Four sudden deaths in men aged between 50 and 75 years with underlying cardiac conditions were reported to public health authorities in Israel over a period of six days between 15 and 21 October 2006. The first three deaths came to the attention of the authorities because they had all occurred in the same locality and the patients were known to have received influenza vaccine (Vaxigrip) from the same vaccine lot in the days before death. Concerns were raised at a national level about the safety of the vaccine, and details of similar cases were requested by health authorities from elsewhere in Israel. The fourth death came to the attention of the authorities shortly after this. The time period between immunisation and death ranged from a few hours to six days. The clinical features in all four cases were consistent with death from cardiac disease. There was no history of sudden collapse directly after immunisation (anaphylaxis) and no indications of sepsis.

As a precaution, the Israeli Ministry of Health temporarily suspended the annual influenza vaccination programme on 22-23 October while they undertook a formal investigation. The initial conclusions were that, despite the geographical and/or temporal clustering of the cases, it was very unlikely that there was a causal link between the deaths and the influenza vaccinations. Consequently the vaccination programme in Israel resumed on 24 October.

European response
There was media reporting of the deaths and the possible link to the vaccine. On 23 October, the European Commission circulated an Early Warning Response System (https://webgate.cec.eu.int/ewrs/) message informing of the event and asking member states if a similar event has been reported at national level, and if they were undertaking or planning public health measures to respond to the event.

Authorities in one country temporarily withdrew the Vaxigrip vaccine, although the influenza vaccination programme continued. Using much of the information available to the Israeli authorities, the European Centre for Disease Prevention and Control (ECDC) undertook a risk assessment on 23 October [1], which also concluded that the evidence made a causal relationship very unlikely. The ECDC assessment noted that the patients and the population immunised were at significant risk of sudden cardiac death because of their age and/or other factors, and concluded that there was no reason to change national influenza vaccine programmes in EU member states [1,2]. Influenza vaccination programmes are very busy throughout Europe at this time of year, immunising the regular recommended risk groups of the elderly, those with underlying chronic illnesses and healthcare workers before the winter epidemics begin. The country where the vaccine was withdrawn has now reversed the decision [3,4].

Discussion
Most of the people who are offered influenza vaccination are, due to their age and/or illness status, at higher risk of sudden and unexpected death than people in younger and healthier population groups. The four men who died belonged to a group at high risk of natural sudden death due to their age and/or underlying illnesses. Deaths would be expected to occur in the population over 65 years at a rate of about 1 per 1000 every week (based on estimates for the Swedish population [5]. An even higher death rate would be expected in people over 65 years who have underlying cardiac disease. Therefore, of the around 140 000 elderly and at-risk people who reportedly have received influenza vaccine in Israel this autumn, it could be estimated that at least 140 might die through unrelated causes in the week following immunisation (20 per day). Therefore, these four deaths, although three were clustered geographically, do not necessarily represent an unexpected event.

The vaccine concerned (Vaxigrip) is a trivalent inactivated influenza vaccine produced by Sanofi Pasteur, France, as a sterile suspension with the adult dose containing 15 micrograms of haemagglutinin for the strains according to World Health Organization recommendations [3,4]. The vaccine also contains formaldehyde, Triton? X-100 and neomycin (and thiomersal as a preservative in the multidose preparation). This vaccine has been used since 1968 and is licensed for over 100 countries, where it is marketed under different brand names. Information from WHO indicates that approximately 80 million doses have been produced for the 2006-2007 northern hemisphere influenza season.

WHO is not aware of any other reports of serious events or lethal outcomes following immunisation with Vaxigrip since the start of immunisation in the northern hemisphere during the 2006-2007 influenza season which runs from autumn to spring each winter. WHO has stressed the importance of national authorities being vigilant for serious adverse events occurring within a short period following the administration of seasonal influenza vaccines in elderly people and individuals suffering from underlying chronic conditions. Deaths in chronically ill people that occur following influenza immunization, but which are unrelated to the vaccine, are also likely to be underreported.

The very few deaths that may be caused by influenza immunisation are usually related to anaphylactic shock, which occurs very shortly following the injection. The four deaths in Israel were not related to anaphylactic shock, and moreover, the risk of anaphylactic shock does not increase in cardiac patients. Death may occasionally occur when a patient develops sepsis due to contamination of a vaccine batch by infectious agents, but the clinical pictures in the four cases are not consistent with infection and these vaccines are reported to have been subject to normal quality control.

There is strong scientific evidence that influenza vaccination among cardiac patients is as safe as for the general population, while acute infections – including influenza – are associated with a transient increase in the risk of vascular events [6-10]. Therefore, these four deaths are very unlikely to be linked causally to the immunisations and routine annual influenza immunisation continues to be strongly recommended for the risk groups recognised in most EU countries [4].

Acknowledgements
This report was prepared using information supplied by Ministry of Health of Israel to the Communicable Disease Surveillance and Response and Vaccine-Preventable Diseases and Immunization Units at the WHO Regional Office for Europe.