32-nhanes.Rmd

# National Health and Nutrition Examination Survey (NHANES) {-}

[![Build Status](https://travis-ci.org/asdfree/nhanes.svg?branch=master)](https://travis-ci.org/asdfree/nhanes) [![Build status](https://ci.appveyor.com/api/projects/status/github/asdfree/nhanes?svg=TRUE)](https://ci.appveyor.com/project/ajdamico/nhanes)

The National Health and Nutrition Examination Survey (NHANES) is this fascinating survey where doctors and dentists accompany survey interviewers in a little mobile medical center that drives around the country. While the survey methodologists interview people, the medical professionals administer laboratory tests and conduct a thorough physical examination. The blood work and medical exam allow researchers to answer tough questions like, "how many people have diabetes but don't know they have diabetes?"

* Many tables containing information gathered from the various examinations, generally with one row per individual respondent.

* A complex sample survey designed to generalize to the civilian non-institutionalized population of the United States.

* Released biennially since 1999-2000.

* Administered by the [Centers for Disease Control and Prevention](http://www.cdc.gov/).

## Simplified Download and Importation {-}

The R `lodown` package easily downloads and imports all available NHANES microdata by simply specifying `"nhanes"` with an `output_dir =` parameter in the `lodown()` function. Depending on your internet connection and computer processing speed, you might prefer to run this step overnight.

```{r eval = FALSE }
library(lodown)
lodown( "nhanes" , output_dir = file.path( path.expand( "~" ) , "NHANES" ) )
```

`lodown` also provides a catalog of available microdata extracts with the `get_catalog()` function. After requesting the NHANES catalog, you could pass a subsetted catalog through the `lodown()` function in order to download and import specific extracts (rather than all available extracts).

```{r eval = FALSE , results = "hide" }
library(lodown)
# examine all available NHANES microdata files
nhanes_cat <-
	get_catalog( "nhanes" ,
		output_dir = file.path( path.expand( "~" ) , "NHANES" ) )

# 2015-2016 only
nhanes_cat <- subset( nhanes_cat , years == "2015-2016" )
# download the microdata to your local computer
nhanes_cat <- lodown( "nhanes" , nhanes_cat )
```

## Analysis Examples with the `survey` library \ {-}

Construct a complex sample survey design:

```{r eval = FALSE }

```

```{r eval = FALSE }
options( survey.lonely.psu = "adjust" )

library(survey)

nhanes_demo_df <- 
	readRDS( file.path( path.expand( "~" ) , "NHANES" , "2015-2016/demo_i.rds" ) )

nhanes_tchol_df <- 
	readRDS( file.path( path.expand( "~" ) , "NHANES" , "2015-2016/tchol_i.rds" ) )

nhanes_df <- merge( nhanes_demo_df , nhanes_tchol_df , all = TRUE )

stopifnot( nrow( nhanes_df ) == nrow( nhanes_demo_df ) )

# keep only individuals who took the "mobile examination center" component
nhanes_df <- subset( nhanes_df , ridstatr %in% 2 )

nhanes_design <- 
	svydesign(
		id = ~sdmvpsu , 
		strata = ~sdmvstra ,
		nest = TRUE ,
		weights = ~wtmec2yr ,
		data = nhanes_df
	)
```

### Variable Recoding {-}

Add new columns to the data set:
```{r eval = FALSE }
nhanes_design <- 
	update( 
		nhanes_design , 
		
		one = 1 ,
		
		pregnant_at_interview = 
			ifelse( ridexprg %in% 1:2 , as.numeric( ridexprg == 1 ) , NA ) ,
		
		race_ethnicity = 
			factor( 
				c( 3 , 3 , 1 , 2 , 4 )[ ridreth1 ] ,
				levels = 1:4 , 
				labels = 
					c( 'non-hispanic white' , 'non-hispanic black' , 
						'hispanic' , 'other' )
			) ,
		
		age_category =
			factor(
				findInterval( ridageyr , c( 20 , 40 , 60 ) ) ,
				labels = c( "0-19" , "20-39" , "40-59" , "60+" )
			)
	)
```

### Unweighted Counts {-}

Count the unweighted number of records in the survey sample, overall and by groups:
```{r eval = FALSE , results = "hide" }
sum( weights( nhanes_design , "sampling" ) != 0 )

svyby( ~ one , ~ race_ethnicity , nhanes_design , unwtd.count )
```

### Weighted Counts {-}
Count the weighted size of the generalizable population, overall and by groups:
```{r eval = FALSE , results = "hide" }
svytotal( ~ one , nhanes_design )

svyby( ~ one , ~ race_ethnicity , nhanes_design , svytotal )
```

