nssrn.Rmd

# National Sample Survey of Registered Nurses (NSSRN) {-}

[![License: GPL v3](https://img.shields.io/badge/License-GPLv3-blue.svg)](https://www.gnu.org/licenses/gpl-3.0) <a href="https://github.com/asdfree/nssrn/actions"><img src="https://github.com/asdfree/nssrn/actions/workflows/r.yml/badge.svg" alt="Github Actions Badge"></a> 

The employment, education, and demographics of the RN, NP, and APRN workforce in the United States.

* One table with one row per surveyed registered nurse (RN) or nurse practitioner (NP).

* A complex sample designed to generalize to RNs and NPs at both state and national levels.

* Released every four to ten years since 1977.

* Administered by the [Health Services and Resources Administration](http://www.hrsa.gov/), in partnership with [Census](https://www.census.gov/).

---

## Recommended Reading {-}

Four Example Strengths & Limitations:

✔️ [Sampling frame drawn directly from state licensure bodies](https://data.hrsa.gov/DataDownload/NSSRN/GeneralPUF04/rnsurvey2004.pdf#page=140)

✔️ [Workforce-specific questionnaire](https://bhw.hrsa.gov/sites/default/files/bureau-health-workforce/data-research/nssrn-survey-form.pdf#page=14)

❌ [Small sample sizes of public health registered nurses](https://doi.org/10.1177/00333549231151877)

❌ [Cross-sectional design does not track the same nurse's career over time](https://data.hrsa.gov/Content/Documents/topics/About%20the%20NSSRN%20Survey%20and%20Data%2001092020.pdf)

<br>

Three Example Findings:

1. [Over 1992 to 2008, APRNs in states with more restrictive scope of practice laws worked fewer hours](http://doi.org/10.3386/w26896).

2. [Among US nurses leaving current employment in 2018, 31.5% reported leaving due to burnout](http://doi.org/10.1001/jamanetworkopen.2020.36469).

3. [Almost 2 million RNs used federally assisted student loans to finance a nursing degree](https://bhw.hrsa.gov/sites/default/files/bureau-health-workforce/data-research/nssrn-education-training-report.pdf).

<br>

Two Methodology Documents:

> [2022 NSSRN Methodology Report](https://data.hrsa.gov/DataDownload/NSSRN/GeneralPUF22/Documentation.zip)

> [Frequently Asked Questions](https://www.census.gov/programs-surveys/nssrn/about/faq.html)

<br>

One Haiku:

```{r}
# florence nightingale
# founder of modern nursing
# a statistician
```

---

## Download, Import, Preparation {-}

Download and import the state file:
```{r eval = FALSE , results = "hide" }
library(haven)

nssrn_tf <- tempfile()

nssrn_url <-
	"https://data.hrsa.gov/DataDownload/NSSRN/GeneralPUF22/2022_NSSRN_PUF_Stata_Package.zip"
	
download.file( nssrn_url , nssrn_tf , mode = 'wb' )

nssrn_files <- unzip( nssrn_tf , exdir = tempdir() )

nssrn_dta <- grep( "\\.dta$" , nssrn_files , ignore.case = TRUE , value = TRUE )

nssrn_tbl <- read_dta( nssrn_dta )

nssrn_df <- data.frame( nssrn_tbl )

names( nssrn_df ) <- tolower( names( nssrn_df ) )

nssrn_df[ , 'one' ] <- 1
```

### Save Locally \ {-}

Save the object at any point:

```{r eval = FALSE , results = "hide" }
# nssrn_fn <- file.path( path.expand( "~" ) , "NSSRN" , "this_file.rds" )
# saveRDS( nssrn_df , file = nssrn_fn , compress = FALSE )
```

Load the same object:

```{r eval = FALSE , results = "hide" }
# nssrn_df <- readRDS( nssrn_fn )
```

### Survey Design Definition {-}
Construct a complex sample survey design:

```{r eval = FALSE , results = "hide" }
library(survey)

nssrn_design <- 
	svrepdesign(
		weight = ~rkrnwgta ,
		repweights = 'rkrnwgta[0-9]+' ,
		scale = 4 / 80 ,
		rscales = rep( 1 , 80 ) ,
		mse = TRUE ,
		type = 'JK1' ,
		data = nssrn_df
	)
```

