Near real time vaccine safety monitoring for COVID-19 vaccines

Date of Approval: 
2020-10-19 00:00:00
Lay Summary: 
At its first use as part of a vaccine programme, the safety of any COVID-19 vaccine will already have been well explored in clinical trials conducted in several thousand people. Therefore, we will have a good understanding of the risk of any more common side-effects within the trial population which, if the vaccine is authorised for introduction, will be mild and outweighed by the expected benefits. However, following introduction of a vaccine, and use in larger populations, suggestions of very rare but more serious potential side-effects can start to emerge. These concerns are often unfounded, whereby events are linked with the vaccine due to their timing but are in fact not caused by the vaccine. However, they can also on rare occasions be due to a true side-effect. Therefore, once a COVID-19 vaccine is introduced it is important that the risk of rare events is actively monitored in order to both provide evidence on vaccine safety, mitigating the impact of unfounded scares which can discourage people from being vaccinated and provide them with reassurance, as well as to rapidly detect any true safety concerns. This monitoring study will include weekly analyses to generate signals of potential rare risks where we see higher rates of an event occurring in patients than is expected given the natural occurrence of such events in unvaccinated people. Any concerns will then be further explored to ascertain if they are related to the vaccine and changes made to the vaccination programme to ensure safety if necessary.
Technical Summary: 
Sequential tests have been applied within longitudinal patient records for the timely detection of safety signals related to vaccines in the US. The purpose of this approach is to allow repeated interrogation of the data, starting immediately after the vaccine is introduced and adjusting for multiple analyses, in order to rapidly detect any increases in risk for pre-identified adverse events of special interest (AESI) potentially associated with vaccination. The feasibility of conducting these analyses has been previously explored in the CPRD. In brief, this active monitoring approach uses repeated application of the Maximised Sequential Probability Ratio Test (MaxSPRT) at pre-defined time points to assess the rate of pre-identified AESI compared to a control. A signal is triggered if the relative risk crosses a pre-defined threshold. AESIs are prospectively identified based on clinical trial data, experience with other vaccines, and knowledge on events that occur naturally in the target population and which may be reported in temporal association with vaccination. Several potential COVID-19 vaccines are in development and once one or several are available deployment across a large population could be very rapid. Following introduction, these sequential methods will be implemented for a full range of AESI. This will allow the size of the evidence base on vaccine safety to grow from first use and in line with the exposed population. These methods will be implemented as part of the UK regulatory portfolio of COVID-19 vaccine vigilance activities. The totality of the data will be used to support communications on the safety of the vaccine to mitigate unfounded safety scares, which can be triggered by small numbers of serious unexpected events, and to rapidly detect any true rare risks should they occur informing the need for changes to the vaccination programme to ensure safety if necessary.
Health Outcomes to be Measured: 
Incidence of AESI. The pre-defined list of AESI is: Sudden death (all ages, including SIDS in infants); Guillain-Barré syndrome, and other peripheral and polyneuropathies; Multiple sclerosis, and other demyelinating disorders; Optic neuritis; Encephalitis (including acute disseminated encephalomyelitis); Myasthenia gravis; Bell’s palsy; Seizure disorders (including febrile); Myocardial infarction; Myo/pericarditis; Stroke and other cerebrovascular events; Idiopathic thrombocytopenic purpura, autoimmune thrombocytopenia; Rheumatoid arthritis, polyarthritis; Autoimmune thyroiditis; Kawasaki syndrome (Paediatric multisystem inflammatory syndrome temporally associated with COVID-19) AESI code lists are presented in Annex 1. Note, that this list may be added to depending on emerging safety data from COVID-19 vaccine phase III trials or other source post-licensure. Absolute exposure to a COVID-19 vaccine will also be an outcome of interest.
Application Number: 

Katherine Donegan - Chief Investigator - MHRA
Katherine Donegan - Corresponding Applicant - MHRA
Jemma Walker - Collaborator - London School of Hygiene & Tropical Medicine ( LSHTM )
Jenny Wong - Collaborator - MHRA
Nick Andrews - Collaborator - Public Health England
Philip Bryan - Collaborator - MHRA

HES Admitted Patient Care;ONS Death Registration Data