The Impact of 2009 Pandemic Influenza on Racial/ethnic Disparities in Seasonal Influenza Vaccination

Background:The H1N1 pandemic influenza struck the US from April 2009 to April 2010, with its peak in October 2009. The delayed availability of H1N1 pandemic vaccines (distributed after December 2009) induced the relative shortage of seasonal influenza vaccines during the 2009-2010 season.  Because our previous study (analyzing 2001-2005 seasons) indicated that vaccine supply shortage was associated with increased racial/ethnic disparities in seasonal influenza vaccination, we hypothesized that these racial/ethnic disparities increased during the 2009-2010 season as well (hypothesis 1).  Additionally, our previous studies showed that an increased epidemic level was followed by an increased influenza vaccination rate with a two-week interval (during the same season) or a one-year interval (during the subsequent season).  Therefore, we also hypothesized a positive association between an influenza epidemic level (either pandemic or seasonal) and the influenza vaccination rate for the same and the subsequent seasons (hypothesis 2).  This association was hypothesized to differ across racial/ethnic groups (hypothesis 3).

Methods:Cross-sectional multivariable logistic regression analyses were performed across two consecutive seasons among the same cohort: (period 1) 2008-2009 and 2009-2010 seasons, (period 2) 2009-2010 and 2010-2011 seasons. Self-reported receipt of influenza vaccine was examined among non-Hispanic African-American (AA), non-Hispanic White (W), English-speaking Hispanic (EH) and Spanish-speaking Hispanic (SH) elderly (aged>64 years), from the Medicare Current Beneficiary Survey (MCBS) (un-weighted/weighted N =3,700~3,730/10.7-12.3 million). Key covariates included the number of weeks indicating the elevated influenza epidemic (either pandemic or seasonal) in each state and the interaction terms between these epidemic variables and the racial/ethnic groups. Additional covariates included individual-level demographics and health status.

Results: Our regression models showed statistically significant W-AA disparities (odds ratio (OR) = 0.41~0.48, p<.056) and W-SH disparities (OR=0.10~0.24, p<.01). Due to the relative vaccine supply shortage during the 2009-2010 pandemic season, the overall vaccination rate declined from 72.7% (2008-2009 season) to 70.4% (2009-2010 season) and then increased to 74.5% (2010-2011 season). Our regression model predicted that this overall vaccination rate during the 2009-2010 season could have increased by 2.7 percentage points without the pandemic influenza.  Hypothesis 1 was supported only for W-AA disparities, but not for W-SH disparities. This was partly explained by a positive association between vaccination and the ongoing pandemic epidemic level (OR=1.08~1.16, p<.05), found only among the SH group.  Because of this positive association, our regression model predicted that the vaccination rate among the SH group could have decreased from 55.1% to 20.1% during the 2009-2010 season and from 54.5% to 35.0% during the 2010-2011 season without the pandemic influenza.  Therefore, hypotheses 2 and 3 were supported.

Conclusions: Vaccine supply shortage was associated with increased racial/ethnic disparities, while improved supply was associated with reduced disparities in influenza vaccination rates. These associations may not hold when a high epidemic level (e.g., 2009 pandemic) motivated a racial/ethnic minority to demand a vaccination to a greater extent.  To avoid future widening of racial/ethnic health disparities, policy options during a shortage include preferential delivery of vaccines to safety-net providers serving AA and Hispanic populations who may have an incentive reinforced by an epidemic.