes below it, but the next day goes back over it.

† Since you finished the sidebar, you get the Jeopardy Question! “What was the name of the Berkeley dining establishment where Kary Mullis worked — before he got the Nobel Prize?”

[End of Sidebar]

Reuters listed a number of possible explanations for what was going on. All were very reasonable.

The first, “testing discrepancies,” we just talked about.

The second was reinfection: “convalescing patients might not build up enough antibodies to develop immunity to SARS-CoV-2.” Fair enough.

For the third possibility, reactivation, the Reuters writers were clearly smitten with a word, biphasic.

The SARS-Cov-2 virus might never be cleared the by the immune system but, like herpes (HSV), go into hibernation and reactivate from time to time.

Or Covid-19 could be biphasic, the writers suggested, like anthrax.

Anthrax?

That regained the reader’s flagging attention.

One of the Wuhan Four mentioned to his letter-writing doctor that, even after he’d been discharged from the hospital, he had kept himself away from his family because, in an abundance of caution, he didn’t want to take a chance on inadvertently carrying the virus home from the hospital. Very sensible.

The Reuters writers shortened that to:

likely that some recovered patients would remain carriers even after meeting discharge criteria.

Another word choice: carriers.

Didn’t that official at China’s National Health Commission just say:

such patients have not been found to be infectious.

?

He did. But a story about carriers is so much better.

The Zombie goes on NPR

Now, don’t get me wrong. I like NPR. I listen to all the time. It’s one of my car radio buttons.

But NPR can be, so… earnest.

It must be exhausting, being that earnest all the time. I know I couldn’t do it.

The Zombie lay dormant for a month. (Perhaps he’s biphasic.) Anyway, he surfaced again 27 March, 2020. Link here.

NPR’s headline was:

Mystery In Wuhan: Recovered Coronavirus Patients Test Negative … Then Positive

I so love a good mystery.

A spate of mysterious second-time infection.

Hmm. We need help parsing that line.

“Alexa? How many in a spate?”

She doesn’t know.

Lower down in the story we find out. Take a guess.

Four.

Does it mean? The boys are back?

They’re not only back, they’ve brought company:

some Wuhan residents who had tested positive earlier and then recovered from the disease are testing positive for the virus a second time. Based on data from several quarantine facilities in the city, which house patients for further observation after their discharge from hospitals, about 5%-10% of patients pronounced “recovered” have tested positive again.

NPR doesn’t waste time, but gets right to the scary part;

Some of those who retested positive appear to be asymptomatic carriers — those who carry the virus and are possibly infectious but do not exhibit any of the illness’s associated symptoms.

The “carriers” have morphed into “asymptomatic carriers.” Sneaky buggers!

And morally, these are worse than plain-Jane pre-symptomatic carriers.

A pre-symptomatic has an excuse. They don’t know they have It yet.

These recovered people running are around spreading It telling everybody they are well.

Oh dear. Who’s going to stop them?

In a Post-Truth world, facts are a bummer.

Who wants to face facts, when they tell you your grandchildren may parboil, your children may grow up poorer than you are, and you yourself might die a painful death in a few months?

Give me an Alternative Fictional Universe anytime.

It’s easy to populate this fictional universe with stock characters.

Villians!

The Superspreaders — one of my favorites.

Asymptomatic Carriers. You don’t know who might be one of Them. Your neighbor. Your wife — even your husband.The UPS driver. That nice checker at the grocery store.

And The Immunes will be in for it soon enough. See my essay.

We’ll all be living inside our own Marvel Comics dystopian universe.

This is probably going to happen anyway.

But careful journalists don’t have to contribute to it. That’s all.

So answer the question, already

So what is “the” answer? Can you get reinfected?

Short answer: No.

The long answer: Of course you can.

Glad I cleared that up.

It depends on who’s asking the question, and why.

Can you get reinfected?

“Probably not, but we don’t know for sure yet.” Conservative, correct and acceptable. I feel like Miss Manners.

if you get a large enough dose of new virus up the nose, you are technically going to be reinfected

If you don’t want your frustrated inquirer to go off searching for answers on Facebook or Twitter, I would avoid a flat-out “We don’t know yet.”

But the proper answer takes some thinking about why they are asking. Or what they really are asking.

If there is a #2 phrase for our era — after #1, “A lot of people are saying…” — it’s “A lot of people are asking….”

At lot of people asking this question don’t give a hoot about reinfection. They’re asking if immunity exists.

(Yesterday I heard Anderson Cooper say “if immunity is real…” on CNN.)

Will there be immunity to Covid-19? Miss Manners here: “Almost certainly, but we’re not certain how good it will be, or how long it will last.”

Whether you give someone the long answer depends on their level of understanding of how the immune system works.

I’ll try to give in a few paragraphs, without all the caveats.

If your recovered from Covid-19 (the disease) your body eventually produced antibodies that neutralized lots of copies of SARS-Cov-2 (the virus).

Otherwise, you wouldn’t have recovered.

Your blood can be tested for the presence of specifically these antibodies. Not just any old antibodies. The antibodies that helped you fight off Covid-19.

That’s why there’s this talk of blood serum transfusions from the recovered to the seriously ill.

The shorthand word for the level of antibody in your blood is the titer. The higher titer, the better. (For you in the future. To get a higher level, you may have had to have had the disease worse.)

The chart of titer over time in real, recovered people — like out for 6–12 months — is one we all want to see.

And may want to put a horizontal threshold line on that chart to produce a binary, Yes/No answer to the question, “Do you have antibodies?”

