A Systematic Review and Meta-Analysis on the Real-World Effectiveness of COVID-19 Vaccines against Infection, Symptomatic and Severe COVID-19 Disease Caused by the Omicron Variant (B.1.1.529)

Real-world data on the effectiveness of COVID-19 vaccines against the Omicron variant (B.1.1.529) is limited. This systematic review aimed to investigate the real-world effectiveness and durability of protection conferred by primary course and booster vaccines against confirmed Omicron infection, and severe outcomes. We systematically searched literature up to 1 August 2022. Meta-analysis was performed with the DerSimonian-Laird random-effects model to estimate the pooled vaccine effectiveness (VE). Overall, 28 studies were included representing 11 million individuals. The pooled VE against Omicron infection was 20.4% (95%CI: 12.1–28.7%) and 23.4% (95%CI: 13.5–33.3%) against symptomatic infection with variation based on vaccine type and age groups. VE sharply declined from 28.1% (95%CI: 19.1–37.1%) at three months to 3.9% (95%CI: −24.8–32.7%) at six months. Similar trends were observed for symptomatic Omicron infection. A booster dose restored protection against Omicron infection up to 51.1% (95%CI: 43.8–58.3%) and 57.3% (95%CI: 54.0–60.5%) against symptomatic infection within three months; however, this waned to 32.8% (95%CI: 16.8–48.7%) within six months. VE against severe Omicron infection following the primary course was 63.6% (95%CI: 57.5–69.7%) at three months, decreased to 49% (95%CI: 35.7–63.4%) within six months, and increased to 86% after the first or second booster dose.


Introduction
The Omicron variant (B.1.1.529) was reported to the World Health Organization (WHO) from South Africa in late November 2021 [1]. It was immediately designated as a variant of concern (VOC) [1]. Compared to pre-omicron variants, a number of mutations have been identified in the omicron variant, including multiple mutations in the receptor-binding domain of the spike protein associated with increased transmissibility and immune evasion after natural infection and vaccination [2][3][4]. The omicron variant has rapidly evolved into new sub-lineages or sub-variants: BA. 1  and XBB (a hybrid of two different Omicron BA.2 sub-variants) are Omicron sub-lineages being monitored by the WHO to investigate if these lineages may pose an additional threat to global public health [1].
Several studies have shown diminished neutralization of both Omicron variants by pre-Omicron convalescent sera and by sera of vaccinated individuals [4][5][6]. Recent studies have shown a reduction in COVID-19 vaccine effectiveness against the Omicron variant [7][8][9][10][11], affecting the current COVID-19 vaccination strategy. Recent social media analysis has shown increased public vaccine hesitancy due to the potential lack of effectiveness of ancestral COVID-19 vaccines against the new VOCs [12].
Emerging data on high prevalence of asymptomatic infection, greater risk of reinfection, and reduced vaccine protection during the omicron-dominant period compared to the earlier VOC warrants further investigation on the effectiveness of current COVID-19 vaccines against the Omicron variant [7][8][9][10][11]13]. A systematic review and meta-analysis have recently been published to evaluate the effectiveness of the current COVID-19 vaccines against Omicron infection [14]. This meta-analysis included 15 studies and demonstrated that primary vaccination does not provide sufficient protection against symptomatic Omicron infection [14]. However, the systematic review included studies conducted in the early Omicron era with shorter-term follow up. The real-world long-term effectiveness and durability of protection conferred by primary COVID-19 vaccination course and booster doses against the Omicron variant is not precisely known. To summarize the existing evidence on the effectiveness and the duration of protection conferred by COVID-19 vaccines, data were synthesized from an ongoing systematic review [15]. This systematic review aimed to investigate the real-word effectiveness of primary and booster vaccination against SARS-CoV-2 infection and severe COVID-19 disease due to laboratory-confirmed SARS-CoV-2 Omicron variant. The review also aimed to evaluate the duration of protection following full vaccination and booster doses.

Methods
The systematic review drew data from an ongoing systematic review that aimed to synthesis and evaluate the vaccine effectiveness (VE) of COVID-19 vaccines at preventing SARS-CoV-2 infections and severe COVID-19 disease in real-world settings. The systematic review followed the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guideline and was registered in the Prospective Register of Systematic Reviews (PROSPERO registration ID: CRD42022291375) [15].

Search Strategy
Systematic literature searches were performed on MEDLINE, PubMed, Embase and Cochrane Database of Systematic Reviews website on or before 1 August 2022 with no language restrictions. WHO COVID-19 DATABASE, pre-print severs (medRxiv, bioRxiv) and grey literature were searched. For preprint studies with several versions available, the most recent update published up to 1 August 2022 was included. Reviews and their references are examined for inclusion. Medical subject headings with the following search keywords were used: SARS-CoV-2 OR COVID-19 AND Vaccine OR Vaccination AND OR Vaccine effectiveness. The methods have been previously described in detail in the ongoing systematic review protocol [15] and full search strategies are available in Supplementary Material S1.

