Skip to Main Content

After almost a year of pandemic terror, the end is in sight. But you still have to squint.

The FDA has granted emergency use authorization for two safe and effective vaccines that science has delivered at record speed. The question now is: How do we best distribute them?

The Centers for Disease Control and Prevention’s Advisory Committee on Immunization Practices (ACIP) has published guidance that vaccinations should start with health care personnel and residents of long-term care facilities, followed by other essential frontline workers and those over 75 years old. Mentioned only as a subpriority is how a history of Covid-19 infection should affect one’s place in line: “HCP with documented acute SARS-CoV-2 infection in the preceding 90 days may choose to delay vaccination until near the end of the 90-day period in order to facilitate vaccination of those HCP who remain susceptible.”


Given the low risk for reinfection and the limited supply of vaccine doses, it would be a mistake not to make previous infection a more central consideration in our vaccine prioritization. With an estimated 75 million Americans having already been infected with SARS-CoV-2, but only 24 million knowing it, using wide-scale Covid-19 antibody testing can help better target vaccine allocation to those at highest risk. Doing so can save lives and return us to normalcy sooner.

This strategy builds upon the two biggest discoveries made in efforts against the virus. The first is that after infection, including mild and asymptomatic infections, there appears to be lasting and strong immunity out to six-plus months. The fact that there have been nearly 100 million cases of Covid-19 confirmed worldwide and only a handful of documented reinfections provides convincing evidence of lasting immunity. And even among the rare reinfections, their course will likely be milder thanks to the immune system’s memory.


The second breakthrough is the resounding success of Covid-19 vaccine development.

This combination of lasting immunity and effective vaccines has been the cornerstone of almost all past successes against viruses (HIV, to date, being the key exception). It’s how the scourges of smallpox, polio, measles, mumps, and other infectious diseases have been beaten. And it’s how we are going to beat Covid-19.

But even in a best-case scenario, it will be months before enough vaccine doses have been made to treat everyone. With epidemiologists estimating that two-thirds of the population must be immune for the herd protection needed to quell the pandemic, an antibody-assisted approach would let us reach that threshold faster.

Here’s another reason why an antibody-assisted approach to vaccination is needed: Due to the combination of inadequate testing and asymptomatic infection, most people infected with Covid-19 are never diagnosed with it. This is especially true in states hardest hit by the virus. In New York state, for example, it is estimated that 30% of the population has recovered from Covid-19 while only 7% have been diagnosed with the virus. Underdiagnosis isn’t limited to locales like New York, which had an early surge. More than 36% of North Dakotans are estimated to have been infected, while only 13% have been diagnosed. Given these discrepancies, in states like North Dakota, without the assistance of antibody testing, I estimate that as many as 1 in 4 vaccines could be given to someone who is currently immune to Covid-19.

While the presence of antibodies is not a perfect measure of immunity, thanks to both the rarity of reinfection and the accuracy of current antibody testing (with false positive rates around 1% or lower), those with antibodies can safely be considered a low-risk group. This reality was further confirmed in a recent New England Journal of Medicine report from Oxford University that followed 12,000 health care workers over six months and found no symptomatic infections in those with antibodies to SARS-CoV-2.

But theory and in practice are two different things. With the difficulty the U.S. has had scaling PCR testing, and with early vaccine distribution sputtering, efforts to test swaths of the public for antibodies may sound foolhardy. It isn’t.

In regards to scaling antibody testing, the process is wholly different from the PCR-based testing used to detect acute infection. Antibody tests are more like traditional bloodwork and are processed as automated immunoassays. This means they can be run in large batches on machines almost all functional medical laboratories already own and can use existing lab collection infrastructure for collection and processing. As Benjamin Mazer, a pathologist at Johns Hopkins Hospital, told me, “The delays we’ve faced with PCR testing shouldn’t deter people from antibody tests if they’re needed. The antibody test is far simpler to perform and can be turned around in hours instead of days.”

An easy place to start would be testing for antibodies in people already requiring lab tests for other reasons, such as when they are admitted to a hospital, at the emergency department, or have a clinic appointment. Standing orders paired with canceled copayments for others at clinical and commercial laboratories can further expand access. School- and employer-based batch testing can inform their future vaccination campaigns.

