Cogito, ergo sum
„Effects of Previous Infection and Vaccination on Symptomatic Omicron Infections
List of authors.
Heba N. Altarawneh, M.D., Hiam Chemaitelly, Ph.D., Houssein H. Ayoub, Ph.D., Patrick Tang, M.D., Ph.D., Mohammad R. Hasan, Ph.D., Hadi M. Yassine, Ph.D., Hebah A. Al-Khatib, Ph.D., Maria K. Smatti, M.Sc., Peter Coyle, M.D., Zaina Al-Kanaani, Ph.D., Einas Al-Kuwari, M.D., Andrew Jeremijenko, M.D., et al.
Article
Figures/Media
Metrics
36 References
Abstract
BACKGROUND
The protection conferred by natural immunity, vaccination, and both against symptomatic severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection with the BA.1 or BA.2 sublineages of the omicron (B.1.1.529) variant is unclear.
METHODS
We conducted a national, matched, test-negative, case–control study in Qatar from December 23, 2021, through February 21, 2022, to evaluate the effectiveness of vaccination with BNT162b2 (Pfizer–BioNTech) or mRNA-1273 (Moderna), natural immunity due to previous infection with variants other than omicron, and hybrid immunity (previous infection and vaccination) against symptomatic omicron infection and against severe, critical, or fatal coronavirus disease 2019 (Covid-19).
RESULTS
The effectiveness of previous infection alone against symptomatic BA.2 infection was 46.1% (95% confidence interval [CI], 39.5 to 51.9). The effectiveness of vaccination with two doses of BNT162b2 and no previous infection was negligible (−1.1%; 95% CI, −7.1 to 4.6), but nearly all persons had received their second dose more than 6 months earlier. The effectiveness of three doses of BNT162b2 and no previous infection was 52.2% (95% CI, 48.1 to 55.9). The effectiveness of previous infection and two doses of BNT162b2 was 55.1% (95% CI, 50.9 to 58.9), and the effectiveness of previous infection and three doses of BNT162b2 was 77.3% (95% CI, 72.4 to 81.4). Previous infection alone, BNT162b2 vaccination alone, and hybrid immunity all showed strong effectiveness (>70%) against severe, critical, or fatal Covid-19 due to BA.2 infection. Similar results were observed in analyses of effectiveness against BA.1 infection and of vaccination with mRNA-1273.
CONCLUSIONS
No discernable differences in protection against symptomatic BA.1 and BA.2 infection were seen with previous infection, vaccination, and hybrid immunity. Vaccination enhanced protection among persons who had had a previous infection. Hybrid immunity resulting from previous infection and recent booster vaccination conferred the strongest protection. (Funded by Weill Cornell Medicine–Qatar and others.)
Qatar endured a wave of the omicron (B.1.1.529) variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)1 that started on December 19, 2021, and peaked in mid-January 2022.2-4 The wave was first dominated by the BA.1 subvariant, but within a few days after the onset of the wave, the BA.2 subvariant predominated (Fig. S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org). Although BA.1 and BA.2 remain classified as subvariants of omicron, considerable genetic distance exists between them.5 The protection against these subvariants provided by previous immunity — and whether immunity is induced by previous infection, vaccination, or a hybrid of both — remains to be established.
With the use of data from December 23, 2021, through February 21, 2022, we investigated the protection conferred by previous infection from variants other than omicron, vaccination with two or three doses of the coronavirus disease 2019 (Covid-19) messenger RNA (mRNA) vaccines BNT162b2 (Pfizer–BioNTech)6 or mRNA-1273 (Moderna),7 and hybrid immunity (previous infection and vaccination). Effectiveness against symptomatic BA.1 infection, symptomatic BA.2 infection, and any symptomatic omicron infection was assessed. Protection against any severe (acute-care hospitalization),8 critical (hospitalization in an intensive care unit),8 or fatal9 case of Covid-19 due to BA.1, BA.2, or any omicron infection was also assessed.
Methods
STUDY POPULATION AND DATA SOURCES
The study was conducted in the resident population of Qatar. We analyzed information from the national, federated databases regarding Covid-19 vaccination, laboratory testing, hospitalization, and death. These data were retrieved from the integrated nationwide digital-health information platform. The databases included all SARS-CoV-2–related data and associated demographic information since the start of the pandemic. These databases include, with no missing information, results of all polymerase-chain-reaction (PCR) testing and, more recently, rapid antigen testing conducted at health care facilities on or after January 5, 2022.
All PCR testing (but not rapid antigen testing) performed in Qatar is classified on the basis of symptoms and the reason for testing. Of all the PCR tests conducted during this study, 19.2% were performed because of clinical symptoms. Qatar has an unusually young, diverse population — only 9% of its residents are 50 years of age or older, and 89% are expatriates from more than 150 countries.10 Qatar launched its Covid-19 vaccination program in December 2020 with the BNT162b2 and mRNA-1273 vaccines.11 Further descriptions of the study population and the national databases have been reported previously.4,10-15
STUDY DESIGN
The study assessed the effectiveness of previous infection, vaccination with BNT162b2 or mRNA-1273, and hybrid immunity (previous infection and vaccination) against symptomatic infection with BA.1, BA.2, and any omicron infection.2,15-18 We used a test-negative, case–control design, in which effectiveness estimates were derived by comparing the odds of previous infection or vaccination or both among case participants (persons with a positive PCR test) with that among controls (PCR-negative persons).2,15-18 We also assessed effectiveness against any severe, critical, or fatal case of Covid-19.
To estimate the effectiveness against symptomatic infection, we exactly matched cases and controls that were identified from December 23, 2021, through February 21, 2022. Case participants and controls were matched in a 1:1 ratio according to sex, 10-year age group, nationality, and calendar week of PCR test. Matching was performed to control for known differences in the risk of SARS-CoV-2 exposure in Qatar.10,19,20 Matching according to these factors was previously shown to provide adequate control of differences in the risk of SARS-CoV-2 exposure in studies of different designs, all of which involved control groups, such as test-negative, case–control studies.11,12,15,21,22 To assess effectiveness against any severe, critical, or fatal case of Covid-19, we used a 1:5 matching ratio to improve the statistical precision of the estimates.
Only the first PCR-positive test that was identified for an individual participant during the study period was included, but all PCR-negative tests were included. Controls included persons with no record of a PCR-positive test during the study period. Only PCR tests conducted because of clinical symptoms were used in the analyses.
