|
Abstract:
|
The National Immunization Survey (NIS) is the primary tool for monitoring vaccination coverage for 19-35 month olds in the U.S. Immunization Information Systems (IIS) contain provider-verified, population-based immunization surveillance data, and can be also be used to monitor vaccination coverage. Differences in reported vaccination coverage estimates from the NIS and IIS are difficult to interpret. Differences might reflect bias in one or both of the data systems, or might reflect differences in methodologies for estimating vaccination coverage. We proposed simulating the NIS sample design and analytic methods within a convenience sample of five IIS sentinel sites (IIS-SS) that have high data quality and comparing vaccination coverage between the NIS and IIS-SS. The objective of this analysis is to develop an analytic approach that "replicates" or approximates the NIS methodology using IIS-SS data to more systematically distinguish between system biases and measurement differences for estimating vaccination coverage from IIS. We will simulate the NIS design one hundred times and produce point estimates and 95% interval estimates.
|
