Effectiveness of BNT162b2 and mRNA-1273 covid-19 vaccines against symptomatic SARS-CoV-2 infection and severe covid-19 outcomes in Ontario, Canada: test negative design study

Abstract Objective To estimate the effectiveness of mRNA covid-19 vaccines against symptomatic infection and severe outcomes (hospital admission or death). Design Test negative design study. Setting Ontario, Canada between 14 December 2020 and 19 April 2021. Participants 324 033 community dwelling people aged ≥16 years who had symptoms of covid-19 and were tested for SARS-CoV-2. Interventions BNT162b2 (Pfizer-BioNTech) or mRNA-1273 (Moderna) vaccine. Main outcome measures Laboratory confirmed SARS-CoV-2 by reverse transcription polymerase chain reaction (RT-PCR) and hospital admissions and deaths associated with SARS-CoV-2 infection. Multivariable logistic regression was adjusted for personal and clinical characteristics associated with SARS-CoV-2 and vaccine receipt to estimate vaccine effectiveness against symptomatic infection and severe outcomes. Results Of 324 033 people with symptoms, 53 270 (16.4%) were positive for SARS-CoV-2 and 21 272 (6.6%) received at least one dose of vaccine. Among participants who tested positive, 2479 (4.7%) were admitted to hospital or died. Vaccine effectiveness against symptomatic infection observed ≥14 days after one dose was 60% (95% confidence interval 57% to 64%), increasing from 48% (41% to 54%) at 14-20 days after one dose to 71% (63% to 78%) at 35-41 days. Vaccine effectiveness observed ≥7 days after two doses was 91% (89% to 93%). Vaccine effectiveness against hospital admission or death observed ≥14 days after one dose was 70% (60% to 77%), increasing from 62% (44% to 75%) at 14-20 days to 91% (73% to 97%) at ≥35 days, whereas vaccine effectiveness observed ≥7 days after two doses was 98% (88% to 100%). For adults aged ≥70 years, vaccine effectiveness estimates were observed to be lower for intervals shortly after one dose but were comparable to those for younger people for all intervals after 28 days. After two doses, high vaccine effectiveness was observed against variants with the E484K mutation. Conclusions Two doses of mRNA covid-19 vaccines were observed to be highly effective against symptomatic infection and severe outcomes. Vaccine effectiveness of one dose was observed to be lower, particularly for older adults shortly after the first dose.

asymptomatic terms were presented in a record, the individual was classified as 'symptomatic'.
Among all eligible individuals tested for SARS-CoV-2 during the study period (n=2,137,686), 39% had a test record that noted whether the individual was symptomatic for COVID-19 or asymptomatic at the time of testing. For the remainder, these fields were blank, contained irrelevant information (e.g., indication of test [e.g., pre-op], targeted patient population [e.g., health care worker, close contact]), or recorded symptoms not consistent with COVID-19 (e.g., anxiety, falls).

CONFIDENTIAL -NOT FOR DISTRIBUTION 5 AUG 2021
6 eTable 1: List of covariates used in the analyses.