SARS-CoV-2 reinfection is conventionally defined as a documented infection that occurs at least 90 days after an earlier infection, to avoid misclassification of prolonged PCR positivity as reinfection if a shorter time interval is used.2,23 Previous infection was therefore defined as a PCR-positive test that occurred at least 90 days before the PCR test used in the study. Tests for persons who had PCR-positive tests that occurred within 90 days before the PCR test used in the study were excluded. Accordingly, previous infections in this study were considered to be due to variants other than omicron, since they occurred before the omicron wave in Qatar.2-4
PCR tests for persons who received vaccines other than BNT162b2 or mRNA-1273 and tests for persons who received mixed vaccines were excluded from the analyses. Tests that occurred within 14 days after a second dose or 7 days after a third dose of vaccine were excluded. These inclusion and exclusion criteria were implemented to allow for build-up of immunity after vaccination4,14 and to minimize different types of potential bias, as informed by earlier analyses in the same population.12,22 Every control that met the inclusion criteria and that could be matched to a case was included in the analyses.
We compared five groups with the group that had no previous infection and no vaccination. The five groups were characterized by type of exposure: previous infection and no vaccination, two-dose vaccination and no previous infection, two-dose vaccination and previous infection, three-dose vaccination and no previous infection, and three-dose vaccination and previous infection. The groups were defined on the basis of the status of previous immunologic events (previous infection or vaccination) at the time of the PCR test.
Classification of severe,8 critical,8 and fatal9 Covid-19 cases followed World Health Organization guidelines, and assessments were made by trained medical personnel with the use of individual chart reviews as part of a national protocol applied to hospitalized patients with Covid-19. Details regarding Covid-19 severity, criticality, and fatality classification are provided in Section S1 in the Supplementary Appendix.
LABORATORY METHODS AND SUBVARIANT ASCERTAINMENT
The large omicron wave in Qatar started on December 19, 2021, and peaked in mid-January 2022.2-4 A total of 315 random SARS-CoV-2–positive specimens collected from December 19, 2021, through January 22, 2022, underwent viral whole-genome sequencing on a GridION sequencing device (Nanopore Technologies). Of these specimens, 300 (95.2%) were confirmed to be omicron infections and 15 (4.8%) to be delta (or B.1.617.2)1 infections.2-4 Of the 286 omicron infections with confirmed subvariant status, 68 (23.8%) were BA.1 and 218 (76.2%) were BA.2.
We used the TaqPath COVID-19 Combo Kit (Thermo Fisher Scientific), which tests for the spike (S) gene of SARS-CoV-2 and the 69-70del mutation in the S gene,24 to identify BA.1 and BA.2 infections. An S-gene target failure was used as a proxy for BA.1 infection, and a non–S-gene target failure was used as a proxy for BA.2 infection. Additional details regarding laboratory methods for real-time reverse-transcriptase–quantitative PCR testing are provided in Section S2.
OVERSIGHT
This retrospective study was approved by the institutional review boards at Hamad Medical Corporation and Weill Cornell Medicine–Qatar, with a waiver of informed consent. The reporting of this study follows the Strengthening the Reporting of Observational Studies in Epidemiology guidelines (Table S1). The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript. All the authors contributed to data collection and acquisition, discussion and interpretation of the results, and the writing of the manuscript. All the authors read and approved the final manuscript.
STATISTICAL ANALYSIS
Although all records of PCR testing were examined for selection of cases and controls, only matched samples were analyzed. Cases and controls were described with the use of frequency distributions and measures of central tendency and compared with the use of standardized mean differences. The standardized mean difference was defined as the difference between the mean value of a covariate in one group and the corresponding mean value of a covariate in the other group, divided by the pooled standard deviation, with values of less than 0.1 indicating adequate matching.25
Odds ratios, which compared the odds of previous infection or vaccination or both among cases with that among controls, and associated 95% confidence intervals were derived with the use of conditional logistic regression. This analytic approach, which also incorporated matching according to calendar week of PCR test, minimizes potential bias due to variation in epidemic phase16,26 and roll-out of vaccination during the study period.16,26 Confidence intervals were not adjusted for multiplicity and therefore should not be used to infer definitive differences among exposure groups. Interactions were not investigated. Effectiveness and associated 95% confidence intervals were calculated as 1 minus the odds ratio of previous infection or vaccination or both among cases as compared with controls.16,17 The reference group for all estimates included persons with no previous infection and no vaccination.
An additional analysis was conducted to investigate the effects of previous infection, two-dose vaccination, and three-dose vaccination as a function of time since the immunologic event (previous infection or vaccination). This analysis used the same approach as the primary analysis, but with stratification according to time since the most recent immunologic event.
A person was considered to have had a previous positive test if that test was positive by PCR assay. A sensitivity analysis of effectiveness against any symptomatic omicron infection was conducted, but with previous positive testing being based on positive PCR as well as positive rapid antigen tests, to investigate whether exclusion of rapid antigen–positive tests could have biased our estimates. Statistical analyses were performed with the use of Stata/SE software, version 17.0 (StataCorp).
Results
STUDY POPULATION
From December 23, 2020 (the date that vaccination began in Qatar), through February 21, 2022 (the end of the study), 1,306,862 persons received at least two doses of BNT162b2, and 341,438 of these received a third (booster) dose. The median date of the first dose was May 3, 2021, the median date of the second dose was May 24, 2021, and the median date of the third dose was December 25, 2021. The median interval between the first and second doses was 21 days (interquartile range, 21 to 22), and between the second and third doses was 251 days (interquartile range, 233 to 274). The narrow interquartile range between the first and second doses reflects strict adherence to national policy.
During the study period, 893,671 persons received two doses of mRNA-1273, and 135,050 of these received a third dose. The median date of the first dose was May 28, 2021, the median date of the second dose was June 27, 2021, and the median date of the third dose was January 12, 2022. The median interval between the first and second doses was 28 days (interquartile range, 28 to 30), and between the second and third doses was 236 days (interquartile range, 213 to 260).
Table 1.
Characteristics of the Matched Case Participants and Controls According to Omicron Infection in the BNT162b2 Analysis.
