SAFE Project: Background rates of adverse events of special interest (AESIs) for COVID-19 vaccination

The first COVID-19 vaccines became available at the end of 2020 and mass vaccination programmes were implemented rapidly across many countries. Careful safety monitoring for adverse events of special interest (AESIs) following vaccination was, and continues, to be a critical activity. To understand if increased AESI incidences are attributable to COVID-19 vaccination/s, baseline reference data from before vaccine deployment is required. Baseline data serve as preparation for signal verification and risk assessment should they be required during the programme as well as risk communication. As with other countries, New Zealand did not have the baseline data required to form the basis of robust vaccine safety monitoring.

In 2018 the Coalition for Epidemic Preparedness and Innovation (CEPI) contracted with the Brighton Collaboration, through The Task Force for Global Health, to harmonise the safety assessment of CEPI-funded vaccines via its Safety Platform for Emergency vACcines (SPEAC) project. In response to the pandemic SPEAC established and prioritised a set of AESIs for COVID-19 vaccines. SPEAC and the Brighton Collaboration have also engaged in a long-term programme establishing case definitions for AESIs following vaccination. Internationally agreed case definitions for each of the AESIs enable consistent assessment and diagnosis of AESIs across multiple healthcare networks and countries, support measurement of baseline rates for the conditions and reporting and contributes to harmonised vaccine safety surveillance. The Brighton Collaboration case definitions provide key criteria to define the level of diagnostic certainty that the presentation is a case.

The outcomes of the SAFE Project helped prepare New Zealand for local assessment of COVID-19 vaccine safety by establishing baseline rates for 20 AESIs extracted from the SPEAC prioritised list and myocarditis, pericarditis, multisystem inflammatory syndrome, a range of haematological conditions potentially associated with the newly identified thrombosis with thrombocytopenia syndrome (TTS), herpes zoster (shingles), narcolepsy, and sudden death for the New Zealand population overall and key subgroups from 2008–2019. These background rates may be used as a first step to contextualise data from prospective monitoring studies, spontaneous reports from the Centre for Adverse Reaction Monitoring (CARM) and other databases, and case reports, and thereby form a basis for identifying potential COVID-19 vaccine safety signals. With the baseline established, signals may be verified through the conduction of observed over expected analysis with the same population cohort.

There are two components to the SAFE Project. First, the Background rates of AESIs in New Zealand 2008–2019 reports on the incidence of predefined AESIs and a range of other conditions of interest. Second, the Assessing accuracy of search strategies for AESIs reports on clinical record assessment provides insight into the disease codes used in the administrative health data. The primary objectives were to estimate the incidence of predefined AESIs and other conditions of interest in the general population by calendar year, age band, sex, ethnicity, deprivation, and region over the period 2008–2019 and use clinical record assessment to identify the accuracy (positive predictive value (PPV)) of the ICD-10-AM codes used in the health data to estimate the incidence of AESIs.

The Background rates of AESIs in New Zealand 2008–2019 also reports on the prevalence of predefined high-risk medical conditions for developing severe COVID-19 disease.

 *This project was funded by the New Zealand Ministry of Health.

 Click here to download the report Part 1: Background rates of AESIs in New Zealand 2008–2019, and click here to download the report appendices.

 Click here to download the report Part 2: Assessing accuracy of search strategies for AESI.

You will need Adobe Acrobat Reader to open the files.

Data dashboards

The GVDN® developed protocols for partner sites to implement and determine background rates for health outcomes of interest, carry out rapid cycle analysis and perform observed versus expected analyses for adverse events of special interest (AESI), and implement association studies to identify the relationship between vaccination and a potential vaccine safety signal.

Data from GVDN partner sites conducting these studies will be displayed on interactive dashboards as they become available. The GVDN Background Rates Dashboard with background rates data on 13 AESI from 10 sites/countries can be viewed from our Resources/Data Dashboards page.

The GVDN uses Shiny applications to create interactive data dashboards that allow users to interact with and visualise study data as they become available. Each dashboard is embedded in a GVDN webpage, and is self-contained, hosted by RStudio on shared cloud servers, and accessed via SSL (Secure Sockets Layer). Data for each dashboard has been uploaded to the dashboard by the GVDN.

Dashboard data are deidentified. With data for rare events, to avoid recording potentially identifying low counts or suppression of counts below a threshold, such as counts <6, some age groups will have been aggregated up to 20-year age groups, to increase counts.

The currently available dashboard provides the Background rates of AESIs in New Zealand 2008–2019 from the SAFE Project and can be accessed through the link below.

