Duration of protection of BNT162b2 and mRNA-1273 COVID-19 vaccines against symptomatic SARS-CoV-2 Omicron infection in Qatar

Study population, data sources, and study design

This study was conducted in the resident population of Qatar, applying the same methodology that was used to assess waning of the BNT162b2⁴ and mRNA-1273⁵ COVID-19 vaccine effectiveness in the same population during earlier infection waves. A detailed description of this methodology can be found in Chemaitelly et al.⁴

COVID-19 laboratory testing, vaccination, clinical infection data, and related demographic details were extracted from the national, federated SARS-CoV-2 databases that include all polymerase chain reaction (PCR) testing, COVID-19 vaccinations, and COVID-19 hospitalizations and deaths in Qatar since the start of the pandemic.

Every PCR test conducted in Qatar is classified on the basis of symptoms and the reason for testing (clinical symptoms, contact tracing, surveys or random testing campaigns, individual requests, routine healthcare testing, pre-travel, at port of entry, or other). Qatar has unusually young, diverse demographics, in that only 9% of its residents are ≥50 years of age, and 89% are expatriates from over 150 countries.^6,7 Nearly all individuals were vaccinated in Qatar, but if vaccinated elsewhere, those vaccinations were still recorded in the health system at the port of entry upon return to Qatar.

By February 2, 2022 (end of study), 1,301,203 individuals received ≥2 BNT162b2 doses in Qatar, of whom 301,059 received a booster dose. Meanwhile, 891,810 individuals received ≥2 mRNA-1273 doses, of whom 105,558 received a booster dose. The median dates of first, second, and third doses were May 3, 2021, May 24, 2021, and December 17, 2021 for BNT162b2, and May 28, 2021, June 27, 2021, and January 4, 2022 for mRNA-1273, respectively. The median time between the first and second dose was 21 days (interquartile range (IQR), 21-22 days) for BNT162b2 and 28 days (IQR, 28-30 days) for mRNA-1273, while that between second and booster dose was 250 days (IQR, 232-272 days) for BNT162b2 and 233 days (IQR, 212-254 days) for mRNA-1273.

Vaccine effectiveness during the large SARS-CoV-2 Omicron wave in Qatar, between December 23, 2021 and February 2, 2022, was estimated using the test-negative, case-control study design, a standard design for assessing vaccine effectiveness.^8-16 Cases (PCR-positive persons) and controls (PCR-negative persons) were exact-matched two-to-one by sex, 10-year age group, nationality, and calendar week of PCR test to estimate effectiveness against symptomatic SARS-CoV-2 Omicron infection. Symptomatic SARS-CoV-2 Omicron infection was defined as a PCR-positive swab collected during the Omicron wave because of clinical suspicion due to presence of symptoms compatible with a respiratory tract infection.

The two-to-one matching ratio was done to improve the statistical precision of the estimates. This specific matching ratio was dictated by the higher number of PCR-positive tests than PCR-negative tests during the large Omicron wave (Figure S1). Matching was performed to control for known differences in the risk of exposure to SARS-CoV-2 infection in Qatar.^7,17-20

Vaccine effectiveness was also estimated against any severe, critical, or fatal COVID-19 using the same methodology. Here, however, the matching ratio was one-to-five to improve the statistical precision given the relatively small number of severe, critical, and fatal COVID-19 cases during the Omicron wave.

Only the first PCR-positive test during the study was included for each case, whereas all PCR-negative tests during the study were included for each control. Controls included individuals with no record of a PCR-positive test during the study period. Only PCR tests conducted for clinical suspicion; that is due to presence of symptoms compatible with a respiratory tract infection, were included in the analysis. All persons who received a vaccine other than BNT162b2 or mRNA-1273, or who received mixed vaccines, were excluded. These inclusion and exclusion criteria were implemented to minimize different types of potential bias based on earlier analyses in the same population.^4,5 Every control that met the inclusion criteria and that could be matched to a case was included in the analysis. Both PCR-test outcomes and vaccination status were ascertained at the time of the PCR test.