The study was based on the total population of Qatar; therefore, the population is representative of the internationally diverse, but young and predominantly male, population of the country (Table S2). Figure S2 shows the process for selecting the populations for the analysis of BNT162b2, and Table 1 shows the characteristics of these populations. Figure S3 shows the process for selecting the populations for the analysis of mRNA-1273, and Table S4 shows the characteristics of these populations.
EFFECTIVENESS OF PREVIOUS INFECTION AND BNT162B2 VACCINATION AGAINST BA.1 INFECTION
Figure 1.
Effectiveness of Previous Infection, Vaccination with BNT162b2, and Hybrid Immunity against Symptomatic Omicron BA.1 and BA.2 Infection and against Severe, Critical, or Fatal Covid-19.
Table 2.
Effectiveness of Previous Infection, Vaccination with BNT162b2, and Hybrid Immunity against Symptomatic Omicron Infections and against Severe, Critical, or Fatal Covid-19.
The effectiveness of previous infection and no vaccination against symptomatic BA.1 infection was 50.2% (95% confidence interval [CI], 38.1 to 59.9) (Figure 1A and Table 2). The median interval between the previous infection and the PCR test used in the study was 324.5 days (range, 91 to 643; interquartile range, 274 to 497).
The effectiveness of two doses of BNT162b2 and no previous infection was negligible (−4.9%; 95% CI, −16.4 to 5.4). The median interval between the second dose and the PCR test used in the study was 268 days (range, 15 to 394; interquartile range, 211 to 293). The effectiveness of three doses and no previous infection was 59.6% (95% CI, 52.9 to 65.3). The median interval between the third dose and the PCR test used in the study was 42 days (range, 7 to 291; interquartile range, 28 to 62).
The effectiveness of hybrid immunity (previous infection and two doses of BNT162b2) was 51.7% (95% CI, 43.5 to 58.7), which was similar to the effectiveness of previous infection alone. The effectiveness of previous infection and three doses of BNT162b2 was the highest, at 74.4% (95% CI, 63.4 to 82.2).
Previous infection, vaccination, and hybrid immunity all showed strong effectiveness (>90%) against severe, critical, or fatal Covid-19 due to BA.1 infection, but some of the 95% confidence intervals were wide because of small case numbers (Figure 1B and Table 2). The severity of BA.1 infections was low, and only 0.3% (95% CI, 0.2 to 0.4) of infections progressed to severe, critical, or fatal Covid-19.
EFFECTIVENESS OF PREVIOUS INFECTION AND BNT162B2 VACCINATION AGAINST BA.2 INFECTION
The effectiveness of previous infection and no vaccination against symptomatic BA.2 infection was 46.1% (95% CI, 39.5 to 51.9) (Figure 1C and Table 2). The median interval between the previous infection and the PCR test used in the study was 319 days (range, 90 to 662; interquartile range, 275 to 499).
The effectiveness of two doses of BNT162b2 and no previous infection was negligible (−1.1%; 95% CI, −7.1 to 4.6). The median interval between the second dose and the PCR test used in the study was 270 days (range, 14 to 399; interquartile range, 213 to 296). The effectiveness of three doses of BNT162b2 and no previous infection was 52.2% (95% CI, 48.1 to 55.9). The median interval between the third dose and the PCR test used in the study was 43 days (range, 7 to 322; interquartile range, 26 to 65).
The effectiveness of previous infection and two doses of BNT162b2 was 55.1% (95% CI, 50.9 to 58.9), which is similar to the effectiveness of previous infection alone. The effectiveness of previous infection and three doses of BNT162b2 was the highest, at 77.3% (95% CI, 72.4 to 81.4).
Previous infection, vaccination, and hybrid immunity all showed strong effectiveness (>70%) against severe, critical, or fatal Covid-19 due to BA.2, but some of the 95% confidence intervals were wide because of small case numbers (Figure 1D and Table 2). The severity of BA.2 infections was low, and only 0.3% (95% CI, 0.2 to 0.3) of infections progressed to severe, critical, or fatal Covid-19.
EFFECTIVENESS OF PREVIOUS INFECTION AND BNT162B2 VACCINATION AGAINST ANY OMICRON INFECTION
Figure 2.
Effectiveness of Previous Infection, Vaccination with BNT162b2 or mRNA-1273, and Hybrid Immunity against Any Symptomatic Omicron Infection and against Severe, Critical, or Fatal Covid-19.
The effectiveness of previous infection, BNT162b2 vaccination, and hybrid immunity against any symptomatic omicron infection showed similar patterns to those against BA.1 and BA.2 (Figure 2A and Table 2). The effectiveness against severe, critical, or fatal Covid-19 due to any omicron infection also showed similar patterns to those against these outcomes due to BA.1 and BA.2 (Figure 2B and Table 2).
Figure 3.
Effectiveness of Previous Infection, Vaccination, and Hybrid Immunity against Any Symptomatic Omicron Infection According to Time since Previous Infection or Vaccination.
The analysis of the effectiveness of previous infection, two-dose vaccination, and three-dose vaccination as a function of time since the immunologic event (previous infection or vaccination) showed rapidly waning vaccine protection after the second and third doses but slowly waning protection from previous infection (Figure 3). A sensitivity analysis in which previous positive testing included both PCR-positive and rapid antigen–positive results showed similar findings to those of the main analyses, which indicates that exclusion of previous rapid antigen–positive tests may not have biased our estimates (Table S3).
EFFECTIVENESS OF PREVIOUS INFECTION AND MRNA-1273 VACCINATION AGAINST BA.1, BA.2, AND ANY OMICRON INFECTION
Figure 4.
Effectiveness of Previous Infection, Vaccination with mRNA-1273, and Hybrid Immunity against Symptomatic Omicron BA.1 and BA.2 Infection and against Severe, Critical, or Fatal Covid-19.
The effectiveness of previous infection, vaccination, and hybrid immunity in the analysis of mRNA-1273 showed similar patterns to those of the analysis of BNT162b2 (Figure 2 and Figure 4). Additional information is provided in Table S5.
Discussion
Previous infection with a variant other than omicron was associated with an approximately 50% reduced risk of infection. No difference in the protection of previous infection against BA.1 and BA.2 was discernable. Two-dose vaccination and no previous infection had negligible effectiveness against BA.1 and BA.2, but most persons received their second dose more than 8 months earlier. These findings are explained by the short-lived protection of primary-series vaccination against omicron infections3,27 and the more durable protection from natural infection,2,28 as confirmed by the additional analysis of protection as a function of time after previous infection or vaccination (Figure 3).
