Skip to Main Content

Reinfections from Covid-19 continue to seem rare, and are not responsible for the current, stubbornly high case counts in the United States, according to scientists and the latest findings.

At least, that’s what researchers are left to conclude. Experts say the country and individual states don’t have strong systems to determine how frequently people are getting reinfected — another consequence of the nation’s limited surveillance network. They’re calling for better data collection and analysis around second cases of Covid-19.


The main factors driving coronavirus transmission in the United States are a mix of the old — easing restrictions, people coming into close contact with others — and the new, like the more transmissible variants, experts say. And Caitlin Rivers, an infectious disease epidemiologist at the Johns Hopkins Center for Health Security, said she thought that reinfections are still uncommon.

But without better data, she said, “we don’t know, and I want to know.”

“It’s not the most likely explanation” for what’s spurring cases, she said about reinfections. “But I would like to see more evidence and data around that.”


The Centers for Disease Control and Prevention has outlined how suspected reinfections can be investigated and has a form for jurisdictions to report them, which Rivers called the “building blocks” for tracking cases. Still, she said, not a lot of data is being produced. The U.S. also doesn’t have the level of studies set up like those in England to track reinfections over time.

The notion of reinfections is not some unexpected twist in the rollercoaster pandemic. Scientists anticipate that with a respiratory infection like Covid-19, people won’t be protected for life after recovering from an initial case. Rather, the immunity generated from their first infection will wane over time. But one predominant hypothesis is that whatever reinfections occur with Covid-19 will generally (with some exceptions) be milder than the initial case, perhaps even asymptomatic. Even if the immune system can’t fend off the virus entirely, it still remembers the pathogen well enough to vanquish it quickly.

Researchers don’t yet have estimates for how long protection lasts after a Covid-19 infection, but there isn’t going to be one precise answer. Factors like age and even how sick people get from their initial case can influence that variability. Some people — including those who have really mild or asymptomatic infections the first go — don’t generate that strong of an immune response, and might become susceptible again sooner.

Already, there have already been several dozen documented reinfections around the world, though current numbers are almost certainly an undercount. If subsequent infections are indeed mild or asymptomatic, many of them will go undetected. It can also be difficult to establish two separate infections in the same person.

But so far, the available research suggests just a tiny subset of the more than 30 million people in the United States — and 132 million globally — who’ve had a confirmed Covid-19 case have become infected again. Studies indicate that the vast majority of people mount a robust and long-lasting immune response after being infected for the first time with SARS-CoV-2, the virus that causes Covid-19. A study that followed health care workers in U.K. for six months, for example, found that those who had an initial Covid-19 infection carried protective antibodies for the length of the study period; the few who tested positive again generally had no symptoms (it’s unclear, though, if they could still transmit the virus). A study out of Denmark also found reinfections were rare, though they were more common in people 65 and up.

“We anticipate reinfections will be a part of the epidemiology at some point, but I don’t think they’re accounting for the cases now in any major shape or form,” said Michael Diamond, a viral immunologist at Washington University in St. Louis. He noted in the U.S., the biggest wave of cases occurred only recently, at the end of 2020 and beginning of 2021.

“The majority of people who’ve had prior infection in the recent past” — six months, he said — “are going to have protection against” the forms of the virus that are dominant in the U.S. now.

At this point in the pandemic, there are two potential forces that could increase the number of reinfections, though experts say there’s not clear evidence of either occurring to a great extent.

For one, the first wave arrived in the U.S. more than a year ago. Perhaps enough time has passed that the immunity of some people infected early on has waned. Older people, for example, tend to have weaker immune responses — though at this point, that population is increasingly vaccinated, giving them an even stronger immune boost against the virus. (Health officials say that people who have had Covid-19 should still be vaccinated because the shots appear to generate greater levels of protection than infections.)

Michigan was one of the states hit early in the U.S. last spring, for example, and is again experiencing one of the country’s biggest outbreaks. But experts there think second cases are not playing any major role in the state’s surge.

“The large majority of what we’re seeing in Michigan is not reinfection,” said Anurag Malani, an infectious disease physician at St. Joseph Mercy Health System in Ann Arbor.

The other factor is the presence of certain newer versions of the coronavirus. Two variants that first appeared in South Africa and Brazil — called B.1.351 and P.1, respectively — as well as iterations that cropped up in California — B.1.427 and B.1.429 — appear to be able to sneak past certain elements of the immune response elicited after infections by other forms of the virus. It’s possible then that these variants are more likely to infect a person again who would otherwise be able to block other iterations of the virus entirely.

