Deaths following influenza vaccination--background mortality or causal connection?

In October 2006, four deaths occurred in Israel shortly after influenza immunization, resulting in a temporary halt to the vaccination campaign. After an epidemiologic investigation, the Ministry of Health concluded that these deaths were not related to the vaccine itself and the campaign resumed; however, vaccine uptake was markedly reduced. Estimates of true background mortality in this high-risk population would aid in public education and quell unnecessary concerns regarding vaccine safety. We used data from a large HMO to estimate mortality in influenza vaccine recipients aged 55 and over during four consecutive winters (2003, 2004, 2005 and 2006). Date of immunization was ascertained from patient treatment files, vital status through Israeli National Insurance Institute data. We calculated crude death rates within 7, 14 and 30 days of influenza immunization, and used a Cox Proportional Hazards Model to estimate the risk of death within 14 days of vaccination, adjusting for age and comorbid conditions (age over 75, history of diabetes or cardiovascular disease, status as homebound patient) in 2006. The death rate among influenza vaccine recipients ranged from 0.01 to 0.02% within 7 days and 0.09-0.10% at 30 days. Influenza immunization was associated with a decreased risk of death within 14 days after adjustment for comorbidities (Hazard ratio, 0.33, 95% CI, 0.18-0.61). Our findings support the assumption that influenza vaccination is not associated with increased risk of death in the short term.