Booster vaccination was associated with an approximately 60% reduced risk of infection. No difference in the protection of booster vaccination against BA.1 and BA.2 was discernable. However, most persons received their third dose less than 45 days earlier, perhaps explaining the relatively high effectiveness.3
The protection conferred by hybrid immunity of previous infection and two-dose vaccination was similar to that of previous infection alone, at approximately 50%, which suggests that this protection originated from the previous infection and not from vaccination. This finding is also explained by the short-lived protection of primary-series vaccination against omicron infections.3,27
However, the highest effectiveness was seen with hybrid immunity from previous infection and recent booster vaccination (approximately 80%). This finding provides evidence for the benefit of vaccination, even for persons with a previous infection. Strikingly, this protection is what one would expect if previous infection and booster vaccination each acted independently. Because previous infection reduced the risk of infection by 50% and booster vaccination reduced it by 60%, the reduction in the risk of infection for both combined, if they acted fully independently, would be 1−(1−0.5)×(1−0.6)=0.8, which is an 80% reduction, just as observed. Although this effect needs to be further investigated, this finding may suggest that the combined effect of these two forms of immunity against omicron infection reflects neither synergy nor redundancy of the individual biologic effects of each.
Even though the five forms of immunity investigated showed large differences in protection against symptomatic infection that ranged from 0 to 80%, they all showed strong protection against Covid-19–related hospitalization and death, at an effectiveness of more than 70%. This suggests that any form of previous immunity, whether induced by previous infection or vaccination, is associated with strong and durable protection against Covid-19–related hospitalization and death. Notably, there was no evidence for a difference in severity between BA.1 and BA.2 infections in the study samples.
No notable differences were observed between the effects of BNT162b2 and mRNA-1273 vaccination. The results confirmed other findings that we reported recently, including a protection of approximately 50% for previous infection against reinfection with BA.1,2 a protection of approximately 50% for mRNA boosters as compared with primary series,4 and the finding that mRNA vaccines have negligible effectiveness against omicron infection 6 or more months after the second dose.3
This study has limitations. Ascertainment of BA.1 and BA.2 infections was based on proxy criteria, but this method of ascertainment is well established.24,29,30 Some omicron infections may have been misclassified delta infections, but the incidence of delta was limited during the study period (Section S2). Ascertainment of BA.1 and BA.2 infections was not possible for a minority of infections. However, this may not have biased our results, since both infections with and without BA.1 or BA.2 ascertainment had a similar distribution among exposure categories (Table S6).
Although matching was performed according to sex, age, and nationality, matching was not possible for other factors, such as coexisting conditions. However, matching according to these factors provided demonstrable control of bias in our earlier studies.11,12,15,21,22 The analysis of effectiveness according to time since the most recent immunologic event is possibly at higher risk than the primary analysis for bias because of confounding, since persons who were vaccinated earliest were more likely to have coexisting conditions or to work in high-risk occupations. Effectiveness was assessed with the use of an observational, test-negative, case–control design rather than a design in which cohorts of individual persons were followed up. However, the cohort study design applied earlier to the same population yielded findings similar to those of the test-negative design.14,15,31 Moreover, our recent study of the effectiveness of boosters relative to primary series used a cohort study design and generated results consistent with the results reported here.4
Nonetheless, one cannot rule out the possibility that in real-world data, bias could arise in unexpected ways or from unknown sources, such as subtle differences or changes in test-seeking behavior. For example, with the large omicron wave, use of rapid antigen testing was expanded to supplement PCR testing in Qatar starting on January 5, 2022, and especially so for some of the routine testing such as post-travel testing. However, rapid antigen testing was broadly implemented and probably did not differentially affect PCR testing to introduce bias, as supported by the sensitivity analysis (Table S3) and the minimal differences between PCR and rapid antigen tests according to exposure category (Table S7). With the small proportion of the population of Qatar being 50 years of age or older,10 our findings may not be generalizable to other countries in which elderly citizens constitute a larger proportion of the population.
Notwithstanding these limitations, findings were consistent with those of other studies of vaccine effectiveness against omicron infection (BA.1 or BA.2 subvariants were not specified).27,32-36 Moreover, with the mass scale of PCR testing in Qatar,12 the likelihood of bias is perhaps minimized. Extensive sensitivity and additional analyses were conducted to investigate effects of potential bias in our earlier studies that used similar methods. These included different adjustments and controls in the analysis and different study inclusion and exclusion criteria to investigate whether effectiveness estimates could have been biased.12,22 These analyses showed consistent findings.2,3,12,17,22
No notable differences were observed in the effectiveness against BA.1 and BA.2 of previous infection, vaccination, and hybrid immunity. Protection from previous infection with variants other than omicron against reinfection was moderate and durable, but protection of primary-series vaccination against infection was negligible by 6 months after the second dose. Recent booster vaccination had moderate effectiveness, whereas hybrid immunity from previous infection and recent booster vaccination conferred the strongest protection against infection, at approximately 80%. All five forms of immunity were associated with strong and durable protection against Covid-19–related hospitalization and death.
Supported by the Biomedical Research Program and the Biostatistics, Epidemiology, and Biomathematics Research Core at Weill Cornell Medicine–Qatar; the Ministry of Public Health; Hamad Medical Corporation; and Sidra Medicine. The Qatar Genome Program and the Qatar University Biomedical Research Center provided the reagents needed for the viral genome sequencing.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
This article was published on June 15, 2022, at NEJM.org.
We thank the many dedicated persons at Hamad Medical Corporation, the Ministry of Public Health, the Primary Health Care Corporation, Qatar Biobank, Sidra Medicine, and Weill Cornell Medicine–Qatar for their diligent efforts and contributions to make this study possible.