But Diamond noted that the number of cases of B.1.351 and P.1 remains low in the U.S., suggesting the variants are not spreading widely through reinfecting people. While just a small fraction of cases are sequenced, the available numbers indicate other strains of the virus far outnumber B.1.351 and P.1.

Even if someone contracts a second infection from a variant like B.1.351 or P.1, Diamond said his guess was that the immune system would generally still be able to recognize the virus well enough despite its genetic differences to fend off the most severe outcomes. He added, though, that scientists needed to verify that through additional research, and it’s something experts will be watching for as the variants continue to circulate.

“It may not prevent you from feeling lousy, it may not prevent you from transmitting,” Diamond said, “but it may prevent you from getting serious disease.”

  • Indeed, data collection and analysis on Covid disease and all its aspects could use vast improvements in the US. It is weird that the nation of MicroSoft, Google, etc has not capitalized on recruiting their home-grown IT specialists for serious data assembly and compilation. Huge chunks of knowledge are missed through lack of innovative thinking, and for “patient privacy concerns” . If the power of its IT specialist would be uitilized, the US would not be still in lala-land about re-infections, immunity longevity, long-term effects etc.

  • According to the Wall Street Journal and other media, there are or have been 800+ cases of P1 variant in British Columbia.
    The Canadian vaccine rollout was reported to be behind ours – but it is still not certain this variant is stopped by our vaccines, and even if it is, and we are ahead, it is reported to be several times as contagious as our original type, so we may not outrun it.
    It would be nice to know we are expected to have effective vaccines against it soon.

  • This is so tiresome. Once again, it’s “experts say.” Once again the “experts” and the journalists fall back on the conditional mood and qualifying phraseology, to wit:

    “potential forces that could. . .”
    “It’s possible then. . .
    “Perhaps enough time has passed”
    “it may not prevent”
    “it’s not the most likely”
    “appear to be able”
    “might become susceptible again”

    We don’t need journalists to write this stuff — we can get a Google Algorithm to do the scut work of assembling “expert” quotes and then Andrew and his fellow STAT journos can go do something productive — like wait tables, or code.

    Of course, all our epidemiological betters want better data. Well, duh! The problem, of course, is more or less intractable. It’s not about the quantity of data, it’s whether the data is sufficient in quality and quantity to confirm a falsifiable hypothesis.

    And given the incredibly noisy, miniscule, incomplete, and and inconclusive data about the scary “variants” we got NOTHIN’. Except for anecdotes, suppositions, guesses, mights, and coulds, woulds, and possibles.

    In the meantime lockdowns, social restrictions, and unwarranted fears do their incontrovertible damage.

  • Interesting that the narrative is always “we do not have the data” for questions as to reinfection rates and lifting restrictions after fully vaccinated but never for questions relating to the short and long term safety of new mRNA technology.

    • “Interesting” as in why the F is the richest nation on the planet not testing sufficiently even after all this time into the pandemic? It is definitely interesting the we have a fractured and sub optimal healthcare system that has been strained and exposed and still very few seem to want to make it better.

      Also, understand that some answers can only be known in retrospect regardless of how much data is collected. The big one: when is this over?/when do we reach herd immunity? cannot be known until after it happens. Could be 70%, could be 99%, could be that it is impossible to eradicate entirely and we need booster shots every 6 months. Time will tell. For now, the virus is clearly winning and I am disappointed the US is rushing to reopen everything in the face of rising case rates and more infectious variants.

  • I see nothing to make me worry about reinfection with the same variant, but reinfection by P1 seems to be very common in Brazil. And I can find no hard data indicating the vaccines we use here will stop it.
    South African reports were of high numbers of reinfections and young people becoming seriously ill. Pfizer had clinical trial data indicating their vaccine works against that one, B1351, but nothing on P1 so far.
    And I can find nothing on Moderna vaccine efficacy against any variant, so far. Johnson and Johnson, some efficacy against South African type, but nothing on Brazilian
    It is a big mess. I do not understand why we did not send people into Manaus and South Africa with our vaccines and do a study soon after their outbreaks, or at least after variants were suspected to be causing them.
    Without more hard data, it seems like we do not know if our vaccination program here is stopping the virus, or just delaying it until the variants get to higher numbers.

Comments are closed.