Variable Definition Age
Age was determined from the Registered Persons Database. This variable was included a priori as hypothesized to be directly related to COVID-19 infection risk. [2] Sex Sex was determined from the Registered Persons Database. This variable was included a priori as hypothesized to be directly related to COVID-19 infection risk. [2] Biweekly period of COVID-19 test Based on the index date (i.e. specimen collection date, or date of severe outcome if before specimen collection date): 14 Dec 2020 to 27 Dec 2020 28 Dec 2020 to 10 19 Apr 2021 Chronic heart disease Individuals were defined as having "chronic heart disease" if they had congestive heart failure (CHF), ischemic heart disease, or atrial fibrillation. The definitions for these conditions are as follows: CHF: [3] An ICES-derived CHF database was used to identify patients with CHF, based on 1 NACRS, DAD, SDS, or OHIP claim and a second claim (from either) in 1 year. The CHF database is limited to those aged 40 years or older. This variable was included a priori as hypothesized to be directly related to COVID-19 infection risk. [4] Variable Definition Atrial fibrillation: [6] Individuals with 1 hospitalization or 4 MD visits within a year in the past 5 years with the following codes: ICD-9: 427.31, 427.32 ICD-10: I48 OHIP dxcode: 427 Chronic respiratory disease Asthma: [7] An ICES-specific asthma database was used to identify patients with asthma, based on 2 or more ambulatory care visits and/or 1 or more hospitalizations. This variable was included a priori as hypothesized to be directly related to COVID-19 infection risk, as a result of its relationship to severe COVID-19 outcomes. [4] OHIP OHIP diagnostic code: 493 DAD ICD-9 diagnostic code: 493 ICD-10 diagnostic codes: J45, J46 Chronic obstructive pulmonary disease (COPD): [8] An ICES-specific COPD database was used to identify patients with COPD, based on 1 or more ambulatory care visits and/or 1 or more hospitalizations. The algorithm to identify COPD patients was only validated in those aged 35 years or older. This variable was included a priori as hypothesized to be directly related to COVID-19 infection risk. [4] OHIP OHIP diagnostic codes: 491, 492, 496   DAD  ICD-9 diagnostic codes: 491, 492, 496  ICD-10 diagnostic codes: J41, J42, J43, J44  Hypertension An ICES-specific hypertension database was used to identify patients with hypertension, based on 1 or more DAD diagnoses or 2 or more OHIP diagnoses in a two-year period; or 1 OHIP diagnosis followed by an OHIP/DAD diagnosis within two years. [9] This variable was included a priori as hypothesized to be directly related to COVID-19 infection risk. [4] DAD We included immunosuppressive conditions a priori as hypothesized to be directly related to COVID-19 infection risk. [4] HIV: [11] An ICES-specific HIV database was used to identify patients with HIV, based on 3 physician claims in 3 years with OHIP diagnostic codes: 042, 043 or 044 Solid organ transplant recipients: CORRLINK is an ICES-specific database which links CORR (Canadian Organ Replacement Register) and DAD data. This database only includes patients that have received an organ transplant and does not include dialysis patients.
• For transplants before December 31, 2019: individuals are a transplant recipient if they have a treatment code of 171, where the treatment was before the index date • For transplants on/after January 1, 2020: Identify ICD-10 codes, CCI procedure codes, and OHIP feecodes from DAD, NACRS, and OHIP (codes available upon request) Allogenic/autologous bone marrow transplant recipients: We identified those who had a history of allogenic bone marrow transplant before the index date using the following combination of diagnostic codes: DAD: • CCP procedure codes = 53.0 • CCI procedure codes = 1WY19, 1LZ19HHU7, 1LZ19HHU8 OHIP: • Feecode = Z426 Other immune disorders: Individuals were identified as having disorders of the immune system based on health care encounters recorded in DAD, SDS, NACRS, and OHIP in the 2-years prior to index using Expanded Diagnostic Clusters from the Johns Hopkins ACGⓇ System Version 10. [12] Any hospitalization (any diagnosis field) with the following codes: • Sickle-cell disease (ICD-10 D57.0 -D57.  [15,16] • Psoriasis/psoriatic arthritis [17] o  [18] o Individuals with one hospitalization or 5 physician billings over 2 years. DAD: ICD-9: 340; ICD-10: G35. OHIP: dxcode = 340.

CONFIDENTIAL -NOT FOR DISTRIBUTION 5 AUG 2021
Variable Definition ICD-10: I85.9, I98.2, K70.3,K71.7, K74.6 Defined using the Decompensated Cirrhosis Algorithm 5 (from [21]): One or more physician visits with diagnosis code 571 and (one or more hospital diagnosis or one or more procedure), using the following diagnostic codes: We assigned subjects to a DA using postal code, as recorded in the Registered Persons Database.

Household income quintile
Calculated at the DA level using 2016 Census data by multiplying the median income (before-tax) by the number of households and dividing by the sum of single-person equivalent to obtain income per single person equivalent. [26] For DAs where median income was unavailable,

Variable
Definition neighbouring DAs were used to estimate income per single person equivalent. DA-based income quintiles were constructed separately for each census metropolitan area or census agglomeration (one or more adjacent municipalities integrated via commuting flows). DAs within each such area were ranked from the lowest average income per single-person equivalent to the highest, and DAs were assigned to five groups, such that each group contained approximately one-fifth the total in-scope population of each area.

Persons per dwelling quintile
Average number of persons in private households, calculated at the DA level using the 2016 Census data. [27] DAs across the province were ranked by average number of persons per household into 5 categories (quintiles), such that each group contained approximately one-fifth of the DAs.
Essential worker quintile Calculated at the DA level, using 2016 Census data. [28] For each DA, we calculated the number of individuals ≥15 years old that were working in one of the following Census-defined work categories: Sales and service occupations; trades, transport and equipment operators and related occupations; natural resources, agriculture and related production occupations; and occupations in manufacturing and utilities.
DAs across the province were then ranked by these percentages into quintiles, with the lowest 1/5 of DAs comprising the first quintile, and so on.
DAs across the province were then ranked by these percentages into quintiles, with the lowest 1/5 of DAs comprising the first quintile, and so on. Proportion reported, unless stated otherwise. b The sum of counts does not equal the column total because of individuals with missing information (<1.0%) for this characteristic.