Classification of COVID-19 case severity (acute-care hospitalizations),²¹ criticality (ICU hospitalizations),²¹ and fatality²² followed World Health Organization (WHO) guidelines, and assessments were made by trained medical personnel using individual chart reviews (Section S1). Each person who had a positive PCR test result and COVID-19 hospital admission was subject to an infection severity assessment every three days until discharge or death, regardless of the length of the hospital stay or the time between the PCR-positive test and the final disease outcome. Individuals who progressed to severe,²¹ critical,²¹ or fatal²² COVID-19 between the PCR-positive test result and the end of the study were classified based on their worst outcome, starting with death, followed by critical disease, and then severe disease.

Laboratory methods

Details of laboratory methods for real-time reverse-transcription PCR (RT-qPCR) testing and multiplex RT-qPCR variant screening are found in Section S2. All PCR testing was conducted at the Hamad Medical Corporation Central Laboratory or at Sidra Medicine Laboratory, following standardized protocols.

Oversight

Hamad Medical Corporation and Weill Cornell Medicine-Qatar Institutional Review Boards approved this retrospective study with waiver of informed consent. The study was reported following the STROBE guidelines. The STROBE checklist is found in Table S1.

Statistical analysis

All records of PCR testing in Qatar during the study were examined, but only samples of matched cases and controls that underwent PCR testing for clinical suspicion were included in the analysis. Demographic characteristics of study samples were described using frequency distributions and measures of central tendency. Study groups were compared using standardized mean differences (SMDs), defined as the difference in the mean of a covariate between groups divided by the pooled standard deviation. SMD <0.1 indicated adequate matching.²³

The odds ratio, comparing odds of vaccination among cases versus controls, and its associated 95% confidence interval (CI) were derived using conditional logistic regression, that is factoring the matching in the study design. This matching and analysis approach aims to minimize potential bias due to variation in epidemic phase,^8,24 gradual roll-out of vaccination during the study,^8,24 or other confounders.^25,26 95% CIs were not adjusted for multiplicity and thus should not be used to infer definitive differences between different groups. Interactions were not investigated. Vaccine effectiveness at different time points and its associated 95% CI were calculated by applying the following equation:^8,9 In each analysis for a specific time-since-vaccination stratum, we included only those vaccinated in that specific time-since-vaccination stratum and those unvaccinated (our reference group). Only matched pairs of PCR-positive and PCR-negative persons, in which members of the pair were either unvaccinated or fell within each time-since-vaccination stratum were included in the corresponding vaccine effectiveness estimate. Thus, the number of cases (and controls) varied across time-since-vaccination analyses. Effectiveness after the second dose was estimated by one or more months where one month was defined as 30 days. Effectiveness after the third/booster dose was estimated by one or more weeks where one week was defined as 7 days.

Caveats and limitations

Since the immunization campaigns prioritized vaccination of persons with severe or multiple chronic conditions and by age group, the observed pattern of waning of protection could theoretically be confounded by effects of age and comorbidities. Individual-level data on co-morbid conditions were not available; therefore, they could not be explicitly factored into our analysis. However, only a small proportion of the study population may have had serious co-morbid conditions. Only 9% of the population of Qatar are ≥50 years of age,^6,7 and 60% are young, expatriate craft and manual workers working in mega-development projects.^19,20,27 The national list of persons prioritized to receive the vaccine during the first phase of vaccine roll-out included only 19,800 individuals of all age groups with serious co-morbid conditions. Old age may serve as a partial proxy for co-morbid conditions. A similar pattern of waning of protection was observed for younger and older persons in our earlier waning of vaccine effectiveness studies for the same population.^4,5 However, with the small proportion of Qatar’s population being ≥50 years of age,^6,7 our findings may not be generalizable to other countries in which elderly citizens constitute a larger proportion of the total population.

