key: cord-0260191-xfcqbfse authors: Skowronski, D. M.; Setayeshgar, S.; Febriani, Y.; Ouakki, M.; Zou, M.; Talbot, D.; Prystajecky, N.; Tyson, J. R.; Gilca, R.; Brousseau, N.; Deceuninck, G.; Galanis, E.; Fjell, C. D.; Sbihi, H.; Fortin, E.; Barkati, S.; Sauvageau, C.; Naus, M.; Patrick, D. M.; Henry, B.; Hoang, L. M.; De Wals, P.; Garenc, C.; Carignan, A.; Drolet, M.; Sadarangani, M.; Brisson, M.; Krajden, M.; De Serres, G. title: Two-dose SARS-CoV-2 vaccine effectiveness with mixed schedules and extended dosing intervals: test-negative design studies from British Columbia and Quebec, Canada date: 2021-10-26 journal: nan DOI: 10.1101/2021.10.26.21265397 sha: 7482cd5d25bf7d13698d96ebc883552ed66bf502 doc_id: 260191 cord_uid: xfcqbfse Background The Canadian COVID-19 immunization strategy deferred second doses and allowed mixed schedules. We compared two-dose vaccine effectiveness (VE) by vaccine type (mRNA and/or ChAdOx1), interval between doses, and time since second dose in two of Canada's larger provinces. Methods Two-dose VE against infections and hospitalizations due to SARS-CoV-2, including variants of concern, was assessed between May 30 and October 2, 2021 using test-negative designs separately conducted among community-dwelling adults [≥]18-years-old in British Columbia (BC) and Quebec, Canada. Findings In both provinces, two doses of homologous or heterologous SARS-CoV-2 vaccines were associated with ~95% reduction in the risk of hospitalization. VE exceeded 90% against SARS-CoV-2 infection when at least one dose was an mRNA vaccine, but was lower at ~70% when both doses were ChAdOx1. Estimates were similar by age group (including adults [≥]70-years-old) and for Delta-variant outcomes. VE was significantly higher against both infection and hospitalization with longer 7-8-week vs. manufacturer-specified 3-4-week interval between doses. Two-dose mRNA VE was maintained against hospitalization for the 5-7-month monitoring period and while showing some decline against infection, remained [≥]80%. Interpretation Two doses of mRNA and/or ChAdOx1 vaccines gave excellent protection against hospitalization, with no sign of decline by 5-7 months post-vaccination. A 7-8-week interval between doses improved VE and may be optimal in most circumstances. Findings indicate prolonged two-dose protection and support the use of mixed schedules and longer intervals between doses, with global health, equity and access implications in the context of recent third-dose proposals. In Canada, two mRNA vaccines were approved in December 2020 according to a two-dose schedule with interval between doses of three weeks for BNT162b2 (Pfizer-BioNTech, Comirnaty) and four weeks for mRNA-1273 (Moderna, Spikevax). 1 On February 26, 2021, a chimpanzee adenoviral vectored (ChAdOx1) vaccine (AstraZeneca, Vaxzevria and equivalent COVISHIELD) was authorized with an interval of 4-12 weeks between doses. 1 In randomized controlled trials (RCTs), the efficacy of a single dose exceeded 90% for mRNA and 75% for ChAdOx1 vaccines. [2] [3] [4] [5] In early January 2021, confronted with elevated pandemic activity and constrained vaccine supplies, two provinces, British Columbia (BC) and Quebec, opted to extend the interval between doses (to six weeks and 12 weeks, respectively), to provide the benefits of substantial single-dose protection to as many people as possible, as soon as possible. In early March 2021 (epidemiological week 9), Canada's National Advisory Committee on Immunization (NACI) endorsed a second dose-deferral approach, recommending that the interval between first and second doses of all SARS-CoV-2 vaccines be extended up to 16 weeks. 6 On March 29 (week 13), in response to emerging reports of ChAdOx1-associated thrombosis with thrombocytopenia, NACI recommended ChAdOx1 be used for adults ≥ 55-years-old. 7 On April 23 (week 16), NACI lowered this age limit to ≥ 30 years. 7 After assessing interchangeability, NACI further recommended on June 1 (week 22), that first-dose recipients of ChAdOx1 could be offered either the same product or an mRNA vaccine, and that first-dose mRNA recipients could complete the series with the alternate mRNA vaccine if the same product was not available. 8 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 26, 2021. ; https://doi.org/10.1101/2021.10.26.21265397 doi: medRxiv preprint The first SARS-CoV-2 vaccines in BC and Quebec were prioritized for long-term care facility (LTCF) residents and healthcare workers. Vaccination of community-dwelling adults began with the oldest in March 2021, progressing sequentially to younger age groups. With the recommended deferral, second-dose coverage started to increase among elderly adults in May. As vaccine supply improved, the interval between first and second doses was shortened to eight weeks in late-May, by which time (week 21) about 70% of all adults ≥ 18-years-old in BC and Quebec had received at least one dose and <10% had received two doses. The interval between doses was again shortened in August to four weeks to maximize the number of fully-vaccinated individuals before autumn, with >80% of adults ≥ 18 years