### Variable Recoding {-}

Add new columns to the data set:
```{r eval = FALSE , results = "hide" }
nssrn_design <- 
	update( 
		nssrn_design , 
		
		# all advanced practice registered nurses
		# (including nurse practitioners)
		all_aprn = as.numeric( ed_lcrn == 2 ) ,
		
		age_group =
			factor(
				findInterval( age_gp_puf , c( 0 , 3 , 5 , 7 , 9 ) ) ,
				levels = 1:5 ,
				labels = 
					c( 
						'34 or younger' ,
						'35 to 44' ,
						'45 to 54' ,
						'55 to 64' ,
						'65 or older'
					)
			) ,
			
		primary_position_state =
			factor(
				as.numeric( pn_loc_code_puf ) ,
				levels = 
					c(1L, 2L, 4L, 5L, 6L, 8L, 9L, 10L, 
					11L, 12L, 13L, 15L, 16L, 17L, 18L, 
					19L, 20L, 21L, 22L, 23L, 24L, 25L, 
					26L, 27L, 28L, 29L, 30L, 31L, 32L, 
					33L, 34L, 35L, 36L, 37L, 38L, 39L, 
					40L, 41L, 42L, 44L, 45L, 46L, 47L, 
					48L, 49L, 50L, 51L, 53L, 54L, 55L, 
					56L, 72L,
					# note collapsed geographies from codebook
					500L, 800L) ,
				labels =
					c("Alabama", "Alaska", "Arizona", "Arkansas", "California", 
					"Colorado", "Connecticut", "Delaware", "District of Columbia", 
					"Florida", "Georgia", "Hawaii", "Idaho", "Illinois", "Indiana", 
					"Iowa", "Kansas", "Kentucky", "Louisiana", "Maine", "Maryland", 
					"Massachusetts", "Michigan", "Minnesota", "Mississippi", "Missouri", 
					"Montana", "Nebraska", "Nevada", "New Hampshire", "New Jersey", 
					"New Mexico", "New York", "North Carolina", "North Dakota", "Ohio", 
					"Oklahoma", "Oregon", "Pennsylvania", "Rhode Island", "South Carolina", 
					"South Dakota", "Tennessee", "Texas", "Utah", "Vermont", "Virginia", 
					"Washington", "West Virginia", "Wisconsin", "Wyoming", "Puerto Rico",
					# note collapsed geographies from codebook
					"District of Columbia & Delaware", "Montana & Wyoming")
			)
	)
```

---

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

### Unweighted Counts {-}

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

svyby( ~ one , ~ age_group , nssrn_design , unwtd.count )
```

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

svyby( ~ one , ~ age_group , nssrn_design , svytotal )
```

### Descriptive Statistics {-}

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

svyby( ~ pn_earn_puf , ~ age_group , nssrn_design , svymean , na.rm = TRUE )
```

Calculate the distribution of a categorical variable, overall and by groups:
```{r eval = FALSE , results = "hide" }
svymean( ~ primary_position_state , nssrn_design , na.rm = TRUE )

svyby( ~ primary_position_state , ~ age_group , nssrn_design , svymean , na.rm = TRUE )
```

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

svyby( ~ pn_earn_puf , ~ age_group , nssrn_design , svytotal , na.rm = TRUE )
```

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

svyby( ~ primary_position_state , ~ age_group , nssrn_design , svytotal , na.rm = TRUE )
```

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

svyby( 
	~ pn_earn_puf , 
	~ age_group , 
	nssrn_design , 
	svyquantile , 
	0.5 ,
	ci = TRUE , na.rm = TRUE
)
```

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

### Subsetting {-}

Restrict the survey design to individuals working as RNs or APRNs (excluding RNs working as LPNs):
```{r eval = FALSE , results = "hide" }
sub_nssrn_design <- subset( nssrn_design , pn_lcreq_none == 2 )
```
Calculate the mean (average) of this subset:
```{r eval = FALSE , results = "hide" }
svymean( ~ pn_earn_puf , sub_nssrn_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( ~ pn_earn_puf , nssrn_design , na.rm = TRUE )

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

grouped_result <-
	svyby( 
		~ pn_earn_puf , 
		~ age_group , 
		nssrn_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( nssrn_design )
```

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

# SRS with replacement
svymean( ~ pn_earn_puf , nssrn_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( ~ all_aprn , nssrn_design ,
	method = "likelihood" )
```

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

Perform a design-based t-test:
```{r eval = FALSE , results = "hide" }
svyttest( pn_earn_puf ~ all_aprn , nssrn_design )
```

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

Perform a survey-weighted generalized linear model:
```{r eval = FALSE , results = "hide" }
glm_result <- 
	svyglm( 
		pn_earn_puf ~ all_aprn + primary_position_state , 
		nssrn_design 
	)

summary( glm_result )
```

---

## Replication Example {-}

This example matches statistics and relative standard errors from the "Demographics" tab of [Nursing Workforce 2022 NSSRN Dashboard Data](https://data.hrsa.gov//DataDownload/DD_Files/Nursing_Workforce_2022_NSSRN_Dashboard_Data.xlsx):

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

unwtd_count_result <- svyby( ~ one , ~ age_group , nssrn_design , unwtd.count )

# cells L398 thru L402
stopifnot( coef( unwtd_count_result ) == c( 6693 , 12268 , 10804 , 10538 , 8811 ) )

wtd_n_result <- svytotal( ~ age_group , nssrn_design )

# cells J398 thru J402
stopifnot( round( coef( wtd_n_result ) , 0 ) == c( 861060 , 1078187 , 935778 , 834939 , 639412 ) )

share_result <- svymean( ~ age_group , nssrn_design )

# cells K398 thru K402
stopifnot( round( coef( share_result ) , 3 ) == c( 0.198 , 0.248 , 0.215 , 0.192 , 0.147 ) )

# cells M398 thru M402
stopifnot( 
	round( SE( share_result ) / coef( share_result ) , 4 ) == 
	c( 0.0206 , 0.0155 , 0.0192 , 0.0187 , 0.0125 )
)
```

---

## 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 NSSRN users, this code replicates previously-presented examples:

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

nssrn_srvyr_design %>%
	group_by( age_group ) %>%
	summarize( mean = survey_mean( pn_earn_puf , na.rm = TRUE ) )
```