Relationship of influenza vaccine match and use rate to medically attended acute respiratory illnesses in older residents of Maryland. November 13, 2012

The effectiveness of seasonal influenza vaccine may be influenced by mismatches to circulating influenza viruses. The relationship of these vaccine mismatches to the occurrence of medically attended acute respiratory illnesses (MAARI)-related emergency department (ED) visits and hospitalizations for all Maryland residents aged 50 years or older was examined for seven years (2001-2008). Also, relationships of individual circulating influenza types or subtypes to these MAARI-encounters were investigated. Anonymous, retrospective discharge data from all Maryland hospitals were utilized to determine daily numbers of MAARI-related ED visits and hospitalizations. Rate ratios (RR) of these MAARI-related encounters were then determined between intense influenza outbreak periods and non-influenza season time periods for each year using a Poisson regression model. Center for Disease Control end of season data reports were used to estimate each season's 'match' of each trivalent influenza vaccine component to subsequent circulating influenza viruses in the United States. The overall vaccine match proportion for the three vaccine viruses was then multiplied by reported vaccination rates of Maryland residents aged 50 years or older to determine an 'estimated influenza vaccine impact factor' (EIVIF) for each year. Correlation coefficients (CC) were then calculated comparing RR of MAARI-encounters to the corresponding EIVIF data. Finally, yearly RR of MAARI-encounters were compared to corresponding rates of individual circulating influenza types or subtypes circulating in the Maryland area. Consistent trends were observed whereby increased EIVIF proportions were correlated with decreased RR of MAARI-related ED visits and hospitalizations. This correlation reached statistical significance when comparing EIVIF to RR of MAARI-related ED visits (CC=-0.77, P=0.04). Also, the proportion of A/H3N2 viruses circulating each season was significantly positively correlated to that season's RR of MAARI-related ED visits (CC=0.89) and hospitalizations (CC=0.92); both P<0.01. Our data suggest increased protective effects of influenza immunization on reducing MAARI-related ED visits and hospitalizations in older individuals when the vaccine is well matched to the circulating strains. A/H3N2 viruses were clearly associated with more frequent MAARI-related ED visits and hospitalizations than A/H1N1 and B viruses.