FOI release

Freedom of Information request on the MHRA’s process of checking and evaluating Yellow Cards, specifically with fatal reports showing discrepancies (FOI 21-593)

Published 13 August 2021

24th June 2021 FOI 21/593

Dear

Thank you for your FOI request dated 1st June 2021, whereby you requested information regarding the MHRA’s process of checking and evaluating Yellow Cards, specifically with fatal reports showing discrepancies.

1a) Can you give details of the process by which Yellow Card data are checked for consistency and validity?

1b) Is it checked by a human or is AI used in the evaluation?

1c) Does it happen when the report is submitted, after it or not at all?

Please be reassured that the MHRA takes all reports of adverse reactions with the utmost seriousness, including those reporting fatal events in patients who have received a COVID-19 vaccine. Where these events are reported to us, we follow up to find out full details of the events, the cause of death and, where applicable, further information of any post-mortem findings. Additionally, if a discrepancy is noted in a Yellow Card report we can follow-up with the reporter for confirmation of the information provided. We are dedicated to ensuring our online reporting tool is as easy as possible for people to use. We are working hard to rectify any errors in the data we have received via the Yellow Card scheme as soon as possible.

The MHRA is using artificial intelligence (AI) software to reduce the amount of manual coding for each report, thereby saving resources in processing cases and ensuring they are rapidly available for scientific analysis. The tool is not being used for assessment of data, but to help ensure that all the information from the reporter is well structured to support analysis and is subject to robust quality assessment. This specific AI tool is for the surveillance of COVID-19 vaccines due to the size and scale of the vaccination campaign, rather than because of concerns about the safety profile. Once reports are received, a team of safety experts look at each individual adverse drug report to individually assess each case, together with additional sources of evidence. Additionally, we apply statistical techniques that can tell us if we are seeing more events than we would expect to see based on what is known about background rates of illness in the absence of vaccination. We supplement this form of safety monitoring with other epidemiology studies including analysis of data on national vaccine usage, anonymised GP-based electronic healthcare records and other healthcare data to proactively monitor safety. Furthermore, we also take into account the international experience based on data from other countries using the same vaccines. We also receive independent advice from the Commission on Human Medicines (CHM) which is responsible for advising on the impact of any safety issues including fatal cases on the balance of risks and benefits of COVID-19 vaccines. For further information on the CHM please follow this link: https://www.gov.uk/government/organisations/commission-on-human-medicines/about

2a) How many ADRs are listed across all MHRA regulated products where the reaction is by definition fatal?

2b) Of these, how many were non-fatal?

As you are aware, this request is covered by Section 21 of the Freedom of Information Act (information reasonably accessible to the applicant by other means) and the information you have asked for is therefore exempt from disclosure.

3) Are ADRs involving non-fatal deaths evaluated or investigated?

See response to question 1.

4) If investigations reveal invalid data, what happens to the Yellow Card report(s)?

Yellow card reports require reporter details, a patient identifier, suspect medication/vaccine and adverse reaction details in order for it to be considered a valid report. As reports are reviewed, any reports that are suspected to have provided false reporter or patient information for example can be followed up for confirmation. Any fake reports are removed from the database. However, please note that all reports are kept on the database and assessed routinely unless they lack the information for a valid report or are clearly fake. We encourage everyone to report suspected adverse reactions to the scheme and patient safety is our highest priority.

I hope the information provided is helpful, but if you are dissatisfied with the handling of your request, you have the right to ask for an internal review. Internal review requests should be submitted within two months of the date of this response; and can be addressed to this email address.

Yours sincerely,

FOI Team, Vigilance and Risk Management of Medicines Division