National Immunization Survey (NIS)

The vaccination coverage rate tracker for national, state, and selected local areas.

• One table with one row per sampled toddler.

• A complex sample survey designed to generalize to children aged 19-35 months in the United States.

• Released annually since 1995, plus an adolescent (13-17 years) sample since 2008.

• Administered by the Centers for Disease Control and Prevention.

---

Recommended Reading

Four Example Strengths & Limitations:

✔️ Both parents and medical providers interviewed

✔️ Detailed health insurance questions

❌ Low household response rates and only half-completed provider data during 2019-2023

❌ Although national estimates are precise, estimates for state and local areas should be interpreted with caution because their sample sizes are smaller, confidence intervals wider than national estimates

Three Example Findings:

1. In 2014 in the general population in Ohio, vaccination coverage with at least one dose or at least two doses of MMR among young children and adolescents was 96% and 88%, respectively.

2. Completion of a 7-vaccine series by 19 months of age increased from 52% in 2011 to 59% in 2021.

3. HPV vaccination initiation by age 13 rose from 27% to 70% among those born in 1999 versus 2009.

Two Methodology Documents:

About NIS

National Immunization Survey-Child: A User’s Guide for the 2023 Public-Use Data File

One Haiku:

# i hear babies cry
# protesting lungs of iron
# a wonderful world

---

Download, Import, Preparation

Download the 2023 fixed-width file:

dat_tf <- tempfile()

dat_url <- "https://www.cdc.gov/nis/media/files/2024/11/NISPUF23.DAT"

download.file( dat_url , dat_tf , mode = 'wb' )

Edit then execute the import script provided by the CDC:

library(Hmisc)

r_tf <- tempfile()

r_script_url <- "https://ftp.cdc.gov/pub/Vaccines_NIS/NISPUF23.R"

r_input_lines <- readLines( r_script_url )

# do not let the script do the save()
r_input_lines <- gsub( "^save\\(" , "# save(" , r_input_lines )

# redirect the path to the flat file to the local save location of `dat_tf`
r_input_lines <- gsub( '\\"path\\-to\\-file\\/(.*)\\.DAT\\"' , "dat_tf" , r_input_lines )

# save the edited script locally
writeLines( r_input_lines , r_tf )

# run the edited script
source( r_tf , echo = TRUE )

# rename the resultant data.frame object
nis_df <- NISPUF23

names( nis_df ) <- tolower( names( nis_df ) )

nis_df[ , 'one' ] <- 1

Save Locally  

Save the object at any point:

# nis_fn <- file.path( path.expand( "~" ) , "NIS" , "this_file.rds" )
# saveRDS( nis_df , file = nis_fn , compress = FALSE )

Load the same object:

# nis_df <- readRDS( nis_fn )

Survey Design Definition

Construct a complex sample survey design:

library(survey)

options( survey.lonely.psu = "adjust" )

nis_design <- 
    svydesign(
        id = ~ seqnumhh , 
        strata = ~ stratum , 
        weights = ~ provwt_c , 
        data = subset( nis_df , provwt_c > 0 ) 
    )

Variable Recoding

Add new columns to the data set:

nis_design <- 
    
    update( 
        
        nis_design , 
        
        first_fed_formula =
            ifelse( bf_formr20 %in% 888 , NA , bf_formr20 ) ,
        
        dtap_3p =

            as.numeric(

                ( p_numdah >= 3 ) |
                ( p_numdhi >= 3 ) |
                ( p_numdih >= 3 ) |
                ( p_numdta >= 3 ) |
                ( p_numdtp >= 3 )

            ) ,
        
        dtap_4p =

            as.numeric(

                ( p_numdah >= 4 ) |
                ( p_numdhi >= 4 ) |
                ( p_numdih >= 4 ) |
                ( p_numdta >= 4 ) |
                ( p_numdtp >= 4 )

            )
            
    )

---

Analysis Examples with the survey library  

Unweighted Counts

Count the unweighted number of records in the survey sample, overall and by groups:

sum( weights( nis_design , "sampling" ) != 0 )

svyby( ~ one , ~ state , nis_design , unwtd.count )

Weighted Counts

Count the weighted size of the generalizable population, overall and by groups:

svytotal( ~ one , nis_design )

svyby( ~ one , ~ state , nis_design , svytotal )

Descriptive Statistics

Calculate the mean (average) of a linear variable, overall and by groups:

svymean( ~ first_fed_formula , nis_design , na.rm = TRUE )

svyby( ~ first_fed_formula , ~ state , nis_design , svymean , na.rm = TRUE )

Calculate the distribution of a categorical variable, overall and by groups:

svymean( ~ sex , nis_design , na.rm = TRUE )

svyby( ~ sex , ~ state , nis_design , svymean , na.rm = TRUE )