### Descriptive Statistics {-}

Calculate the mean (average) of a linear variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
svymean( ~ lbxtc , nhanes_design , na.rm = TRUE )

svyby( ~ lbxtc , ~ race_ethnicity , nhanes_design , svymean , na.rm = TRUE )
```

Calculate the distribution of a categorical variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
svymean( ~ riagendr , nhanes_design )

svyby( ~ riagendr , ~ race_ethnicity , nhanes_design , svymean )
```

Calculate the sum of a linear variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
svytotal( ~ lbxtc , nhanes_design , na.rm = TRUE )

svyby( ~ lbxtc , ~ race_ethnicity , nhanes_design , svytotal , na.rm = TRUE )
```

Calculate the weighted sum of a categorical variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
svytotal( ~ riagendr , nhanes_design )

svyby( ~ riagendr , ~ race_ethnicity , nhanes_design , svytotal )
```

Calculate the median (50th percentile) of a linear variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
svyquantile( ~ lbxtc , nhanes_design , 0.5 , na.rm = TRUE )

svyby( 
	~ lbxtc , 
	~ race_ethnicity , 
	nhanes_design , 
	svyquantile , 
	0.5 ,
	ci = TRUE ,
	keep.var = TRUE ,
	na.rm = TRUE
)
```

Estimate a ratio:
```{r eval = FALSE , results = "hide" }
svyratio( 
	numerator = ~ lbxtc , 
	denominator = ~ ridageyr , 
	nhanes_design ,
	na.rm = TRUE
)
```

### Subsetting {-}

Restrict the survey design to respondents aged 60 or older:
```{r eval = FALSE , results = "hide" }
sub_nhanes_design <- subset( nhanes_design , age_category == "60+" )
```
Calculate the mean (average) of this subset:
```{r eval = FALSE , results = "hide" }
svymean( ~ lbxtc , sub_nhanes_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:
```{r eval = FALSE , results = "hide" }
this_result <- svymean( ~ lbxtc , nhanes_design , na.rm = TRUE )

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

grouped_result <-
	svyby( 
		~ lbxtc , 
		~ race_ethnicity , 
		nhanes_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:
```{r eval = FALSE , results = "hide" }
degf( nhanes_design )
```

Calculate the complex sample survey-adjusted variance of any statistic:
```{r eval = FALSE , results = "hide" }
svyvar( ~ lbxtc , nhanes_design , na.rm = TRUE )
```

Include the complex sample design effect in the result for a specific statistic:
```{r eval = FALSE , results = "hide" }
# SRS without replacement
svymean( ~ lbxtc , nhanes_design , na.rm = TRUE , deff = TRUE )

# SRS with replacement
svymean( ~ lbxtc , nhanes_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:
```{r eval = FALSE , results = "hide" }
svyciprop( ~ pregnant_at_interview , nhanes_design ,
	method = "likelihood" , na.rm = TRUE )
```

### Regression Models and Tests of Association {-}

Perform a design-based t-test:
```{r eval = FALSE , results = "hide" }
svyttest( lbxtc ~ pregnant_at_interview , nhanes_design )
```

Perform a chi-squared test of association for survey data:
```{r eval = FALSE , results = "hide" }
svychisq( 
	~ pregnant_at_interview + riagendr , 
	nhanes_design 
)
```

Perform a survey-weighted generalized linear model:
```{r eval = FALSE , results = "hide" }
glm_result <- 
	svyglm( 
		lbxtc ~ pregnant_at_interview + riagendr , 
		nhanes_design 
	)

summary( glm_result )
```

## Analysis Examples with `srvyr` \ {-}

The R `srvyr` library calculates summary statistics from survey data, such as the mean, total or quantile using [dplyr](https://github.com/tidyverse/dplyr/)-like syntax. [srvyr](https://github.com/gergness/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](https://cran.r-project.org/web/packages/srvyr/vignettes/srvyr-vs-survey.html) details the available features. As a starting point for NHANES users, this code replicates previously-presented examples:

```{r eval = FALSE , results = "hide" }
library(srvyr)
nhanes_srvyr_design <- as_survey( nhanes_design )
```
Calculate the mean (average) of a linear variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
nhanes_srvyr_design %>%
	summarize( mean = survey_mean( lbxtc , na.rm = TRUE ) )

nhanes_srvyr_design %>%
	group_by( race_ethnicity ) %>%
	summarize( mean = survey_mean( lbxtc , na.rm = TRUE ) )
```

---

## Replication Example {-}

```{r eval = FALSE , results = "hide" }

```