Study setting and study population
The study was conducted in the State of Qatar between January 1, 2021 and June 12, 2022. Qatar has a robust, centralized national mortality tracking system to examine all deaths in the country. All deaths in Qatar are individually reviewed by trained licensed physicians dedicated to mortality reviews who extract medical data from the electronic health records and assign direct, antecedent, and contributing causes of death for each case using the World Health Organization recommended definitions13. All deaths are processed in the public sector health facilities in Qatar which use an interconnected single electronic health records platform.
For the current study, medical records of all decedents during the study period who received any SARS-CoV-2 vaccine were reviewed. COVID-19 vaccination data for all decedents were retrieved from the Qatar National COVID-19 Database3,4,5,6. Those who had died within 30 days of receiving a SARS-CoV-2 vaccine dose were eligible for inclusion in the study.
Vaccination and COVID-19 testing data
This study was conducted in the resident population of Qatar. COVID-19 laboratory testing, vaccination, and clinical infection data were extracted from the integrated, nationwide, digital-health information platform that hosts the national, federated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) databases. These databases are complete with no missing information for polymerase chain reaction testing, COVID-19 vaccinations, COVID-19 hospitalizations, and basic demographic details, and have captured all SARS-CoV-2-related data since epidemic onset. Nearly all individuals were vaccinated in Qatar, through the universal public healthcare system for all nationals and residents of Qatar. For rare individuals who received COVID-19 vaccination outside Qatar, vaccination details were recorded in the health system at the port of entry upon return to Qatar, in order to fulfill national requirements and to benefit from privileges associated with vaccination, such as quarantine exemption5,14,15,16,17,18.
Ascertainment of probability of death related to vaccination
We developed an algorithm to determine the probability of a death being related to COVID-19 vaccine administration using a framework similar to the one proposed by the World Health Organization, which takes into account evidence of any other cause of death, temporal association with vaccination, and lack of any other possible explanation19,20 (Fig. 2). Four licensed physicians were trained on the algorithm criteria to independently assign the probability of a death being related to vaccine administration. Final categorization was based on at least three reviewers assigning the same level of probability. Where less than three reviewers assigned the same probability of association, one additional reviewer adjudicated the outcome.
We created four categories of probability of association with SARS-CoV-2 vaccination: Not Related; Low Probability; Intermediate Probability; and High Probability. Cases with more than 2 levels of probability assigned by the four primary reviewers were classified as indeterminate. If a clear alternate and completely unrelated cause of death was identified (e.g., road traffic accident, suicide/homicide, widely metastatic cancer under palliative care), the death was classified as being not related to the vaccination. Presence of one or more severe underlying conditions associated with high risk of mortality (e.g., chronic advanced heart failure, pre-existing atherosclerotic heart disease with prior major adverse cardiovascular events) and physician documentation in the medical records of those contributing directly to death were used to assign low probability, while presence of one or more stable underlying conditions with low risk of short-term mortality and study reviewers’ confirmation of no clear alternate cause of death were used to assign intermediate probability of relationship to vaccination. Those with no underlying conditions and no plausible risk factor for death noted in the medical records or identified by reviewers were further subclassified based on time of death from last SARS-CoV-2 vaccination. Deaths occurring within 5 days of vaccination were classified as having high probability, deaths between 6–15 days as intermediate probability, and deaths occurring 16–30 days as having low probability of being causally associated with vaccination (Fig. 2).
All cases determined to have high or intermediate probability of association with SARS-CoV-2 vaccination were further reviewed to ascertain the underlying and contributory cause(s) of death by two trained mortality reviewers using the World Health Organization classification system. This review was specifically intended to identify the cause(s) that directly led to the precipitation of the terminal event that resulted in death.
Data were collected and analyzed in Microsoft Excel for Microsoft 365, Microsoft Corporation, Redmond, WA, USA.
The study was approved by the Institutional Review Board at Hamad Medical Corporation. A waiver of informed consent was granted for the study since only decedents records were reviewed and all identifiers were subsequently removed.
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.