Learning Objectives for this Presentation:
By the end of the presentation, participants will be able to understand issues surrounding registry estimation of local immunization rates.


Background:
Immunization registries contain significant promise for their ability to determine immunization rates on a local basis and to find pockets of need. However three problems that have dogged registries in this effort are record scattering, adjusting for early childhood mobility, and incomplete reporting by providers.

Objectives:
To present an approach for determining immunization rates that addresses registry strengths and challenges.

Methods:
Demographic and immunization records were pulled from the ALERT registry for children born in 2002, for immunizations received from birth through age two. For children born in Oregon, a model of immunization factors was applied to produce estimated county level rates for the 4:3:1 series, based on compliance with observed periodicity of shot seeking by region. Factors in the model including migration, local mobility, non-shot seeking, record scattering, and partial reporting. For children in ALERT not born in Oregon, missing histories were imputed from Oregon births matched on region and observed immunizations. A metric is also applied to the model to adjust rates at the local, county and state level.

Results:
Overall immunization and demographic records were pulled for 56,392 children from ALERT. Through review against birth and demographic data, 3,905 records were merged, resulting in an ALERT population of 52,487 two year olds. Of 44,482 births to Oregon residents, 40,462 had a record in the 4:3:1 series. Another 8,005 children without an Oregon birth record had an immunization in the 4:3:1 series and an Oregon address in ALERT by age two. Overall the model adjusted raw rates upwards by more than 10% in all scenarios.

Conclusions:
Use of this model, with adjustments for mobility and partial record capture, will significantly increase calculated registry-based immunization rates on state, county and local levels.