Study Selection
Observational (non-randomized) studies including cohort studies (prospective or retrospective), cross-sectional studies, case control including test-negative design (TND), regression discontinuity design studies, post-licensure observational studies that investigated the effectiveness of COVID-19 vaccines against documented, symptomatic, severe COVID-19 disease (defined as hospitalization, ICU admission, intubation or mechanical ventilation, or death) were included. COVID-19 cases were defined as being due to the Omicron variant infection, based on S target-negative results on PCR or whole-genome sequencing. VE studies that evaluated effectiveness of the primary vaccine course and booster vaccination compared to no vaccination were included. Outcomes of interest were VE against Omicron infection of "any type" (i.e., studies did not indicate underlying symptoms), "symptomatic COVID-19", and "severe COVID-19" due to Omicron infection. We only included studies that evaluated VE ≥14 days after the primary vaccination course, and ≥7 days after the booster dose. No restrictions were applied to the age of participants, the types of vaccination, or the number of participants. Heterologous primary schedules were considered. Studies that did not report VE data or did not use any confounder adjustment strategies were excluded.
All the relevant records were screened by title and abstract. The retrieval results were screened with the help of Endnote and duplicate studies were eliminated. Potentially relevant publications underwent full-text examination and disagreements on eligibility were solved through discussion. The full texts suitable for the quantitative synthesis were collected in an excel spreadsheet for data extraction.

Data Extraction
Data were extracted by three (H.M., D.D.P.-T. and Z.Y.M.Y.) independent reviewers to identify eligible studies that met pre-specified inclusion criteria. The following information: study design, year of publication, country, age, population type, type of vaccines, time period post primary series or booster doses and study follow-up period, were extracted from the eligible studies. VE data were stratified according to SARS-CoV-2 vaccination course, ≥14 days after completion of the primary vaccination course and ≥7 days after receiving the first or second booster doses. Within each subgroup, the vaccination course was classified according to the vaccine type or technology. Duration of effectiveness of SARS-CoV-2 vaccines was assessed in intervals of three, six, and longer than six months after the primary vaccination series, whereas for the booster vaccination, shorter time intervals were considered (seven or more days, within three months, three to six months) due to less follow-up time since introduction.

Quality Assessment
The Joanna Briggs Institute (JBI) tools [16] were used to assess risk of bias of the included studies (Supplementary Material S2).

Data Analysis
Descriptive statistics were used to summarize the characteristics of studies included in this review. VE was quantified as the risk reduction of any or severe Omicron infection, expressed as a percentage, compared to the unvaccinated group. VE estimates were derived from regression models (Logit, Poisson, and Cox regression models) and calculated as (1 − IRR) × 100, where IRR = incidence rate ratio; (1 − HR) × 100, where HR = Hazard ratio; (1 − RR) × 100, where RR = Risk ratio; and (1 − OR) × 100, where OR = Odds ratio is the ratio of the rate of COVID-19 in the vaccinated group to the corresponding rate in the unvaccinated group. The DerSimonian-Laird random-effects model with Hartung-Knapp-Sidik-Jonkman variance correction was used to combine VE estimates. We used the I 2 test to quantify the heterogeneity between studies. I 2 values were defined as low (≤50%), moderate (50-75%), or high heterogeneity (>75%). To estimate the duration of protection following the primary vaccination series, we modeled days since completing the primary course as a continuous effect, allowing for nonlinearity by using restricted cubic splines. The analysis was carried out using Stata 17.

VE of a Second Booster or Fourth Dose
One study evaluated the VE of a fourth dose (second booster dose) of mRNA-1273 in older residents of long term care facilities in Ontario, Canada [24]. The pooled VE estimate of a fourth dose of mRNA-1273 followed by any combination of three mRNA vaccines at "≈7 days" against Omicron infection was 50.3% (95%CI: 47.1-53.6%), 69.7% (95%CI: 65.3-74.2%) against symptomatic Omicron infection, and 86% (95%CI: 81-90%) against severe outcomes (Figures S3-S5). There was no VE data for other time periods.

Subgroup Analyses
Subgroup VE analyses were performed by age, and sub-lineage of Omicron. For some subgroup analyses, there were not enough data points to estimate VE or duration of protection at different time points.
Additionally, we performed sensitivity analyses by study designs (cohort or casecontrol) and statistical methods employed to estimate the VE (Logit, Poisson, and Cox regression models). These subgroup analyses did not reveal any meaningful differences to the overall VE findings against all outcomes (noting small numbers in some subgroups) (Data not shown).