Author Affiliations
From the Infectious Disease Epidemiology Group and the World Health Organization Collaborating Center for Disease Epidemiology Analytics on HIV/AIDS, Sexually Transmitted Infections, and Viral Hepatitis, Weill Cornell Medicine–Qatar, Cornell University, Education City (H.N.A., H.C., L.J.A.-R.), the Mathematics Program, Department of Mathematics, Statistics, and Physics, College of Arts and Sciences (H.H.A.), the Biomedical Research Center (H.M.Y., H.A.A.-K., M.K.S., P.C., G.K.N.), and the Departments of Biomedical Science (H.M.Y., H.A.A.-K., M.K.S., G.K.N.) and Public Health (H.F.A.-R., L.J.A.-R.), College of Health Sciences, QU Health, Qatar University, the Department of Pathology, Sidra Medicine (P.T., M.R.H.), Hamad Medical Corporation (P.C., Z.A.-K., E.A.-K., A.J., A.H.K., A.N.L., R.M.S., A.A.B., A.A.-K.), Primary Health Care Corporation (M.G.A.-K.), and the Ministry of Public Health (H.E.A.-R., M.H.A.-T., R.B.) — all in Doha, Qatar; the Departments of Population Health Sciences (H.N.A., H.C., A.A.B., L.J.A.-R.) and Medicine (A.A.B.), Weill Cornell Medicine, Cornell University, New York; and the Wellcome–Wolfson Institute for Experimental Medicine, Queens University, Belfast, United Kingdom (P.C.).
Dr. Chemaitelly can be contacted at hsc2001@qatar-med.cornell.edu or at Weill Cornell Medicine–Qatar, P.O. Box 24144, Qatar Foundation, Education City, Doha, Qatar. Dr. Abu-Raddad can be contacted at lja2002@qatar-med.cornell.edu or at Weill Cornell Medicine–Qatar, P.O. Box 24144, Qatar Foundation, Education City, Doha, Qatar.“
The New England Journal of Medicine (NEJM)
https://www.nejm.org/doi/full/10.1056/NEJMoa2203965
— Or —
„Protection and Waning of Natural and Hybrid Immunity to SARS-CoV-2
List of authors.
Yair Goldberg, Ph.D., Micha Mandel, Ph.D., Yinon M. Bar-On, M.Sc., Omri Bodenheimer, M.Sc., Laurence S. Freedman, Ph.D., Nachman Ash, M.D., Sharon Alroy-Preis, M.D., Amit Huppert, Ph.D., and Ron Milo, Ph.D.
Article
Figures/Media
Metrics
25 References
Abstract
BACKGROUND
Infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) provides natural immunity against reinfection. Recent studies have shown waning of the immunity provided by the BNT162b2 vaccine. The time course of natural and hybrid immunity is unknown.
METHODS
Using the Israeli Ministry of Health database, we extracted data for August and September 2021, when the B.1.617.2 (delta) variant was predominant, on all persons who had been previously infected with SARS-CoV-2 or who had received coronavirus 2019 vaccine. We used Poisson regression with adjustment for confounding factors to compare the rates of infection as a function of time since the last immunity-conferring event.
RESULTS
The number of cases of SARS-CoV-2 infection per 100,000 person-days at risk (adjusted rate) increased with the time that had elapsed since vaccination with BNT162b2 or since previous infection. Among unvaccinated persons who had recovered from infection, this rate increased from 10.5 among those who had been infected 4 to less than 6 months previously to 30.2 among those who had been infected 1 year or more previously. Among persons who had received a single dose of vaccine after previous infection, the adjusted rate was low (3.7) among those who had been vaccinated less than 2 months previously but increased to 11.6 among those who had been vaccinated at least 6 months previously. Among previously uninfected persons who had received two doses of vaccine, the adjusted rate increased from 21.1 among those who had been vaccinated less than 2 months previously to 88.9 among those who had been vaccinated at least 6 months previously.
CONCLUSIONS
Among persons who had been previously infected with SARS-CoV-2 (regardless of whether they had received any dose of vaccine or whether they had received one dose before or after infection), protection against reinfection decreased as the time increased since the last immunity-conferring event; however, this protection was higher than that conferred after the same time had elapsed since receipt of a second dose of vaccine among previously uninfected persons. A single dose of vaccine after infection reinforced protection against reinfection.
Although a decline in protection against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection after two doses of BNT162b2 vaccine (Pfizer–BioNTech) has been observed in several studies,1-3 the level of protection remains unclear, as does the presence or extent of waning of natural immunity. Several studies have shown that 6 or more months after infection, persons still have substantial natural immunity against SARS-CoV-2.4-8 However, one recent study showed that messenger RNA (mRNA)–based vaccines confer a level of protection against hospitalization that is five times as high as that provided by previous infection.9
Waning of the humoral response of the immune system is well documented in vaccinated persons and in those who have been infected with SARS-CoV-2.10,11 In addition, studies of seasonal coronaviruses have shown waning of natural immunity and the possibility of reinfection.12,13 It is also unclear how natural immunity interacts with immunity conferred by vaccination. Some laboratory studies have indicated that “hybrid immunity” (i.e., immunity conferred by the combination of previous infection and vaccination) offers greater broad-spectrum protection,14 elicits higher levels of neutralizing antibodies,15 and provides greater protection against infection16 than immunity conferred by vaccination or infection alone. The durability of immunity resulting from SARS-CoV-2 infection and how this immunity compares with that conferred by vaccination are essential questions both at the level of an individual person and at the national level.
In this study, we estimated the incidence of confirmed SARS-CoV-2 infection in the following cohorts: previously infected, unvaccinated persons; previously infected persons who had also received the BNT162b2 vaccine; and vaccinated persons who had not been previously infected. For each cohort, we quantified the association between the time that had passed since infection or vaccination and the rate of confirmed infection. By comparing the rates of infection among these groups, we were able to assess the level of protection afforded by hybrid immunity as compared with that afforded by natural immunity or immunity conferred by vaccination.
Methods
STUDY POPULATION
Our analysis, which was based on data from the national database of the Israeli Ministry of Health, focused on infections that were confirmed during the study period, from August 1 to September 30, 2021. During this period, Israel was in the midst of a fourth pandemic wave that was dominated by the B.1.617.2 (delta) variant.17 Israel had already conducted a campaign offering two doses of the BNT162b2 vaccine and had initiated a campaign offering third and fourth booster doses (see the Supplementary Methods 1 section in the Supplementary Appendix, available with the full text of this article at NEJM.org). In addition, beginning in March 2021, unvaccinated persons who had recovered from coronavirus disease 2019 (Covid-19) at least 3 months previously were eligible to receive a single dose of BNT162b2 vaccine.
