County-level assessment of United States kindergarten vaccination rates for measles mumps rubella (MMR) for the 2014–2015 school year

Vaccine
Volume 35, Issue 47, Pages 6355-6468 (7 November 2017)
http://www.sciencedirect.com/science/journal/0264410X/35/47?sdc=1

County-level assessment of United States kindergarten vaccination rates for measles mumps rubella (MMR) for the 2014–2015 school year
Original Research Article
Pages 6444-6450
Sheryl A. Kluberg, Denise P. McGinnis, Yulin Hswen, Maimuna S. Majumder, Mauricio Santillana, John S. Brownstein
Abstract
United States kindergarten measles-mumps-rubella (MMR) vaccination rates are typically reported at the state level by the Centers for Disease Control and Prevention (CDC). The lack of local MMR data prevents identification of areas with low vaccination rates that would be vulnerable to the spread of disease. We collected county-level vaccination rates for the 2014–2015 school year with the objective of identifying these regions.
We requested county-level kindergarten vaccination data from state health departments, and mapped these data to visualize geographic patterns in achievement of the 95% MMR vaccination target. We aggregated the county-level data to the state level for comparison against CDC state estimates. We also analyzed the relationship of MMR vaccination level with county-level and state-level poverty (using U.S. census data), using both a national mixed model with state as a random effect, and individual linear regression models by state.
We received county vaccination data from 43 states. The median kindergarten MMR vaccination rate was 96.0% (IQR 89–98) across all counties, however, we estimated that 48.4% of the represented counties had vaccination rates below 95%. Our state estimates closely reflected CDC values. Nationally, every 10% increase in under-18 county poverty was associated with a 0.24% increase in MMR vaccination rates (95% CI: −0.07%; 0.54%), but the direction of this relationship varied by state.
We found that county data can reveal vaccination trends that are unobservable from state-level data, but we also discovered that the current availability of county-level data is inadequate. Our findings can be used by state health departments to identify target areas for vaccination programs.