Discussion
Our meta-analysis of 28 studies, which included nearly 11 million individuals, provides evidence on VE and duration of protection of COVID-19 vaccines against SARS-CoV-2 Omicron infection and severe COVID-19. Our data suggest that primary vaccination series are not sufficiently protective against the Omicron infection and protection wanes substantially over time from 28% at three months to 4% at six months. Similar trends were observed for symptomatic Omicron infection following full vaccination, broadly consistent with recent review findings [14,40,41]. The waning of primary COVID-19 vaccination course was less pronounced against severe Omicron disease, decreasing from 64% at three months to 49% after six months, consistent with recent findings [14,40,41].
Our meta-analysis suggests that first booster dose restores and provides additional protection for all outcomes. VE of the first booster dose against any (51%) or symptomatic (57%) Omicron infection remained moderate for at least 3 months. Although there was limited data for longer follow-up, VE of the first booster dose against symptomatic Omicron infection waned to 33% at six months, falling below the WHO's minimal criteria of 50% when considering the outcomes of infection and symptomatic disease. This suggests the waning effect is also present for booster vaccination, consistent with recent studies conducted during the Omicron-dominant period [42][43][44]. However, our review suggests that protection against severe Omicron cases remained robust up to 86% after a single dose of booster for at least up to six months, corroborating recent findings [14,40,41].
Vaccine waning following booster vaccination was lower among the younger age group compared to the older adults aged ≥60 years. VE of the second booster dose against symptomatic Omicron infection declined more rapidly in older adults aged ≥60 years from 58% at three months to 14% at six months [24]. However, high level protection against severe disease still remained up to 78% for at least six months after the second booster dose in older adults, who are more vulnerable to severe COVID-19 outcomes. As of 28 November 2022, 68.5% of the world population have not received booster doses [45]. Future research should continue to evaluate the VE of booster vaccination with longer follow-ups to determine the duration of protection against the Omicron variant.
This meta-analysis had several limitations. The included studies were highly heterogeneous in terms of study populations, statistical approaches employed, definitions of symptomatic or severe COVID-19 used, analysed time points after vaccination, and vaccination schedules and regimes. All these factors may contribute to the discrepancy in our VE estimates and limit the generalizability of our results. Although the included studies made some sort of adjustments to their final VE estimates, not all accounted for important confounders, such as previous SARS-CoV-2 infection, underlying comorbidities, socio-economic parameters and COVID mitigation strategies.

Conclusions
This meta-analysis from a wide variety of study types, settings and populations demonstrates that primary COVID-19 vaccination courses were limited in preventing infections and severe disease caused by the SARS CoV-2 Omicron variant. Our review highlights the importance of booster doses for protection against Omicron infection and, more importantly, in providing high levels of protection against severe Omicron disease, particularly among the elderly population. Further research on the real-world performance of the existing vaccines including the new Omicron-specific vaccines is needed.  Figure S1: Scatterplot of VE against SARS-CoV-2 infection of the Omicron variant plotted according to time from the primary vaccination course. Figure S2: Scatterplot of VE against severe COVID-19 due to Omicron infection plotted according to time from the primary vaccination course. Figure S3: VE estimates against SARS-CoV-2 infection of the Omicron variant after two booster dose in older adults aged ≥60 years. Figure S4: VE estimates against symptomatic Omicron infection after two booster dose in older adults aged ≥60 years. Figure S5: VE estimates against severe COVID-19 due to Omicron infection after two booster dose in older adults aged ≥60 years. Figure S6: VE estimates against SARS-CoV-2 infection of the Omicron variant after the primary vaccination course, by age group. Figure S7: VE estimates against symptomatic Omicron infection after the primary vaccination course, by age group. Figure S8: VE estimates against symptomatic Omicron infection after one booster dose, by age group. Figure S9: VE estimates against severe COVID-19 due to Omicron infection after the primary vaccination course, by age group. Figure S10: VE estimates against SARS-CoV-2 infection of the Omicron variant after the primary vaccination course, by age group. Figure S11: VE estimates against SARS-CoV-2 infection of the Omicron variant after the primary vaccination course in older adults aged ≥60 years, by time intervals. Figure S12: VE estimates against severe COVID-19 due to Omicron infection after the primary vaccination course, by age group. Figure S13: VE estimates severe COVID-19 due to Omicron infection after one booster dose in older adults aged ≥60 years, by time intervals. Figure S14: VE estimates against symptomatic Omicron infection after the primary vaccination course, by Omicron sub-variants. Figure S15: VE estimates against symptomatic Omicron infection after one booster dose, by Omicron sub-variants. Figure S16: VE estimates against severe Omicron infection after the primary vaccination course, by Omicron sub-variants. Figure   Data Availability Statement: All data generated from the study have been included in the manuscript. Data were obtained from the primary studies included in the review.