in both provinces having received two doses by early-October 2021. We report two-dose vaccine effectiveness (VE) against infection and hospitalization, including due to the Delta variant of concern (VOC), among adults ≥ 18-years-old in BC and Quebec. Adjustments to the COVID-19 vaccination program in response to changing pandemic conditions in BC and Quebec provided further unique opportunity to compare two-dose VE by vaccine type, both homologous and heterologous; by interval between doses; and by time since the second dose. There are ~4 million adults ≥ 18-years-old in BC, the westernmost province of Canada, and ~7 million in Quebec, located in eastern Canada ~5000 km apart. About half the adult population in BC and Quebec are women with similar age distributions 18-49, 50-69 and ≥ 70 years: 51% and 49%, 32% and 33%, and 16% and 17%, respectively. A publicly-funded, mostly symptom-based approach for PCR-based SARS-CoV-2 diagnostic testing is broadly accessible in both provinces. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Each province experienced a third pandemic wave that peaked in mid-April 2021 (weeks [14] [15] then subsided to stable low levels from early-June (weeks 22-23) before gradually increasing again with the start of a fourth pandemic wave in late-July/early-August (weeks 29-31). 9,10 Whereas during the third wave the Alpha variant predominated in Quebec, and co-dominated with Gamma in BC, during the fourth wave the Delta variant became dominant in both provinces. 9,10 Specimens with collection dates between weeks 22-39 (May 30 to October 2) were eligible. Cases included any RT-PCR-confirmed SARS-CoV-2 infection. Hospitalized cases were admitted on or ≤ 30 days after specimen collection. Individuals could contribute a single test-positive specimen and were censored from any contribution thereafter. Controls included all specimens that were RT-PCR-negative for SARS-CoV-2 and met inclusion/exclusion criteria. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. In variant-specific analyses, cases were categorized as Alpha, Gamma, or Delta VOC. Methods and sampling frame for genetic characterization evolved in response to changing epidemic conditions, case load and laboratory capacity, as described in Supplementary Table 1 . Clients with record of having received two doses of BNT162b2, mRNA-1273 or ChAdOx1 on or before the specimen collection date were considered vaccinated; those with no record of vaccination prior to specimen collection were considered unvaccinated. Among respiratory specimens collected for SARS-CoV-2 testing that had both dates available, the median interval between illness onset and respiratory specimen collection was two days with interquartile range of 1-4 days in both provinces. We based primary two-dose VE analyses on second-dose receipt ≥ 14 days before specimen collection, excluding those vaccinated 0-13 days prior. Specimens were sampled from respective provincial databases capturing all RT-PCR testing for SARS-CoV-2 along with client, collection and testing details. Hospitalized cases were identified through linkage with notifiable disease lists, supplemented in Quebec by the administrative hospitalization database. Vaccination information was obtained from provincial immunization registries capturing all SARS-CoV-2 vaccinations along with client and vaccination details. Individual-level database linkages were achieved through unique personal identifiers. Specimens with invalid or missing information were excluded as were specimens collected from individuals: identified as cases before the analysis period; residents of LTCFs, assisted-living or independent-living facilities; vaccinated with a single dose or product other than BNT162b2, All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 26, 2021. ; https://doi.org/10.1101/2021.10.26.21265397 doi: medRxiv preprint mRNA-1273 or ChAdOx1; or when tested outside of the public-funding scheme owing to systematically lower likelihood of test-positivity. 9 The latter criterion also excludes individuals routinely screened for travel. Data linkages and analyses were authorized by the Provincial Health Officer (BC) and National Director of Public Health (Quebec) under respective provincial public health legislation without requirement for research ethics board review. Provincial health authorities provided funding and had no role in the design, results, interpretation or decision to submit. In total, 380,532 specimens contributed to VE analyses in BC including 27,439 (7%) test-positive cases of whom 1582 (6%) were hospitalized (Table 1 ). In Quebec, 854,915 specimens contributed with 17,234 (2%) test-positive cases of whom 878 (5%) were hospitalized. More than 85% of all cases in both provinces accrued during the fourth wave between weeks 31-39. About two-thirds of participating case viruses in each province were genetically-characterized overall, and of those, 91% in BC and 85% in Quebec were the Delta variant (Supplementary Table 1 ). Cases and controls by age and sex were similar in both provinces. Hospitalized cases were older and more often male (Table 1 ). Compared to their share of the population, younger adults and females contributed disproportionately to controls. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. By week 39, >80% of controls in BC and Quebec had been vaccinated with >90% having received two mRNA doses, 3% ChAdOx1 and 5% mixed ChAdOx1/mRNA products (Table 1) . Among mixed ChAdOx1/mRNA recipients, >99% in each province had received ChAdOx1 first. About two-thirds of vaccinated controls in BC and Quebec received BNT-162b; 11% and 20% respectively, received mRNA-1273; and 8% and 3%, respectively, received a mix of either mRNA product of whom 80% and 70%, respectively, received BNT162b2 first. Follow-up periods are shown in Table 1 and by vaccine type in Supplementary Tables 2 and 3. ChAdOx1 recipients in both provinces were generally older and with less follow-up time after their second dose. Table 4 ). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Estimates of mRNA VE against infection improved with a longer interval between first and second doses. With the manufacturer-specified interval of 3-4 weeks, VE in BC was 85% All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 18 ). Confidence intervals were wide for ChAdOx1; in Quebec, but not BC, VE showed gradual increase with longer interval ( Figure 5 ). Recognizing that shorter intervals between doses may have been associated with longer time since second dose, we also explored VE stratified simultaneously for period effects. This is shown for BNT162b2 in Figure 6 and for all vaccine types in Supplementary From two of the larger provinces of Canada, located at nearly opposite ends of the country, we report concordant findings of two-dose SARS-CoV-2 VE against infection and hospitalization, including with mixed vaccines and extended dosing intervals. In both provinces, two homologous or heterologous doses were associated with about 95% reduction in the risk of hospitalization. VE also exceeded 90% against SARS-CoV-2 infection when at least one dose was an mRNA vaccine, but was lower at about 70% when both doses were ChAdOx1. The Delta variant was not All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 26, 2021. ; https://doi.org/10.1101/2021.10.26.21265397 doi: medRxiv preprint associated with compromised VE. Vaccine protection was even better when the spacing between doses was longer than the 3-4 weeks recommended by manufacturers. Over the 5-7-month period after the second dose, mRNA VE was maintained at about 95% against hospitalizations, showing some decline but remaining about 80% or more against infections. With shorter follow-up time for the smaller two-dose ChAdOx1 subset, VE was also maintained to at least the fourth month postvaccination. To date there are no head-to-head RCT comparisons of homologous and heterologous SARS-CoV-2 vaccine efficacy. In immunogenicity studies, heterologous ChAdOx1 followed by mRNA vaccination induced antibody titers exceeding homologous vector vaccination and comparable to homologous mRNA vaccination. 11-13 For pragmatic and immunologic reasons, some countries (e.g. France, Germany, Spain, the United Kingdom (UK)) have, like Canada, recommended and accepted mixed schedules as providing valid two-dose protection. Other countries strictly adhere to the homologous schedules submitted for regulatory approval by sponsor manufacturers. Modification of regulatory approval thereafter typically requires submission of updated data by the sponsor. However, to meet public health needs during a rapidly evolving crisis, decision-makers must be able to respond to emerging post-marketing evidence. In Canada, real-time expert committee recommendation, rather than regulatory review, allowed mixed schedules. By removing the requirement to maintain half of available doses in reserve for homologous series completion, this greatly simplified vaccine deployment and sped vaccine coverage. The findings presented here strongly reinforce vaccine interchangeability to complete the two-dose SARS-CoV-2 vaccine series. Mixed vaccine schedules could reduce logistical hurdles and support nimble vaccination campaigns in countries elsewhere that are still struggling with low vaccine supplies and/or coverage. Global recognition of homologous or heterologous doses as valid proof of All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 26, 2021. ; https://doi.org/10.1101/2021.10.26.21265397 doi: medRxiv preprint vaccination could facilitate a more rapid end to the pandemic and more equitable opportunities for re-opening, travel and return to normal socio-economic interactions, everywhere. Our estimates of two-dose VE against infection exceeding 90% for mRNA but lower at about 70% for ChAdOx1 vaccine are consistent with respective gold-standard RCT findings, [3] [4] [5] and in particular with maximal ChAdOx1 two-dose efficacy against any infection of 66% in pooled RCT meta-analysis. 5 In extended follow-up of participants in the pivotal BNT162b2 RCT, two-dose efficacy against clinical infection peaked at 96% during the first two months but remained >80% between four months and the end of follow-up, similar to the sustained protection we report. 