To be clear, it is both safe and beneficial for those previously infected with SARS-CoV-2 to be vaccinated (just like adults who had the chickenpox need a booster to prevent shingles). It is paramount proper investments be made to support both testing and vaccination. These efforts must be complements, not competitors. And if access to antibody testing is not readily accessible, vaccination should never be delayed. Lastly, once we have enough supply to meet public demand, everyone should be vaccinated, regardless of antibody status.

I could close with an argument about how an antibody-assisted approach would allow the U.S. to reach herd immunity faster. Or revive our economy quicker. Or protect more frontline workers — nurses, teachers, grocers, delivery drivers, firefighters, and others — sooner.

But for me, and I suspect for you too, it’s much less abstract than that. For every vaccine we save by using antibody testing, there will be one more we can give to a higher-risk individual anxiously awaiting her or his turn in line. And we all have loved ones standing in line: an elderly grandparent, an immunocompromised mom, or a cousin fighting cancer.

Given all we have done so far to keep them safe — deferred dining, canceled vacations, and missed hugs — we must employ every weapon in our armamentarium against this plague. This includes antibody testing.

Michael Rose is a resident physician in internal medicine and pediatrics at Johns Hopkins University School of Medicine.

  • Both my husband and I contracted COVID 19 on June 19, 2020. We both had mild symptoms and recovered 5 days later. We also developed antibodies. Struggling with the decision of taking the vaccine.

  • [disclaimer: engineer here, not biomed specialist. Also: aerospace engineer, so used to think about most conservative, worst case scenarios!] I think that the idea makes a lot of sense. The issue is how quickly it could get implemented. I assume there are levels of immune system protection, it’s not white/black yes/no. So for a data driven decision, shouldn’t there be if not a full clinical trial, at least a retrospective study to quantify that level? In the absence of that critical piece information, I would assume that my antibody level is not high enough and I would line up for a vaccine just in case.

    Also those data would have to be digested by the local health authorities, converted into guidance, and implemented in the vaccination schedule/appointment software, and I wouldn’t expect that to happen quickly.

  • I have been writing about this for weeks, trying to get someone to listen. I am 69 and had covid in March. I decided to have an antibody test 3 weeks ago and my number is still high. Now people have to start talking about memory B cells.

  • Thank you for your thoughtful comments Steve. I agree the strategy’s merits would need to be reconsidered if a variant indeed was found to evade “natural” immunity but not vaccine immunity. I too am eagerly awaiting such data collection and share in the collective “fear” of immune escape. I do hope that there would be some immunity, even if imperfect, provided by current vaccines and/or previous infection but time and clinical data will tell.

  • As I said in the previous Comment, the behavior of the South African variant is crucial, if it gets here – if it will infect people who were previous infected with the first strain, but not those who got the vaccines, we need to just give out the vaccines as planned – or, maybe, since this one is supposed to be more contagious, change the prioritization to try to get the spreaders more immune – people who come into contact with large numbers of other people, not necessarily the most vulnerable.
    I also want to say – we need to ban travel from South Africa – Israel, UK, Germany, Turkey, Taiwan, Hong Kong all reportedly have done so- if the variant there is as bad as feared we will have another epidemic on our hands, but this one will be far worse.

  • Another possible problem – the reports out of South Africa of the variant there is that it may escape immunity conferred by prior infection, AND be much more contagious, AND attack young people with severe illness.
    But so far I see fear, but not confirmation, it will evade the vaccines we have.
    If that variant does get here, but does not evade vaccine derived immunity, then doling out the vaccine only to those with no prior immunity would not be a good strategy.

  • I do not at all disagree with Dr. Rose but I have to wonder if we can get this together fast enough to make it a more effective strategy than vaccinating whoever is willing to show up. In California, the ongoing surge is both more reason to target the vaccines to the highest risk – and maybe also less? – because if we can get a lot of them out, and suppress the surge, that will also help the most vulnerable right? It might make sense for an organized organization to set priorities and use the tools on hand to carry them out – but that does not seem to be what is happening here.

Comments are closed.