By Stan Deresinski, MD, FACP, FIDSA

Clinical Professor of Medicine, Stanford University

SYNOPSIS: A study of healthcare workers demonstrated that the presence of antibody to SARS-CoV-2 spike protein or to nucleocapsid provides strong protection against infection with this virus for up to six months.

SOURCE: 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; Dec 23:NEJMoa2034545. doi: 10.1056/NEJMoa2034545. [Online ahead of print].

Lumley and colleagues at the Oxford University Hospitals prospectively evaluated the incidence of the occurrence of SARS-CoV-2 positive polymerase chain reaction (PCR) tests in healthcare workers (HCWs) who, at baseline, either had or lacked antibody to the virus to assess the protection provided by the presence of antibody.

At baseline, 11,364 (90.6%) of 12,541 HCWs had negative tests for the presence of immunoglobulin G (IgG) anti-spike antibody, while 1,177 (9.4%) were seropositive and another 88 seroconverted during the 31 weeks of the study. HCWs were followed for a median of 200 days after a negative test and 139 days after a positive test. Follow-up testing of the 11,364 HCWs who were seronegative for antibody to spike protein at baseline detected 223 who subsequently had a positive PCR, for an incidence of 1.09 per 10,000 days at risk. The positive PCR occurred in the presence and absence of symptoms in 123 (55.2%) and 100 (44.8%) subjects, respectively.

Among the 1,265 who were seropositive (including the 88 who first became seropositive after the baseline assessment), only two subsequently had a positive PCR test, each at a time when they were asymptomatic, for an incidence of 0.13 per 10,000 days at risk. Comparing those who were seropositive to those who were seronegative, the risk ratio for a subsequent positive PCR test was 0.12 (95% confidence interval [CI], 0.03 to 0.47; P = 0.002). There were no symptomatic PCR-confirmed infections in the seropositive cohort, while, among the seronegatives, the incidence was 0.60 per 10,000 days at risk.

Similar risk ratio results were seen with analyses of anti-nucleocapsid IgG antibody. Confirming the protective role of antibodies, the incidence of occurrence of positive PCR tests was inversely correlated with the titers of antibody to the spike protein (P < 0.001). In fact, this protection extended to individuals who had detectable antibody but at levels too low to meet the threshold used to declare a test as positive.


This study confirms that the presence of antibody to SARS-CoV-2 as the result of natural infection is protective against subsequent infection, especially against symptomatic infections. This protective effect lasts at least six months, a finding consistent with recent evidence of the persistence of immunological memory for at least eight months.

These findings have important implications for understanding protection associated with COVID-19 vaccines. It also provides information that is valuable to the safety of HCWs with naturally acquired immunity as well as to their patients. The rare occurrence of a positive PCR test, however, indicates that protection is not complete (something already known from the very rare occurrence of reinfection in individuals who have recovered from COVID-19), and there is a potential danger that they may transmit infection even when they are asymptomatic.