In this study, reinfection with SARS-CoV-2 was defined as a positive polymerase-chain-reaction (PCR) test in a person who had had a positive test of a sample obtained at least 90 days before the study day.18 The definition of severe Covid-19 was consistent with that of the National Institutes of Health19 — that is, a resting respiratory rate of more than 30 breaths per minute, an oxygen saturation of less than 94% while the person was breathing ambient air, or a ratio of partial pressure of arterial oxygen to fraction of inspired oxygen of less than 300. The Israeli Ministry of Health database includes, for all residents who have received a Covid-19 vaccine, been tested for Covid-19, or been previously infected with SARS-CoV-2, basic demographic information such as sex, age, place of residence, and population sector, as well as full records of vaccinations and confirmed infections.
Figure 1.
Study Population.
Using these data at the individual resident level, we studied confirmed infections among persons 16 years of age or older who had tested positive for SARS-CoV-2 infection before July 1, 2021, or who had received at least two doses of BNT162b2 vaccine at least 7 days before the end of the study period. We excluded from the analysis the following persons: those whose data did not include information on age or sex; those who had tested positive for SARS-CoV-2 between July 1 and July 31, 2021; those who had recovered from a PCR-confirmed SARS-CoV-2 infection and then received more than one dose of BNT162b2 vaccine (a small group with limited follow-up data); those who had received more than one dose of BNT162b2 vaccine and then recovered from a PCR-confirmed SARS-CoV-2 infection (a small group); those who had spent the entire study period abroad; and those who had received a vaccine other than BNT162b2 before August 1, 2021 (Figure 1).
STUDY DESIGN AND OVERSIGHT
We compared the incidences of confirmed infection over the study period among cohorts of persons with various histories of immunity-conferring events (i.e., infection or vaccination). The recovered, unvaccinated cohort involved persons who had had a confirmed infection 90 or more days before the study day. There were two “hybrid” cohorts (i.e., cohorts with participants who had both natural immunity and immunity from vaccination); the recovered, one-dose cohort consisted of persons who had recovered from Covid-19 and had later received a single dose of vaccine at least 7 days before the study day, and the one-dose, recovered cohort involved those who had received a single dose of vaccine, followed by a confirmed infection at least 90 days before the study day. The two-dose cohort was composed of persons who had not been infected before the beginning of the study and who had received the second dose of vaccine at least 7 days before the study day, and the three-dose cohort was composed of those who had not been infected before the start of the study and who had received the third (booster) dose of vaccine at least 12 days before the study day.
These cohorts were divided into subcohorts according to the time that had elapsed since the last immunity-conferring event. We used 2 months as the basic time interval to define the subcohorts, but we combined months 12 to 18 for the recovered, unvaccinated cohort and omitted the period of 8 to less than 10 months for the vaccinated and hybrid cohorts because of the small number of persons in those cohorts.
A person could contribute follow-up days to different subcohorts and could also move from one cohort to another according to the following rules. A person who had recovered from Covid-19 and who received a first dose of BNT162b2 vaccine during the study period exited the recovered, unvaccinated cohort on the day of vaccination and entered the recovered, one-dose cohort 7 days later. A person who had recovered from Covid-19 and who had received a first vaccine dose but then received a second dose during the study period exited the recovered, one-dose cohort at the time of the second vaccination. A person in the two-dose cohort who received a third (booster) dose during the study period exited the two-dose cohort on the day of the booster dose and entered the three-dose cohort 12 days later.20 A person with a positive test for SARS-CoV-2 infection between May 1 and June 30, 2021, and who also received a single dose of BNT162b2 vaccine entered either the recovered, one-dose cohort or the one-dose, recovered cohort (according to whether or not confirmed infection predated vaccination) 90 days after the positive test. A person who received a vaccine other than BNT162b2 exited the study on the day of that vaccination.
Studies often compare infection rates among recovered or vaccinated persons with those among unvaccinated persons who have not been previously infected. However, owing to the high vaccination rate in Israel, the latter cohort is small and not representative of the overall population. Furthermore, the Israeli Ministry of Health database does not include complete information about such persons. Therefore, we did not include unvaccinated, previously uninfected persons in our analysis.
The study was approved by the institutional review board at the Sheba Medical Center. The Israeli Ministry of Health and Pfizer have a data-sharing agreement, but only the final results of this study were shared.
STATISTICAL ANALYSIS
To analyze the data, we used methods similar to those used in our previous studies.8,20,21 We assumed that the hazard of infection in each cohort would be independent of the sojourn time in previous cohorts (i.e., the time spent in the cohort before a confirmed infection), and we focused on the relationship between the proportional-hazards survival model and the Poisson regression model22 (see the Supplementary Methods 2 section). Specifically, the number of confirmed infections and the number of person-days at risk during the study period were counted for each subcohort.
A Poisson regression model was fitted, with adjustment for age group as of January 1, 2021 (16 to 39 years, 40 to 59 years, or ≥60 years), sex, population sector (general Jewish, Arab, or ultra-Orthodox Jewish), calendar week, and an exposure risk measure. The latter was calculated for each person on each follow-up day according to the rate of new confirmed infections during the previous 7 days in the person’s area of residence; this continuous measure was then categorized into 10 risk groups according to deciles.20 An average exposure risk was imputed to persons with missing data on residency. In order to ensure that the effect of missing data was small, a descriptive comparison of persons who had missing data with those who did not have missing data, as well as a multiple-imputation analysis, were performed (see the Supplementary Analysis 1 section). Goodness of fit of the model was checked by examining Pearson residuals across the categories.
In a supplementary analysis, we fitted a model with an interaction between age group and subcohort in order to estimate age-specific incidence rates in each subcohort. Each case of infection contributed an event to the respective subcohort. On the basis of the estimated parameters of the fitted regression model, the incidence rate in each subcohort, adjusted for the confounders, was estimated as the expected number of events per 100,000 days if all the person-days at risk were included in that subcohort (see the Supplementary Methods 3 section). The 95% confidence intervals were calculated with the use of a bootstraplike simulation approach23 without adjustment for multiplicity. We repeated the analysis of subcohorts with 1-month intervals (instead of 2-month intervals) to better distinguish between persons who chose to be vaccinated earlier and those who chose to be vaccinated later (or between those who were infected earlier and those who were infected later).
To examine the effect of misclassification of persons into cohorts owing to undocumented infections, we conducted a sensitivity analysis with the assumption that either 50% or 70% of true infections were undocumented. There were too few cases for an in-depth comparison of the incidences of severe disease within and between the cohorts with natural immunity and those with hybrid immunity; thus, only a descriptive analysis was performed. The results of a comparison of the incidences of severe Covid-19 between persons who had received two doses of BNT162b2 vaccine and those who had received a third (booster) dose are reported elsewhere.21
Results
STUDY POPULATION AND DESCRIPTIVE STATISTICS
Figure 2.
