KEYWORDS:
child, immunization, vaccination coverage

BACKGROUND:
Sponsored by the Centers for Disease Control and Prevention, the National Immunization Survey (NIS) is used to monitor vaccination coverage at the local, state, and national levels. The NIS is a two phase survey. Phase one is a random-digit-dialing telephone survey that interviews parents/guardians of children aged 19 to 35 months about their child's vaccination history and request permission to contact the providers. In phase two, questionnaires are mailed to providers to ascertain the child's vaccination history. The coverage estimates reported from NIS are based on the provider data with adjustments for provider non-response.

OBJECTIVE(S):
To describe and compare the characteristics of children whose providers reported vaccination histories and children whose providers did not report vaccination histories.

METHOD(S):
We used the 2000 NIS data to classify children into three groups; 1) those with parent/guardian consent and provider data, 2) those with parent/guardian consent and no provider data, and 3) those whose parent/guardian did not give consent to contact providers. Using the chi-square test, we compared the demographic and socio-economic characteristics of these three groups.

RESULT(S):
The sample size for the 2000 NIS was 34,087, of which 29,028 (85.2%) gave consent to contact providers. Among children whose parent/guardian gave consent, providers failed to respond for 5,613 (19.3%) children. Differences by race/ethnicity, age of mother, educational status of mother, and shot card usage by the respondent during the interview were noted in the three groups.

CONCLUSIONS(S):
Our findings show differences among participants with and without provider data. These differences must be considered when analyzing data from the NIS and other surveys that rely on providers to confirm patient-reported receipt of services.

LEARNING OBJECTIVES:
1) Describe the National Immunization Survey,
2) Describe major differences between children with and without provider data, and
3) Describe the potential impact of these differences on coverage estimates generated from NIS.