[Skip to Navigation]
Sign In
Editor's Note
February 24, 2021

How to Advise Persons Who Are Antibody Positive for SARS-CoV-2 About Future Infection Risk

Author Affiliations
  • 1NYC Health and Hospitals, New York, New York
JAMA Intern Med. 2021;181(5):679. doi:10.1001/jamainternmed.2021.0374

As a physician working in New York, New York, where coronavirus disease 2019 (COVID-19) hit hard in March and April of 2020, people often ask me how to interpret their severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) antibody results. Many people have positive test results for the antibody, some of them received a diagnosis of COVID-19, some of them had symptoms that were consistent with COVID-19 but were never tested because of a limited availability of testing, and some were never symptomatic but learned that they were positive for the antibody on a subsequent laboratory test. If they are positive, they want to know whether they are protected from a future infection with the virus.

Underlying the question of whether the presence of antibodies provides protection against future infections are 3 questions: are antibodies protective, how good are the available tests for accurately detecting antibodies, and how long does protection last? To address the first question, we know that most patients who recover from COVID-19 have antibodies and that reinfection (as opposed to extended symptoms or ongoing viral shedding) is rare, at least at this date. However, even if antibodies are protective, there remains a question of how accurate commercial tests are for detecting antibodies.

The study in this issue of JAMA Internal Medicine by Harvey and colleagues1 provides reassuring answers to the first and second questions. Using a national database with more than 3 million unique patients, the authors found that patients with a positive antibody test result were more likely than those with a negative test result to have a subsequent positive nucleic acid amplification test result for SARS-CoV-2 in the first 30 days from the test result (viral shedding), but that starting at beyond 30 days, the risk of a positive nucleic acid amplification test declined every 30 days, until the risk ratio for those with an initial positive test at 90 days or greater follow-up was only 0.10 compared with those with a negative test. Antibody tests, in this study, appeared accurate and the antibodies protective. Their findings are consistent with a study of health care workers that found that the incidence of SARS-CoV-2 infection in 1265 workers with antispike antibodies was 0.13 per 10 000 days at risk compared with 1.09 for 11 364 workers who were seronegative for these antibodies.2

Unfortunately, neither study can answer how long antibody protection will last because of natural infection. For this reason, vaccination against SARS-CoV-2 is recommended regardless of antibody status. How long the antibody protection provided by vaccines will last is also unknown. To know how long protection will last with antibodies because of natural infection or vaccination is something only time will tell.

Back to top
Article Information

Published Online: February 24, 2021. doi:10.1001/jamainternmed.2021.0374

Corresponding Author: Mitchell H. Katz, MD, NYC Health and Hospitals, 125 Worth St, Room 514, New York, NY 10013 (mitchell.katz@nychhc.org).

Conflict of Interest Disclosures: None reported.

Harvey  RA, Rassen  JA, Kabelac  CA,  et al.  Association of SARS-CoV-2 seropositive antibody test with risk of future infection.   JAMA Intern Med. Published online February 24, 2021. doi:10.1001/jamainternmed.2021.0366Google Scholar
Lumley  SF, O’Donnell  D, Stoesser  NE,  et al; Oxford University Hospitals Staff Testing Group.  Antibody status and incidence of SARS-CoV-2 infection in health care workers.   N Engl J Med. 2020. doi:10.1056/NEJMoa2034545PubMedGoogle Scholar
3 Comments for this article
Vaccine prioritization – where do the facts really lead us?
Manish Joshi, MD | CAVHS
The COVID-19 pandemic continues to ravage humankind; the United States just reached a grim milestone of more than 500.000 deaths due to the disease. The arrival of 2 vaccines (Pfizer and Moderna) has promised some light at the end of the tunnel. Unfortunately, the supply is meager, and <15% of  the US population has been vaccinated to date. Prioritization – deciding who is “at the front of the line” - is of utmost importance. Katz suggests “vaccination against SARS-CoV-2 is recommended regardless of antibody status.” I would disagree. First, the data in the vaccine trials (both Pfizer and Moderna) do not suggest efficacy among participants with evidence of previous SARS-CoV-2 infection. (https://wwmt.com/news/nation-world/questions-arise-over-cdc-guidance-on-covid-19-vaccines). Second, the article in this issue of JAMA Internal Medicine that Katz is commenting on (https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2776810) provides strong evidence for protection due to natural infection.

Katz argues that the duration of immunity due to natural infection is unknown, but so is the duration of immunity due to available vaccines. In my opinion, the wisest application of current knowledge and reason would be to hold on vaccination of those who have already had COVID-19 and to give those doses to the most vulnerable – hospital workers, essential workers, the elderly, and those with pre-existing conditions such as obesity.

As a frontline ICU physician, I have witnessed more deaths than many of my other colleagues working in other departments. Swift and appropriate deployment of this valuable resource – the vaccine – can make a real difference. 
Inferences vs Clinical Results
Joe Psotka, PHD in Psychology | US Army Research Institute
I am concerned that so many physicians are unwilling to make obvious inferences from widely affirmed clinical understanding. Although specific clinical studies are the gold standard, obvious inferences from past understanding is the hallmark of clinical practice at the individual level. No one would have  qualms about making clinical judgments about disease and basing treatment on those judgments when treating an individual. Yet, when it comes to providing general advice to the public,  physicians tread lightly and refuse to make the same clinical judgments, as if speaking to the general public was  different from doing their best for individuals. Yet, this cannot be. The same advice must be followed in either case. So, in the instance of whether antibodies are essential to immune system functioning, on the whole the answer must be yes. The evidence is also clear that immunity persists for several months. In the study that the Editor's Note is about (https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2776810), immune effects continue well beyond 90 days. Current evidence for vaccines suggest a similar length of immunity beyond three months. Is should not be argued that we do not know how long immunity lasts when certain elements are clear already.
Infection with SARS-CoV-2 provides robust protection from re-infection out to at least 35 weeks.
Edward Powell, Ph.D. Astrophysics | Ed Powell Consulting
The preprint, "SARS-CoV-2 reinfection in a cohort of 43,000 antibody-positive individuals followed for up to 35 weeks" by Abu-Raddad (Cornell) et al. ( https://www.medrxiv.org/content/10.1101/2021.01.15.21249731v2 ) followed 43,044 patients who had confirmed COVID-19 cases for 35 weeks (though the study is ongoing) to determine the rate of reinfection. "Reinfection was next investigated using viral genome sequencing. Applying the viral-genome-sequencing confirmation rate, the risk of reinfection was estimated at 0.10% (95% CI: 0.08-0.11%).The incidence rate of reinfection was estimated at 0.66 per 10,000 person-weeks (95% CI: 0.56-0.78). Incidence rate of reinfection versus month of follow-up did not show any evidence of waning of immunity for over seven months of follow-up." And, "Reinfection is rare. Natural infection appears to elicit strong protection against reinfection with an efficacy >90% for at least seven months."

The authors are modest. Their data show that original infection is 99.9% protective of reinfection, whereas the various vaccines are only effective at 95%. ("Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine" (N Engl J Med 2021; 384:1576-1578 DOI: 10.1056/NEJMc2036242) (https://www.nejm.org/doi/full/10.1056/NEJMc2036242)

The appropriate response to a person with positive antibodies, and who had symptoms consistent with COVID-19 in the past, is that they should remain vigilant, but that they are as protected as if they had gotten the vaccine.