Does English fluency affect ethnic disparities in influenza vaccination among the Medicare elderly during seasons with vaccine supply problems?

Research Objective: The persistent racial/ethnic disparities in influenza vaccination rates have been reported among the Medicare elderly. White-Hispanic disparities have been suggested to be largely explained by English fluency among the elderly and middle-aged populations in past studies. We tested hypotheses that worse vaccine supply was associated with increased White-Hispanic disparity during two consecutive seasons with different vaccine supply levels, that the magnitude of such disparity change was greater among the Hispanic with limited English fluency.

Methods: Cross-sectional multivariate logistic regression analyses were performed for four seasons: (period 1) 2000-2001 and 2001-2002 seasons; and (period 2) 2003-2004 and 2004-2005 seasons. We examined self-reported receipt of influenza vaccine over consecutive years among a same cohort of community dwelling non-Hispanic African American, non-Hispanic White (W), and Hispanic elderly aged 65 or older, enrolled in the Medicare Current Beneficiary Survey (MCBS) (un-weighted N = 2,504 and 2,411, weighted N = 11.9 million and 12.6 million for periods 1 and 2, respectively). Analyses included two variables distinguishing English-speaking Hispanic (EH) and Spanish-speaking Hispanic (SH) beneficiaries who used Spanish in at least one of three MCBS interviews per year: around 50% in each period. Vaccination rates were estimated with adjustment for other factors such as health status and socioeconomic characteristics

Principal Findings: During period 1, when vaccine supply improved, adjusted W-EH disparity decreased by 6.6% (from 11.5% to 4.9%; p<.001), and adjusted W-SH disparity decreased 10.9% (from 32.6% to 21.7%; p<.001; significantly greater than 6.6% of W-EH disparity), as hypothesized. During period 2, when vaccine supply declined, adjusted W-EH disparity increased by 5.6% (from 4.1% to 9.7%; p<.001), as hypothesized. However, adjusted W-SH disparity “decreased” by 6.2% (from 35.8% to 29.5%; p<.001), against our hypothesis. The latter exceptional case could have been a floor-effect, i.e., because the adjusted vaccination rate among SH was as low as 41.2% (that is 35% lower than W) in the first season in period 2, this rate declined only by 0.5% among SH (while 7% among W) when vaccine supply decreased in the subsequent season. Namely, the adjusted vaccination rate among SH did not increase but declined during period 2, although the W-EH disparity decreased.

Conclusion: Improved vaccine supply was associated with reduced ethnic disparities in influenza vaccination rates, while worse supply appeared to be associated with increased disparities. Accounting for English fluency is important to accurately measure ethnic disparities and identify subpopulations at a higher risk of missing vaccination.