14 In the pivotal mRNA-1273 RCT, efficacy against COVID-19 illness was 93%, without indication of waning across a median of 5·2 months. 15 Of note, following unblinding of the mRNA-1273 RCT, a two-year open-label study was initiated: participants who had originally received placebo were vaccinated (mRNA-1273p) and compared to participants who were earlier randomized to vaccination (mRNA-1273e). 16 24 In the UK, VE against Delta hospitalization was 77% for ChAdOx1 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. week period available (spanning to early or mid-October, respectively), crude surveillance-based RR estimates of 10 and 7 for infection, respectively, and 52 and 24 for hospitalization, respectively, 9,27 correspond to VE estimates >85% against infection and >95% against hospitalization, reassuringly similar to the estimates we report here. We observed improved VE with interval between first and second doses longer than the 3-4 weeks recommended by manufacturers. Immunogenicity studies have shown that longer intervals are conducive to more complete maturation of the immune response after the first dose, stronger response to the second dose, and ultimately higher and more durable SARS-CoV-2 antibody levels. 28, 29 In the UK, VE was consistently higher with an interval >45 days compared to [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] days between BNT162b2 doses, with more erratic findings for ChAdOx1. 28 In the pooled-analysis All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted October 26, 2021. ; https://doi.org/10.1101/2021.10.26.21265397 doi: medRxiv preprint of ChAdOx1 RCTS, however, efficacy was much higher with an interval >8 weeks. 5 Also from the UK, Pouwels et al. found no evidence that BNT162b2 or ChAdOx1 effectiveness varied when comparing dichotomous intervals <9 or ≥ 9 weeks between doses; 30 however, their broad categorization <9 weeks may have diluted the lower VE associated with shorter intervals. The optimal interval between doses ultimately represents a balance between rapid and enhanced protection. During a surging pandemic wave, rapid administration of the second dose may prevent some cases that would occur with a longer wait but otherwise, surveillance data in most countries suggest <1% of the unvaccinated population were infected during a given four-week period of the pandemic. With substantial single-dose protection against hospitalization, the absolute risk of severe outcome associated with waiting a few more weeks for the second dose would be small in that context. Conversely, the more durable immunity and approximate 5-10% increment in VE we observed with 7-8-week interval between doses could pay dividends into the future, ultimately preventing more cases and hospitalizations (depending upon evolving incidence and duration of protection). In most instances, therefore, an interval of 7-8 weeks between the first and second dose seems optimal for mRNA vaccines, not only to maximize single-dose coverage in the context of vaccine scarcity but also to optimize the second-dose booster response. This study, based upon general laboratory submissions, surveillance registries and administrative data, has limitations. In particular, such data are subject to missing or incomplete information, misclassification and selection bias. The test-negative design partially standardizes for healthcare seeking behaviours, but testing indications for SARS-CoV-2 are broad and discretionary, and case ascertainment will have varied between population sub-groups and over time. With increasing vaccine coverage, the subset of remaining unvaccinated individuals may be less comparable in their likelihood of testing positive, with the direction of such bias unknown, and likely to vary All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. with other public health measures. VE estimates were adjusted for calendar time (and age, sex, region of residence) but information pertaining to other potential confounders such as comorbidity and socio-economic status were not readily available. We cannot rule out residual confounding. Higher-risk healthcare workers or immunocompromised individuals targeted for more rapid second-dose receipt may have contributed to the lower VE associated with shorter interval between doses; however, weighted by their small percentage of the population, such under-estimation would be minor. Further reduced sample size with additional stratification affects the precision of VE estimates, requiring cautious interpretation. Finally, to address differential likelihood of vaccination and exposure risk, our studies were conducted in community-dwelling adults; results may not be generalizable to residents of LTCFs and nursing homes. In conclusion, two doses of mRNA and/or ChAdOx1 vaccines provided powerful and persistent protection against hospitalization, including due to the Delta variant, without sign of decline by 5-7 months post-vaccination among community-dwelling adults, including older adults. VE against infection declined from an earlier post-vaccination peak above 90% but still prevented 80% or more of infections by the seventh month post-vaccination. Extending the interval between first and second doses may have optimized booster dose protection in Canada. Given these findings, the need and timing of a third dose warrant serious reflection by decision-makers, especially since two-dose, or even one-dose, coverages still remain low in many areas of the world. Our findings support mixed SARS-CoV-2 vaccine schedules and extended intervals between doses, each of which may improve vaccine coverage and have health, equity and access implications globally. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Provincial health authorities provided funding but had no role in the design, results, interpretation or decision to submit. We thank Shinhye Kim at the BC Centre for Disease Control for support in manuscript assembly and the summary tabulation of findings. Manish Sadarangani acknowledges general salary support provided to him by awards from the BC Children's Hospital Foundation, the Canadian Child Health Clinician Scientist Program and the Michael Smith Foundation for Health Research. Denis Talbot was recipient of a Career Award from the Fond de recherche du Québec-Santé. Finally, we thank the many frontline, regional and provincial practitioners, including clinical, laboratory and public health providers, epidemiologists, Medical Health Officers, laboratory staff, vaccinators, participants and others who contributed to the epidemiological, virological and genetic characterization data underpinning these analyses. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. [Accessed 16 October 2021]. Available from: https://www.canada.ca/en/publichealth/services/immunization/national-advisory-committee-on-immunizationnaci/extended-dose-intervals-covid-19-vaccines-early-rollout-population-protection.html All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 4 (overall) and Supplementary Tables 8 and 9 (Delta and other variants of concern). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Shown are adjusted vaccine effectiveness (VE) estimates and 95% confidence intervals (CI) against infection (blue) and hospitalization (orange) by time between receipt of the second dose of any mRNA vaccine and specimen collection, among adults ≥ 70 years old in British Columbia (BC) and Quebec, the latter displayed as dashed lines. All VE estimates are adjusted for age group (70-79, ≥ 80 years); sex (men, women); individual epidemiological week of the analysis period (weeks 22-39, categorical); and region of the province (5 in each province) except in Quebec for which calendar time adjustment was tri-weekly for the hospitalization outcome owing to small sample size. For additional details including corresponding sample sizes and precise unadjusted and adjusted estimates and 95%CI, see Supplementary Table 16 (by mRNA vaccine type). All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Shown are adjusted vaccine effectiveness (VE) estimates and 95% confidence intervals (CI) against infection (timed on specimen collection date) by interval between the first and second dose (3-4 weeks in purple; 5-6 weeks in dashed yellow; 7+ weeks in green) and time since the second dose among adults ≥ 18 years old who were vaccinated with two BNT162b2 vaccines in the provinces of British Columbia (BC) and Quebec. All VE estimates are adjusted for age group (18-49, 50-69, 70-79, ≥ 80 years); sex (men, women); individual epidemiological week of the analysis period (weeks 22-39, categorical); and region of the province (5 in each province). For additional details including corresponding sample sizes and precise unadjusted and adjusted estimates and 95%CI, see Supplementary Table 19 (by vaccine type, including mRNA combined and by mRNA vaccine type, as well as ChAdOx1). Recommendations on the use of COVID-19 vaccines. Ottawa: NACI Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine Immune responses against SARS-CoV-2 variants after heterologous and homologous ChAdOx1 nCoV-19/BNT162b2 vaccination Six month safety and efficacy of the BNT162b2 mRNA COVID-19 vaccine. medRxiv preprint Efficacy of the mRNA-1273 SARS-CoV-2 vaccine at completion of blinded Phase Covid-19 in the phase 3 trial of mRNA-1273 during the Delta-variant surge. medRxiv 2021 Effectiveness of Covid-19 Vaccines in Ambulatory and Inpatient Care Settings Effectiveness of Covid-19 vaccines against the B.1.617.2 (Delta) variant Prevention of COVID-19 by mRNA-based vaccines within the general population of California FDA-authorized mRNA COVID-19 vaccines are effective per real-world evidence synthesized across a multi-state health system 1 Unless otherwise specified, displayed is the percentage of cases, hospitalized cases or controls who received a second vaccine dose by on or before specimen collection, by row category, regardless of time since second dose. 2 Displayed is the percentage of cases who were hospitalized. 3 All percentages displayed below this row are column % 4 One case twice vaccinated with unspecified mRNA vaccines in British Columbia. 5 Twelve controls twice vaccinated with unspecified mRNA vaccines in British Columbia