Dynamic Inclusion of Persons in the Study Cohorts.
Table 1.
Demographic and Clinical Characteristics of the Study Cohorts.
Our analysis was based on more than 5.7 million persons who contributed days in the five main cohorts (Figure 1). Figure 2 shows the dynamic inclusion of persons in the cohorts over time. Table 1 shows the number of events (confirmed SARS-CoV-2 infections and cases of severe Covid-19) according to the cohorts and demographic characteristics of the persons as well as the distribution of person-days at risk according to sex, age group, and population sector in the five cohorts. The sex distribution was similar in the five cohorts, with only slightly more person-days at risk for women than for men. There were clear differences among the cohorts in the distribution of the other covariates. Although persons who were 60 years of age or older contributed 53.4% of the person-days at risk in the three-dose cohort, persons of this age contributed only 8.3% of the person-days at risk in the recovered, unvaccinated cohort, 13.8% of the person-days at risk in the recovered, one-dose cohort, 14.3% of the person-days at risk in the one-dose, recovered cohort, and 12.6% of the person-days at risk in the two-dose cohort. The distributions of person-days at risk according to population sector also differed among the cohorts because the Arab and ultra-Orthodox Jewish groups have had a higher incidence of infection during the Covid-19 pandemic, resulting in higher percentages of these groups in the cohorts of recovered persons than in the cohorts of persons who were not previously infected. Figure S4 in the Supplementary Appendix shows the distribution of time between infection and vaccination in the hybrid cohorts.
Tables S1 through S4 provide a more detailed tabulation of the data, with each cohort divided into subcohorts according to the time that had elapsed since infection or vaccination. As expected, the differences in the distributions of covariates among the subcohorts within each cohort were smaller than those among the cohorts. The most prominent differences among subcohorts related to the tendency of older persons to receive vaccination earlier, according to the Israeli vaccination prioritization schedule. The numbers of person-days at risk in the subcohorts of persons who had recovered from Covid-19, regardless of whether they were vaccinated, were much smaller than those in the two-dose and three-dose subcohorts. The numbers of cases of severe Covid-19 among persons in each of the subcohorts of the recovered, unvaccinated cohort and in each of the subcohorts of the two hybrid cohorts were small (<10), so reliable quantification of the levels of protection against severe disease in each of these three cohorts was precluded. We therefore focused on comparing the incidences of confirmed infection among the subcohorts. WANING IMMUNITY AGAINST REINFECTION Table 2. Results of the Poisson Regression Analysis of Confirmed SARS-CoV-2 Infections. Figure 3. Estimated Covariate-Adjusted Rates of Confirmed Infections per 100,000 Person-Days at Risk. Table 2 and Figure 3 summarize the results of the Poisson regression analysis and show the estimated numbers of confirmed infections per 100,000 person-days at risk in each subcohort, with adjustment for age, sex, population sector, calendar week, and risk of exposure. Table 2 also provides two sets of rate ratios for each subcohort — one rate ratio that is relative to the reference subcohort of previously uninfected persons who had been vaccinated within the previous 2 months, and one rate ratio that is relative to the subcohort with the most recent immunity-conferring event within the cohort. The complete set of parameter estimates of the regression model is provided in Table S7. The adjusted incidence rates within age groups (16 to 39 years, 40 to 59 years, and ≥60 years) are provided in Table S8 and Figure S1. Figure S2 shows plots of the Pearson residuals indicating an overall satisfactory fit of the model to the data, with somewhat poorer fit in the cohorts vaccinated with two and three doses. Figure S3 shows the rates when the subcohorts were defined according to 1-month periods. We found evidence of waning immunity in all cohorts (Figure 3), with a steady decrease in protection over time. The adjusted rate of confirmed infections among recovered, unvaccinated persons 4 to less than 6 months after infection was 10.5 per 100,000 person-days at risk (95% confidence interval [CI], 8.8 to 12.4); this rate increased to 30.2 (95% CI, 28.5 to 32.0) among persons in this cohort 12 months or more after infection. In the two-dose cohort, the rate was 21.1 (95% CI, 20.0 to 22.4) among persons vaccinated within the previous 2 months, and this rate increased to 88.9 (95% CI, 88.2 to 89.5) among those vaccinated 6 to less than 8 months previously; in the recovered, one-dose cohort with the same times since vaccination, the rates were 3.7 (95% CI, 3.1 to 4.5) and 11.6 (95% CI, 10.0 to 13.5), respectively. In the subcohorts of the recovered, unvaccinated cohort, the adjusted rates of confirmed infection were similar to those of the recovered, one-dose and one-dose, recovered subcohorts when the time elapsed since the last immunity-conferring event (either infection or vaccination) was the same (Figure 3). For example, at 4 to less than 6 months since the last immunity-conferring event, the rates per 100,000 person-days at risk were 10.5 (95% CI, 8.8 to 12.4) in the recovered, unvaccinated cohort, 10.3 (95% CI, 9.4 to 11.4) in the recovered, one-dose cohort, and 10.6 (95% CI, 7.6 to 15.0) in the one-dose, recovered cohort. At 6 to less than 8 months, the rates were 14.0 (95% CI, 13.3 to 14.8), 11.6 (95% CI, 10.0 to 13.5), and 16.2 (95% CI, 14.0 to 18.5), respectively. These rates were lower than those in the two-dose cohort 4 to less than 6 months after vaccination (69.4; 95% CI, 68.7 to 69.9) and 6 to less than 8 months after vaccination (88.9; 95% CI, 88.2 to 89.5). However, the protection conferred by two doses of vaccine was restored with the administration of a third dose; our study showed a rate of 8.2 (95% CI, 8.0 to 8.4) less than 2 months after booster vaccination (Table 2). The sensitivity analysis for misclassification owing to unreported infections revealed that the rates of confirmed infection in the two-dose and three-dose cohorts as described above may have be underestimated by approximately 10% when the misclassification rate was 50% and by approximately 20% when the misclassification rate was 70%. However, such misclassification did not have a substantial effect on the estimates of waning protection (see the Supplementary Analysis 2 section). ANALYSIS OF CASES OF SEVERE COVID-19 The number of cases of severe Covid-19 was small in the cohorts of previously infected persons, with 25 in the recovered, unvaccinated cohort, 13 in the recovered, one-dose cohort, and 1 in the one-dose, recovered cohort. In the two-dose cohort, there were 1372 cases of severe Covid-19, and in the three-dose cohort, there were 178 cases (Table 1). The resulting crude rates of severe disease among persons 60 years of age or older, without consideration of the time since the last immunity-conferring event, were 0.6 per 100,000 person-days at risk in the recovered, unvaccinated cohort, 0.5 in the recovered, one-dose cohort, 0.5 in the one-dose, recovered