Evaluating Washington State\'s immunization information system as a research tool.


OBJECTIVE: Immunization information systems (IISs) are powerful public health tools for vaccination activities. To date- however- their use for public health research has been limited- in part as a result of insufficient understanding on accuracy and quality of IIS data. We evaluated the completeness and accuracy of Washington State IIS (WAIIS) data- with particular attention to data elements of research interest. METHODS: We analyzed all WAIIS records on all children born between 2006 and 2010 with at least 1 vaccination recorded in WAIIS between 2006 and 2010. We assessed all variables for completeness and tested selected variables for internal validity. To assess external validity- we matched WAIIS data to records from Group Health- a large integrated health care organization in Washington State. On these children- we compared vaccination data in WAIIS with vaccination data from Group Health's immunization registry. RESULTS: The WAIIS data included 486-265 children and 8-670-234 unique vaccinations. Variables required by WAIIS (such as date of vaccination) were highly complete- but optional variables were often missing. For example- most records were missing data on route (80.7%) and anatomic site (81.7%) of vaccination. WAIIS data- when complete- were highly accurate relative to the Group Health immunization registry- with 96% to 99% agreement between fields such as vaccination code and anatomic site. CONCLUSIONS: Required data elements in WAIIS are highly complete and have both internal and external validity- suggesting that these variables are useful for research. Research requiring nonrequired variables should use additional validity checks before proceeding.