Key Information

Global Vaccine Data Network™ (GVDN®) is a multinational, investigator-led research consortium with primary interests in the area of vaccine safety and effectiveness.

We provide the Dashboard(s) for informational purposes only. We cannot guarantee that the information is suitable for any particular purpose or that it is error-free.

Do not rely on Dashboard Information for any health decision, except in consultation with your clinician.

Where you use any Dashboard Information, you must ensure that any research and conclusions you draw reflect generally accepted scientific methodologies and other principles of responsible research conduct, and that any reports or publications are fair and accurate and not likely to mislead or deceive any person.

We take breaches of these Terms of Use seriously.  Where you breach these Terms of Use, we may take legal action against you to recover compensation for any harm you cause as a result.

If you have any questions about our Terms of Use or our Privacy Policy please contact us. We have provided an overview of our Terms of Use below.

You can read our full Terms of Use here, and our Privacy Policy here.


Please click on the links to find out more detail on any particular topic.

  • Introduction and Acceptance of Terms of Use: These Terms of Use and our Privacy Policy set out how we will work with you. You accept these Terms of Use when you access the GVDN Website and/or Dashboard(s) and/or the Protocol(s) and and/or the Blog(s).
  • Changes to these Terms of Use: We may update these Terms of Use at any time.
  • Definitions: Capitalised words used in these Terms of Use have certain meanings.
  • GVDN: We provide the Dashboard(s) for informational purposes only. We cannot guarantee that the information is suitable for any particular purpose or that it is error-free.
  • Use of Dashboard Information and Protocols, and/or Blog(s): Dashboard Information and Protocol(s) may be used only for academic research and other non-commercial purposes. Blog(s) may be reproduced or republished only for the purpose of science communication and the source must be acknowledged with the original URL link.
  • Publications: Reports, publications, extracts and summaries must comply with certain requirements.
  • Blog(s) Content: In submitting a Blog for publication by GVDN, the author(s) warrants they have the appropriate expertise to comment on the subject matter, the Blog is original work and evidence-based, and that the Blog contains no violation of existing copyright it other third party right.
  • Conditions of Use: Use of the GVDN Website and/or Dashboard(s) and/or Protocol(s) and/or (Blog(s) is conditional on a number of requirements guarding against misuse, including prohibitions against commercialisation, copying, linking to the GVDN Website, re-identifying individuals, scraping, hijacking and other harms.
  • Other Content: Blogs and articles on the GVDN Website are the views of the author(s), not our views.
  • Exclusions of Liability: We do not accept liability for your use of the GVDN Website and/or any Dashboard(s) Information and/or Protocol(s) and/or Blog(s).
  • Rights of Access: Your right to access the GVDN Website and Dashboard(s) is temporary and not guaranteed.
  • Indemnity: You indemnify us against loss, etc. resulting from or incurred by us in connection with your breach of these Terms of Use.
  • Fees: There are no fees for using Dashboard Information or Protocol(s) for academic research and other non-commercial purposes. Licence fees and terms apply to use for commercial purposes.
  • Ownership of Content and Trade Marks: We own all rights, including intellectual property rights, in the GVDN trade marks, and the content of the GVDN Website including the Dashboard(s), Dashboard Information, Protocol(s), and Blog(s). You must not use these except as stated or with our consent.
  • Privacy: We will use and protect your personal information as set out in our Privacy Policy.
  • General: These Terms of Use are governed by New Zealand law.

Brief guide for using the dashboard

After selecting the site from a dropdown menu, data can be viewed in graph or table format by AESI and patient type as combined data, or by age, sex, or age-sex, and by year(s). Information describing the site, data source(s), active population, and patient type has been provided by each site and is available in the third tab.

  To resize the dashboard to fit within your viewing screen, click in the dashboard area, then hold down the Control button on your keyboard and scroll up or down.

Clicking on the question icon beside each data set provides additional information related to the variables in that set. 

In the graph view, tab one, clicking on the camera icon will download a picture (PNG) file of the plot. It is possible to zoom into the plot, click and hold the right mouse button while drawing a marquee around the area of the plot you wish to zoom into, then release the button. Clicking on the house icon will reset the plot axes.  While in the graph view, hovering the mouse over a data point will show the closest data numerically.

Data provided in the table view, tab two, can be downloaded as a spreadsheet (XLXS) file by clicking on this button .

In tab three, information describing the site, data source(s), active population, and patient type has been provided by each site.

The data dashboard will time-out and the screen will become grey if the dashboard is left idle for too long. Refreshing the webpage will activate the dashboard again, previous data selections will need to be re-entered.

SAFE Project: Background rates of AESIs in New Zealand 2008–2019 dashboard