cohort, 4.6 in the two-dose cohort, and 0.4 in the three-dose cohort. Discussion We evaluated the waning level of protection against confirmed infection with SARS-CoV-2 among persons who had recovered from previous infection and among previously uninfected persons who received the BNT162b2 vaccine. We compared protection in these groups with that in persons who had been vaccinated with a single dose and later infected with SARS-CoV-2 and with that in persons who had recovered from SARS-CoV-2 infection and later received a single vaccine dose. Previous studies showed higher protection in previously infected persons with or without an additional vaccine dose than in previously uninfected persons who had received two doses of mRNA vaccines.6,7 Our study quantifies the waning of natural and hybrid immunity at the national level in a real-world setting. Waning immunity was evident in all the cohorts. This pattern of waning immunity was evident across all age groups. The adjusted rates of confirmed infection among the recovered, unvaccinated subcohorts were lower than those among the two-dose subcohorts when the time since the last immunity-conferring event was similar; nevertheless, the protection in the two-dose cohort could be restored by the administration of a booster shot. In findings that were consistent with those of other studies,6,7,24 after several months, persons with hybrid immunity were better protected against reinfection than uninfected persons who had previously received two doses of vaccine (the two-dose cohort). Furthermore, we found that a single dose of the vaccine administered to a previously infected person or a booster dose administered to an uninfected person who had received two doses of vaccine restored the level of protection to the level in the early months after recovery or vaccination. The timing of vaccination after infection affects the protection.6 We did not have enough data to evaluate the level of protection as a function of time between infection and vaccination, while taking the waning effect into account. The results reported here are in line with those of a study conducted by an Israeli health maintenance organization.7 That study showed that previously infected persons with or without one vaccine dose have better protection than uninfected persons who have received two doses of vaccine 3 to less than 8 months after the last immunity-conferring event. Our data on hospitalized patients who had severe Covid-19 did not contain enough cases for a definitive analysis but did not appear to support the findings in a recent report9 that suggested that vaccinated persons are more protected than previously infected persons 3 to less than 6 months after an immunity-conferring event. In the recovered, unvaccinated cohort and the hybrid cohorts, the first infections were primarily infections with the original Wuhan-Hu-1 isolate and the B.1.1.7 (alpha) variant.17 If protection provided by previous infection depends on the variant, its effect is confounded with the effect of time since infection. Because a single variant was dominant in Israel during each of the pandemic waves,17 this study cannot disentangle the two effects. Moreover, during the study period, most infections were delta variant infections, and our analysis provides no information regarding protection against newer variants such as B.1.1.529 (omicron). This was an observational study in which persons elected to receive a vaccine at different times, and there was no control for the probable differences in health care–seeking or risk-averse behavior of individual persons. Although the regression approach corrects for confounders for which data are available, including data on exposure risk, the possibility of residual bias remains. The residuals analysis revealed an overall reasonable fit, with a few large residuals in the cohorts vaccinated with two or three doses. These cohorts had large sample sizes, leading to substantial sensitivity to even a modest lack of fit. Our results pertained to the rate of confirmed infection, so they were sensitive to detection bias due to different tendencies to perform PCR testing in the study cohorts. During the study period, the same official PCR testing policy applied to both previously infected persons and those who had received two doses of vaccine — namely, mandatory PCR testing on contact with an infected person. Although differences in testing rates among cohorts and among subcohorts within specified cohorts were observed, their overall magnitude was relatively small. The rate of PCR testing was typically lower in the recovered, unvaccinated cohort than in the other cohorts, so the level of protection in this cohort as compared with that in the two-dose cohort may have been overestimated. The data regarding severe disease were not affected by this bias. Another source of potential bias was cohort misclassification. To be classified as a recovered person in our study, a PCR test must have been performed and found to have been positive. However, not all infected persons had received a diagnosis,25 and some of these persons had been vaccinated. Thus, some of the persons who were classified as being in the two-dose cohort or the three-dose cohort should have been considered to have had hybrid immunity. Under simple assumptions about the misclassification mechanism, we found that misclassification may have led to a 10% or even a 20% underestimation of the infection rate among vaccinated, uninfected persons, depending on the misclassification rate. Although the magnitude of the bias depends on our assumptions, the bias toward underestimation of the infection rate among vaccinated, uninfected persons is real if those who had recovered from Covid-19 and had been misclassified as belonging to the vaccinated cohorts were more protected from reinfection than their uninfected counterparts. An understanding of the rates of waning immunity after immunity-conferring events is important for policy making regarding the need for and the timing of additional vaccine doses. We found that protection against the delta variant waned over time in both vaccinated and previously infected persons and that an additional vaccine dose restored protection. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. Drs. Goldberg and Mandel and Drs. Huppert and Milo contributed equally to this article. This article was published on May 25, 2022, at NEJM.org. We thank Ofra Amir for productive feedback on an earlier version of the manuscript. Author Affiliations From the Faculty of Industrial Engineering and Management, Technion–Israel Institute of Technology, Haifa (Y.G.), the Department of Statistics and Data Science, Hebrew University of Jerusalem (M.M.), and the Israeli Ministry of Health (O.B., N.A., S.A.-P.), Jerusalem, the Department of Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot (Y.M.B.-O., R.M.), the Biostatistics and Biomathematics Unit, Gertner Institute for Epidemiology and Health Policy Research, Sheba Medical Center, Ramat Gan (L.S.F., A.H.), and the Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv (A.H.) — all in Israel. Dr. Goldberg can be contacted at yairgo@technion.ac.il or at the Faculty of Industrial Engineering and Management, Technion–Israel Institute of Technology, Haifa 3200003, Israel.“
https://www.nejm.org/doi/full/10.1056/NEJMoa2118946