redcaptools

library(redcaptools)

redcaptools assists in exporting data from REDCap via the API, converting variable types (e.g. radio-buttons to factors), labelling etc.

Exporting from REDCap

REDCap exports generally come in a single files, with repeating elements as new rows, lots of empty variables (e.g. because they don’t exist for a given visit). To use the data, it must often be subsetted by form or event, variables selected, etc.

redcaptools eases this by exporting the data by form from the beginning with the redcap_export_byform function.

token <- "my-redcap-token"
url <- "https://redcap.mydomain.com/api/"

data <- redcap_export_byform(token, url)

Other export functions are redcap_export_meta (for downloading various metadata objects - currently the data dictionary, form-event-mapping and event tables) and redcap_export_tbl, which can be used to use whatever parameters you like for the API (see the REDCap documentation for details). The dataset that REDCap exports through the website interface is accessed for instance via:

record <- redcap_export_tbl(token, url, "record")

Using manual exports

REDCap also allows exporting data manually through the web interface. While redcaptools is primarily intended to used with the API, there much of it will also work with manually exported datasets. Important to note is that the data dictionary must be provided manually in this case.

For example, the redcap_toform function tries to split a manually downloaded dataset into it’s constituent forms. As less information is available, it’s not as robust as the API version, but it may still be useful.

# a small manually downloaded testing dataset
data <- readRDS(system.file("extdata/test.rda", package = "redcaptools"))
# the data dictionary
dd <- read.csv(system.file("extdata/DataDictionary.csv", package = "redcaptools"))

formdata <- redcap_toform(data, dd)

Processing REDCap data

To make the data more usable in R, it’s often nice to label data, create factor, convert dates etc. redcaptools has functions to help with that too. Each of the following functions also label variables.

The redcap_prep function is the easiest to use. It wraps all of the functions in the following section.

meta <- redcap_export_meta(token, url)
# could also use a data dictionary exported manually

# where all data is in a single dataframe
prepped <- redcap_prep(record, meta$metadata)

# for data split by forms, use sapply
prepped <- sapply(data, redcap_prep, metadata = meta$metadata)

Create factors

REDCap has both single and multiple choice factor type variables. These need handling separately, so there are two sets of functions.

Lists of the options can be obtained with the singlechoice_opts and multichoice_opts functions. These are rather for your information.

singlechoice_opts(meta$metadata)
mutlichoice_opts(meta$metadata)

Factors can be created easily with

prepped <- sapply(data, singlechoice_factor, metadata = meta$metadata)
prepped <- sapply(prepped, multichoice_factor, metadata = meta$metadata)

Create `Date`s and `POSIX` variables

As with single and multiple choice variables, Date and POSIX variables also need handling separately.