Vaccinated persons presumably have a higher social contact rate than unvaccinated persons, and they may also adhere less strictly to safety measures.^28-30 This behavior could reduce real-world effectiveness of the vaccine compared to its biological effectiveness, possibly explaining the waning of protection. Public health restrictions in Qatar were eased gradually, but differently for vaccinated and unvaccinated persons. Many social, work, and travel activities presently require evidence of vaccination (a “health pass”) that is administered through a mandatory mobile app (the Ehteraz app). However, risk compensation is unlikely to solely explain the observed waning of protection over time.

PCR testing in Qatar is done on a mass scale, such that about 5% of the population are tested every week.⁴ With the large demand for testing during the Omicron wave, use of rapid antigen testing (RAT) was expanded to supplement PCR testing at health care facilities starting from January 5, 2022, but information on reason for testing was not available for the RAT testing to be included in our study.

Effectiveness was assessed using an observational, test-negative, case-control study design,^8,9 rather than a design in which cohorts of vaccinated and unvaccinated individuals were followed up. However, the cohort study design applied earlier to the same population of Qatar yielded findings similar to those of the test-negative case-control design,^13-16 supporting the validity of this standard approach in assessing vaccine effectiveness for respiratory tract infections.^8-16 Moreover, our recent study of the effectiveness of booster vaccination against SARS-CoV-2 symptomatic infection, relative to that of the primary series, used a cohort study design³ and its results are consistent with the results generated in the present study using the test-negative, case-control study design.

With the high COVID-19 vaccine coverage in Qatar (>80%), the majority of those unvaccinated were children or young persons, and thus the reference group of those unvaccinated is not representative of the total population demographics. However, matching by age was implemented in all analyses to minimize potential bias arising from the young population structure of those unvaccinated.

To rapidly scale up vaccination, some vaccination campaigns are conducted outside healthcare facilities; thus, records of vaccination are not immediately uploaded into the CERNER system, which tracks all vaccination records at the national level. This administrative time delay can introduce a misclassification bias of those vaccinated versus those unvaccinated. A sensitivity analysis investigating the impact of such potential bias, by assuming a 10% misclassification bias of those vaccinated, found a difference of, at most, few percentage points in estimated effectiveness.⁵ A key strength of the test-negative, case-control study design is that it is less susceptible to this form of bias.^8,9

Nonetheless, one cannot exclude the possibility that in real-world data, bias could arise in unexpected ways, or from unknown sources, such as subtle differences in test-seeking behavior or changes in the pattern of testing with introduction of other testing modalities, such as RAT. For example, with the large Omicron wave in Qatar, use of RAT was expanded to supplement PCR testing starting from January 5, 2022. However, RAT was broadly implemented in the population and thus may not have differentially affected PCR testing to introduce bias.

With the small number of PCR-positive symptomatic infections in some of the time-since-vaccination strata for those vaccinated with mRNA-1273, it was not possible to estimate mRNA-1273 effectiveness month-by-month after the second dose. Furthermore, with the relatively small number of cases and controls that progressed to severe, critical, or fatal COVID-19 during the Omicron wave, confidence intervals for vaccine effectiveness against hospitalization and death were either very wide or could not be estimated, especially so for the mRNA-1273 analyses.

Notwithstanding these limitations, consistent findings were reached indicating rapid waning of vaccine protection against symptomatic Omicron infection that are broadly consistent with findings of other studies.^31-39 Moreover, with the mass scale of PCR testing in Qatar,⁴ the likelihood of bias is perhaps minimized. Extensive sensitivity and additional analyses were conducted to investigate effects of potential bias in our earlier studies for the BNT162b2⁴ and mRNA-1273⁵ vaccines, which used the same methodology as the present study. These included different adjustments in the analysis, different approaches for factoring prior infection in the analysis, and different study inclusion and exclusion criteria to investigate whether the effectiveness estimates could have been biased.^4,5 All analyses showed consistent findings.^4,5