Calculate the sum of a linear variable, overall and by groups:

svytotal( ~ first_fed_formula , nis_design , na.rm = TRUE )

svyby( ~ first_fed_formula , ~ state , nis_design , svytotal , na.rm = TRUE )

Calculate the weighted sum of a categorical variable, overall and by groups:

svytotal( ~ sex , nis_design , na.rm = TRUE )

svyby( ~ sex , ~ state , nis_design , svytotal , na.rm = TRUE )

Calculate the median (50th percentile) of a linear variable, overall and by groups:

svyquantile( ~ first_fed_formula , nis_design , 0.5 , na.rm = TRUE )

svyby( 
    ~ first_fed_formula , 
    ~ state , 
    nis_design , 
    svyquantile , 
    0.5 ,
    ci = TRUE , na.rm = TRUE
)

Estimate a ratio:

svyratio( 
    numerator = ~ bf_exclr06 , 
    denominator = ~ bf_endr06 , 
    nis_design ,
    na.rm = TRUE
)

Subsetting

Restrict the survey design to toddlers up to date on polio shots:

sub_nis_design <- subset( nis_design , p_utdpol == 1 )

Calculate the mean (average) of this subset:

svymean( ~ first_fed_formula , sub_nis_design , na.rm = TRUE )

Measures of Uncertainty

Extract the coefficient, standard error, confidence interval, and coefficient of variation from any descriptive statistics function result, overall and by groups:

this_result <- svymean( ~ first_fed_formula , nis_design , na.rm = TRUE )

coef( this_result )
SE( this_result )
confint( this_result )
cv( this_result )

grouped_result <-
    svyby( 
        ~ first_fed_formula , 
        ~ state , 
        nis_design , 
        svymean ,
        na.rm = TRUE 
    )
    
coef( grouped_result )
SE( grouped_result )
confint( grouped_result )
cv( grouped_result )

Calculate the degrees of freedom of any survey design object:

degf( nis_design )

Calculate the complex sample survey-adjusted variance of any statistic:

svyvar( ~ first_fed_formula , nis_design , na.rm = TRUE )

Include the complex sample design effect in the result for a specific statistic:

# SRS without replacement
svymean( ~ first_fed_formula , nis_design , na.rm = TRUE , deff = TRUE )

# SRS with replacement
svymean( ~ first_fed_formula , nis_design , na.rm = TRUE , deff = "replace" )

Compute confidence intervals for proportions using methods that may be more accurate near 0 and 1. See ?svyciprop for alternatives:

svyciprop( ~ dtap_3p , nis_design ,
    method = "likelihood" )

Regression Models and Tests of Association

Perform a design-based t-test:

svyttest( first_fed_formula ~ dtap_3p , nis_design )

Perform a chi-squared test of association for survey data:

svychisq( 
    ~ dtap_3p + sex , 
    nis_design 
)

Perform a survey-weighted generalized linear model:

glm_result <- 
    svyglm( 
        first_fed_formula ~ dtap_3p + sex , 
        nis_design 
    )

summary( glm_result )

---

Replication Example

This example matches the statistics and standard errors from Data User’s Guide Table 4:

results <-
    svyby( 
        ~ p_utd431h314_rout_s , 
        ~ raceethk , 
        nis_design , 
        svymean
    )

coefficients <- results[ , "p_utd431h314_rout_sUTD" , drop = FALSE ]

standard_errors <- results[ , "se.p_utd431h314_rout_sUTD" , drop = FALSE ]

stopifnot( round( coefficients[ "HISPANIC" , ] , 3 ) == .674 )
stopifnot( round( coefficients[ "NON-HISPANIC WHITE ONLY" , ] , 3 ) == .716 )
stopifnot( round( coefficients[ "NON-HISPANIC BLACK ONLY" , ] , 3 ) == .666 )
stopifnot( round( standard_errors[ "HISPANIC" , ] , 3 ) == .017 )
stopifnot( round( standard_errors[ "NON-HISPANIC WHITE ONLY" , ] , 3 ) == .008 )
stopifnot( round( standard_errors[ "NON-HISPANIC BLACK ONLY" , ] , 3 ) == .023 )

---

Analysis Examples with srvyr  

The R srvyr library calculates summary statistics from survey data, such as the mean, total or quantile using dplyr-like syntax. srvyr allows for the use of many verbs, such as summarize, group_by, and mutate, the convenience of pipe-able functions, the tidyverse style of non-standard evaluation and more consistent return types than the survey package. This vignette details the available features. As a starting point for NIS users, this code replicates previously-presented examples:

library(srvyr)
nis_srvyr_design <- as_survey( nis_design )

Calculate the mean (average) of a linear variable, overall and by groups:

nis_srvyr_design %>%
    summarize( mean = survey_mean( first_fed_formula , na.rm = TRUE ) )

nis_srvyr_design %>%
    group_by( state ) %>%
    summarize( mean = survey_mean( first_fed_formula , na.rm = TRUE ) )