What to know
The National Immunization Survey-Child (NIS-Child) uses random-digit dialing to survey households with children aged 19 to 35 months. The survey's goal is to gather information about immunization. The survey also collects information on breastfeeding. Learn about the response rates, methods, and effects of changes made in the NIS sampling method.
About the survey
CDC’s National Center for Immunizations and Respiratory Diseases (NCIRD) conducts the National Immunization Survey-Child (NIS-Child) each year. The NIS-Child uses random-digit dialing to survey households with children aged 19 to 35 months. The telephone survey asks questions about childhood immunization. In 2001 and 2002, a subsample of respondents was asked about breastfeeding. Since January 2003, all respondents with children aged 19 to 35 months have been asked the breastfeeding questions.
Response rates
The Council of American Survey and Research Organizations landline sample response rates for NIS-Child years 2001 to 2017 ranged from 51.9% to 76.1%. Response rates for the cellular telephone sample for NIS-Child years 2011 to 2022 ranged from 21.1% to 33.5%. You can find a more detailed description of the methods at the National Immunization Survey Website.
Effect of changes in NIS sampling method
A cellular telephone sample of respondents was included in the 2011 NIS. This sample was in addition to the landline telephone sampling frame used for all previous survey years. Using both cellular and landline telephone sampling is referred to as dual-frame sampling. CDC evaluated the impact on estimates of breastfeeding rates when the NIS added a cellular phone sample. The evaluation included comparing differences in breastfeeding estimates between landline only and dual-frame samples using data from the same birth cohort. Full 2014 NIS Report related to breastfeeding.
The NIS sampling frame was changed in 2018 to include only a cellular phone sample (referred to as single cell-frame sampling). To estimate the effect of this sampling frame change on breastfeeding rates, CDC examined data from the 2016 and 2017 NIS surveys. CDC looked at the breastfeeding rate differences from the dual-frame and single cell-frame for infants born in 2015. Full 2019 NIS Report.
Key findings for the change to a cell phone-only sample:
- Among infants born in 2015, the change from dual-frame to single cell-frame sampling of respondents had a minimal effect on national estimates of all six breastfeeding indicators examined. These indicators included ever breastfed, breastfed at 6 and 12 months, exclusive breastfeeding through 3 and 6 months, and supplementing with formula before 2 days among breastfed infants. National prevalence estimates differed by 0.2 percentage points or less.
- The sampling change had a slightly larger effect on breastfeeding estimates at the state level. Some states showed higher rates from the dual-frame compared to the single cell-frame estimates. Some states showed lower rates. Across the six indicators examined, the difference magnitude ranged from 0.0 to 5.4 percentage points. It was less than 2.0 percentage points for about 90% of the state estimates.
- CDC cautions against comparing current single cell-frame estimates of breastfeeding rates to previous dual-frame estimates at the state level.