The “white stick” antibody tests have a one-size-fits-all threshold level that is mass-manufactured into them — a problem I’ve written about here.

Months on, if you get a large enough dose of new virus right up the nose, you are, technically, going to get reinfected. That’s why the second answer is, Yes.

The difference this time will be in your immune response.

As this second infection tries to get going and the virus starts to replicate, your body will wake up and say, “Wait, I’ve seen this one before.”

Your body has a type of memory T-Cell that remembers the recipe for the antibody. (The antibody neutralizes the virus by “locking on” to its outside.) The memory B-Cells tells your immune system, “Hey! Start making this! Here’s the recipe that worked last time.”

The newly manufactured antibodies start neutralizing the newly growing viruses, and with luck snuff out the colony in the bud.

Best case, all this happens and you don’t even notice.

Which is what the average consumer wants. Except they think of immunity as a shield, a physical barrier, when it’s really more like a defense in depth.

(In World War I, those insane attacks across No Man’s Land most of the time did take the front line of the enemy’s trenches, but both sides kept floating reserves behind the lines to bring up in a counterattack, and toss the invaders out.)

In a middling case you might feel something and maybe have some of the old familiar symptoms, but it should not be anywhere as bad as the first time.

But note that the early days of a reinfection — even mild or even unnoticed — things may show up on tests looking the way they do in a first-time infection.

So, strictly speaking, yes, you can always get reinfected.

As to reactivation, I have my doubts. But I have to say it’s possible. We’ll see what the research comes up with. Reactivation would imply the virus has some place in the body where it can hide out, like HSV hides out in nerve cells.

Where is The Zombie going now?

South Korea.

Huh?

No, it’s good. South Korean researchers may be the ones to put a stake through his heart.

Here he is:

South Korean officials on Friday [April 10] reported 91 patients thought cleared of the new coronavirus had tested positive again.

He just won’t quit, will he?

I’m not too worried about the Zombie now. The KCDC is investigating. They are very competent. They’ve got it.

For one, they are taking the time to do The Right Thing.

See, we don’t care if the PCR test came back positive — if all it’s amplifying is viral debris.

We care whether the virus is “live” — can replicate. Then we need to isolate.

The time-honored test is trying growing it in a petri dish. And that’s what the South Korean researchers are taking the time to do.

A few early dispatches are encouraging:

The KCDC has re-investigated three cases from the same family where patients tested positive after recovering, Kwon says. In each of these cases, scientists tried to incubate the virus but weren’t able to — that told them there was no live virus present.

Stay tuned. More Zombie sightings will be added here, as they happen!

Update April 23, 2020

The South Korean researchers haven’t killed the Zombie, but it looks like they are getting him cornered.

A dispatch in the South China Morning Post (Hong Kong) reports that in the KCDC’s small group of test subjects (a) all have blood antibodies and (b) none have live virus, that is, they were unable to culture it.

Laboratory cultivation tests using samples from the 12 patients produced no virus, indicating the virus may have already died but the highly sensitive RT-PCR tests picked up its fragments, producing positive results.

But, they are doing it all over again. I told you they are good:

Yoo Cheon-kwon at the KCDC said a second cultivation test would verify whether the virus was alive or its genetic remnants were being detected at RT-PCR tests.

And an interesting number for anyone worried:

However, health authorities plan to extend self-quarantine for recovered patients beyond the current two weeks because in some extreme cases, cured patients tested positive in RT-PCR tests 35 days after being cured.

[Update 29 April 2020]

The last nail may have been driven into the Zombie’s coffin.

Or maybe it’s a stake through his heart. I’ve always been a little fuzzy about how you kill a Zombie.

Anyway, “Tests in recovered patients found false positives, not reinfections, experts say.”

I’ll quote the Zombie’s epitaph at length, to make sure he gets a proper send-off:

South Korea’s infectious disease experts said Thursday that dead virus fragments were the likely cause of over 260 people here testing positive again for the novel coronavirus days and even weeks after marking full recoveries.

Oh Myoung-don, who leads the central clinical committee for emerging disease control, said the committee members found little reason to believe that those cases could be COVID-19 reinfections or reactivations, which would have made global efforts to contain the virus much more daunting.

“The tests detected the ribonucleic acid of the dead virus,” said Oh, a Seoul National University hospital doctor, at a press conference Thursday held at the National Medical Center.

He went on to explain that in PCR tests, or polymerase chain reaction tests, used for COVID-19 diagnosis, genetic materials of the virus amplify during testing, whether it is from a live virus or just from fragments of dead virus cells that can take months to clear from recovered patients.

The PCR tests cannot distinguish whether the virus is alive or dead, he added, and this can lead to false positives.

“PCR testing that amplifies genetics of the virus is used in Korea to test COVID-19, and relapse cases are due to technical limits of the PCR testing.”

May 19, 2020. KCDC gets a last word

On 19 May, 2020. the South Korean CDC published a summary letter, Findings from investigation and analysis of re-positive cases.

“Virus isolation cell culture result was negative,” meaning they were unable to grow live virus from the fragments found in PCR-positive patients.

These recovered patients were not actually infected, nor were they contagious. And 96% were positive for neutralizing antibodies.

The last nail in the Zombie’s coffin was that the KCDC changed its terminology. Instead of calling these recovered patients “re-positive” (which implies who knows what), it now calls them “PCR re-detected after discharge from isolation.”

R.I.P.

Until next time.

You can follow these posts via Twitter @Will_Bates_sci.

Will Bates writes about science, technology, and business. His journalism has appeared in the New York Times, the Wall Street Journal, and numerous magazines.