We used two approaches to estimate the effect of vaccination on the delta variant. First, we used a negative test case-control model to estimate vaccine efficacy against symptomatic disease caused by the delta variant, compared to the alpha variant, over the period of circulation of the delta variant. This approach has been described in detail elsewhere.ten In short, we compared the vaccination status of people with symptomatic Covid-19 with the vaccination status of people who reported symptoms but tested negative. This approach makes it possible to control biases related to health-seeking behavior, access to tests and identification of cases.
For the secondary analysis, the proportion of people with cases caused by the delta variant compared to the main circulating virus (the alpha variant) was estimated according to vaccination status. The underlying assumption was that if the vaccine had some efficacy and was equally effective against each variant, a similar proportion of cases with either variant would be expected in unvaccinated and vaccinated people. Conversely, if the vaccine was less effective against the delta variant than against the alpha variant, the delta variant should represent a higher proportion of cases occurring more than 3 weeks after vaccination than in unvaccinated persons. Details of this analysis are described in Section S1 of the Supplementary Annex, available with the full text of this article on NEJM.org. The authors guarantee the accuracy and completeness of the data and the reliability of the trial to the protocol.
Data on all people in England who have been vaccinated with Covid-19 vaccines are available in a National Immunization Management System. Data for vaccinations that had taken place up to May 16, 2021, including the date of receipt of each vaccine dose and the type of vaccine, were retrieved on May 17, 2021. Vaccination status was categorized as receipt one dose of vaccine among people who onset of symptoms occurred 21 days or more after receiving the first dose until the day before receiving the second dose, such as receiving the second dose in people whose onset of symptoms occurred 14 days or more after receiving the second dose, and as receiving the first or second dose in people whose symptoms began 21 days or more after receiving the first dose ( including any period after receiving the second dose).
Polymerase chain reaction (PCR) testing for SARS-CoV-2 in the UK is performed by hospital and public health laboratories, as well as community testing using drive-thru or at-home testing , which are available to anyone with symptoms compatible with Covid-19 (high temperature, continued new cough, or loss or change in smell or taste). Data from all positive PCR tests between October 26, 2020 and May 16, 2021 have been extracted. Data on all negative community tests recorded among those reporting symptoms were also extracted for negative test case-control analysis. Children under the age of 16 on March 21, 2021 have been excluded. Data were limited to those who reported symptoms, and only those who had tests within 10 days of symptom onset were included, to account for the reduced sensitivity of PCR tests beyond this. period.25
Whole genome sequencing was used to identify the delta and alpha variants. The proportion of all positive samples that were sequenced increased from around 10% in February 2021 to around 60% in May 2021.4 Sequencing is performed in a network of laboratories, including the Wellcome Sanger Institute, where a high proportion of samples have been tested, and whole genome sequences are assigned to Public Health England variant definitions on the basis of mutations.26
The target status of the Spike gene on PCR was used as a second approach to identify each variant. The laboratories used the TaqPath test (Thermo Fisher Scientific) to test three target genes: spike (S), nucleocapsid (NOT), and the open reading frame 1ab (ORF1ab). In December 2020, it was noted that the alpha variant was associated with negative tests on the S target, therefore S the target-negative status was then used as a proxy for the identification of the variant. The alpha variant represents between 98% and 100% of S negative target results in England. Among the sequenced samples that tested positive for S Target, the delta variant was present in 72.2% of samples in April 2021 and in 93.0% in May (as of May 12, 2021).4 For the negative test case-control analysis, only samples that had been tested in laboratories with the use of the TaqPath assay were included.
The three data sources described above were related to the use of the National Health Service number (a unique identifier for each person receiving medical care in the UK). These data sources were also linked with data on date of birth, last name, first name, postal code, and patient sample identifiers and dates.
Multiple covariates that may be associated with the likelihood of being offered or accepted a vaccine and the risk of exposure to Covid-19 or specifically to one of the variants analyzed were also extracted from the National Immunization Management System and test data. These data included age (in 10-year age groups), sex, multiple deprivation index (a national indication of the level of deprivation based on small geographic areas of residence,27 assessed in quintiles), race or ethnicity, nursing home status, overseas travel history (i.e. outside the UK or Ireland) , geographic region, time period (calendar week), social and health worker status, and status of being in a clinically extremely vulnerable group.28 In addition, for the negative test case-control analysis, a history of SARS-CoV-2 infection before the start of the vaccination program was included. People were considered to have traveled if, when requesting a test, they claimed to have traveled outside the UK and Ireland within the previous 14 days or if they had been tested in a quarantine hotel or so ‘they were in home quarantine. Postal codes were used to determine the multiple deprivation index, and unique property reference numbers were used to identify nursing homes.29
For the negative test case-control analysis, a logistic regression was used to estimate the odds of having a symptomatic case of Covid-19 confirmed by PCR among vaccinated people compared to unvaccinated people (control). The cases were identified as having the delta variant by sequencing or if they were S positive target on the TaqPath PCR assay. The cases were identified as having the alpha variant by sequencing or if they were S negative target on the TaqPath PCR test.
If a person had tested positive multiple times during a 90-day period (which may represent a single episode of illness), only the first positive test was included. A maximum of three randomly selected negative test results were included for each person. Negative tests in which the sample was taken within 3 weeks before a positive result or after a positive result could have been false negatives; therefore, these were excluded. Tests that had been administered within 7 days of a previous negative result were also excluded. People who had already tested positive before the analysis period were also excluded in order to estimate the effectiveness of the vaccine in fully susceptible people. All covariates were included in the model as done with previous test negative case-control analyzes, with calendar week included as a factor and no interaction with region.
In regards to S target status positive or negative, only people who had tested positive on the other two PCR genetic targets were included. Assignment to the delta variant based on S the target status has been limited to the week starting April 12, 2021 and beyond in order to aim for a high specificity of S positive target test for the delta variant.4
The vaccine’s effectiveness for the first dose was estimated in people whose symptom onset date was 21 days or more after receiving the first dose of vaccine, and the effects of the vaccine for the second dose were estimated in people whose symptom onset date was 14 days or more after receiving the second dose. A comparison was made with unvaccinated people and with people who experienced symptoms within 4 to 13 days after vaccination to help explain the differences in underlying risk of infection. The period from the day of vaccine administration (day 0) to day 3 has been excluded because reactogenicity of the vaccine may lead to an increase in testing which skews the results, as previously described.ten