This is an R package to interact with Public Health England’s
Fingertips data tool. Fingertips is a
major public repository of population and public health indicators for
England. The site presents the information in many ways to improve
accessibility for a wide range of audiences ranging from public health
professionals and researchers to the general public. The information
presented is a mixture of data available from other public sources, and
those that are available through user access agreements with other
organisations. The source of each indicator presented is available using
the indicator_metadata()
function.
This package can be used to load data from the Fingertips API into R for further use.
Get the latest released, stable version from rOpenSci:
# Enable repository from ropensci
options(repos = c(
ropensci = 'https://ropensci.r-universe.dev',
CRAN = 'https://cloud.r-project.org'))
# Download and install fingertipsR in R
install.packages('fingertipsR')
You can install the latest development version from github using remotes:
# install.packages("remotes")
remotes::install_github("rOpenSci/fingertipsR",
build_vignettes = TRUE,
dependencies = "suggests")
This is an example of a workflow for downloading data for the indicator on Healthy Life Expectancy at Birth from the Public Health Outcomes Framework profile.
The profiles()
function presents all of the available profiles:
library(fingertipsR)
profs <- profiles()
profs <- profs[grepl("Public Health Outcomes Framework", profs$ProfileName),]
head(profs)
#> # A tibble: 6 x 4
#> ProfileID ProfileName DomainID DomainName
#> <int> <chr> <int> <chr>
#> 1 19 Public Health Outcomes Framework 1000049 A. Overarching indicato~
#> 2 19 Public Health Outcomes Framework 1000041 B. Wider determinants o~
#> 3 19 Public Health Outcomes Framework 1000042 C. Health improvement
#> 4 19 Public Health Outcomes Framework 1000043 D. Health protection
#> 5 19 Public Health Outcomes Framework 1000044 E. Healthcare and prema~
#> 6 19 Public Health Outcomes Framework 1938132983 Supporting information
This table shows that the ProfileID
for the Public Health Outcomes
Framework is 19. This can be used as an input for the indicators()
function:
profid <- 19
inds <- indicators(ProfileID = profid)
print(inds[grepl("Healthy", inds$IndicatorName), c("IndicatorID", "IndicatorName")])
#> # A tibble: 2 x 2
#> IndicatorID IndicatorName
#> <int> <fct>
#> 1 90362 A01a - Healthy life expectancy at birth
#> 2 93505 A01a - Healthy life expectancy at 65
Healthy Life Expectancy at Birth has the IndicatorID
equal to 90362.
Finally, the data can be extracted using the fingertips_data()
function using that IndicatorID
:
indid <- 90362
df <- fingertips_data(IndicatorID = indid, AreaTypeID = 202)
head(df)
#> IndicatorID IndicatorName ParentCode ParentName AreaCode
#> 1 90362 Healthy life expectancy at birth <NA> <NA> E92000001
#> 2 90362 Healthy life expectancy at birth <NA> <NA> E92000001
#> 3 90362 Healthy life expectancy at birth E92000001 England E12000001
#> 4 90362 Healthy life expectancy at birth E92000001 England E12000002
#> 5 90362 Healthy life expectancy at birth E92000001 England E12000003
#> 6 90362 Healthy life expectancy at birth E92000001 England E12000004
#> AreaName AreaType Sex Age CategoryType
#> 1 England England Male All ages <NA>
#> 2 England England Female All ages <NA>
#> 3 North East region Region Male All ages <NA>
#> 4 North West region Region Male All ages <NA>
#> 5 Yorkshire and the Humber region Region Male All ages <NA>
#> 6 East Midlands region Region Male All ages <NA>
#> Category Timeperiod Value LowerCI95.0limit UpperCI95.0limit
#> 1 <NA> 2009 - 11 63.02647 62.87787 63.17508
#> 2 <NA> 2009 - 11 64.03794 63.88135 64.19453
#> 3 <NA> 2009 - 11 59.71114 59.19049 60.23179
#> 4 <NA> 2009 - 11 60.76212 60.39880 61.12544
#> 5 <NA> 2009 - 11 60.84033 60.38649 61.29417
#> 6 <NA> 2009 - 11 62.60207 62.07083 63.13332
#> LowerCI99.8limit UpperCI99.8limit Count Denominator Valuenote RecentTrend
#> 1 NA NA NA NA <NA> <NA>
#> 2 NA NA NA NA <NA> <NA>
#> 3 NA NA NA NA <NA> <NA>
#> 4 NA NA NA NA <NA> <NA>
#> 5 NA NA NA NA <NA> <NA>
#> 6 NA NA NA NA <NA> <NA>
#> ComparedtoEnglandvalueorpercentiles ComparedtoRegionvalueorpercentiles
#> 1 Not compared Not compared
#> 2 Not compared Not compared
#> 3 Worse Not compared
#> 4 Worse Not compared
#> 5 Worse Not compared
#> 6 Similar Not compared
#> TimeperiodSortable Newdata Comparedtogoal Timeperiodrange
#> 1 20090000 <NA> <NA> 3y
#> 2 20090000 <NA> <NA> 3y
#> 3 20090000 <NA> <NA> 3y
#> 4 20090000 <NA> <NA> 3y
#> 5 20090000 <NA> <NA> 3y
#> 6 20090000 <NA> <NA> 3y
Please see the vignettes for information on use.
browseVignettes("fingertipsR")