This vignette covers the the less-typical uses of REDCapR to interact with REDCap through its API.
There is some information that is specific to a REDCap project, as opposed to an individual operation. This includes the (1) uri of the server, and the (2) token for the user’s project. This is hosted on a machine used in REDCapR’s public test suite, so you can run this example from any computer. Unless tests are running.
Other than PHI-free demos, we strongly suggest storing tokens
securely and avoiding hard-coding them like below. Our recommendation is
to store tokens in
a database. If that is not feasible for your institution, consider
storing them in a secured csv and retrieving with REDCapR::retrieve_credential_local().
Disclaimer: Occasionally we’re asked for a longitudinal dataset to be converted from a “long/tall format” (where typically each row is one observation for a participant) to a “wide format” (where each row is on participant). Usually we advise against it. Besides all the database benefits of a long structure, a wide structure restricts your options with the stat routine. No modern longitudinal analysis procedures (e.g., growth curve models or multilevel/hierarchical models) accept wide. You’re pretty much stuck with repeated measures anova, which is very inflexible for real-world medical-ish analyses. It requires a patient to have a measurement at every time point; otherwise the anova excludes the patient entirely.
However we like going wide to produce visual tables for publications, and here’s one way to do it in R. First retrieve the dataset from REDCap.
library(magrittr)
suppressPackageStartupMessages(requireNamespace("dplyr"))
suppressPackageStartupMessages(requireNamespace("tidyr"))
events_to_retain <- c("dose_1_arm_1", "visit_1_arm_1", "dose_2_arm_1", "visit_2_arm_1")
ds_long <- REDCapR::redcap_read_oneshot(redcap_uri = uri, token = token_longitudinal)$data#> 18 records and 125 columns were read from REDCap in 0.2 seconds. The http status code was 200.#> # A tibble: 18 × 6
#> study_id redcap_event_name pmq1 pmq2 pmq3 pmq4
#> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 100 enrollment_arm_1 NA NA NA NA
#> 2 100 dose_1_arm_1 2 2 1 1
#> 3 100 visit_1_arm_1 1 0 0 0
#> 4 100 dose_2_arm_1 3 1 0 0
#> 5 100 visit_2_arm_1 0 1 0 0
#> 6 100 final_visit_arm_1 NA NA NA NA
#> 7 220 enrollment_arm_1 NA NA NA NA
#> 8 220 dose_1_arm_1 0 1 0 2
#> 9 220 visit_1_arm_1 0 3 1 0
#> 10 220 dose_2_arm_1 1 2 0 1
#> 11 220 visit_2_arm_1 3 4 1 0
#> 12 220 final_visit_arm_1 NA NA NA NA
#> 13 304 enrollment_arm_2 NA NA NA NA
#> 14 304 deadline_to_opt_ou_arm_2 NA NA NA NA
#> 15 304 first_dose_arm_2 0 1 0 0
#> 16 304 first_visit_arm_2 2 0 0 0
#> 17 304 final_visit_arm_2 NA NA NA NA
#> 18 304 deadline_to_return_arm_2 NA NA NA NAWhen widening only one variable (e.g., pmq1),
the code’s pretty simple:
ds_wide <-
ds_long %>%
dplyr::select(study_id, redcap_event_name, pmq1) %>%
dplyr::filter(redcap_event_name %in% events_to_retain) %>%
tidyr::pivot_wider(
id_cols = study_id,
names_from = redcap_event_name,
values_from = pmq1
)
ds_wide#> # A tibble: 2 × 5
#> study_id dose_1_arm_1 visit_1_arm_1 dose_2_arm_1 visit_2_arm_1
#> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 100 2 1 3 0
#> 2 220 0 0 1 3In some scenarios, multiple variables (e.g.,
pmq1 - pmq4) can be widened in a single tidyr::pivot_wider()
operation. This example contains the additional wrinkle that the REDCap
event names “first_dose” and “first_visit” are renamed “dose_1” and
“visit_1”, which will help all the values be dose and visit values be
proper numbers.
pattern <- "^(\\w+?)_arm_(\\d)$"
ds_wide <-
ds_long %>%
dplyr::select(study_id, redcap_event_name, pmq1, pmq2, pmq3, pmq4) %>%
dplyr::mutate(
event = sub(pattern, "\\1", redcap_event_name),
event = dplyr::recode(event, "first_dose"="dose_1", "first_visit"="visit_1"),
arm = as.integer(sub(pattern, "\\2", redcap_event_name))
) %>%
dplyr::select(study_id, event, arm, pmq1, pmq2, pmq3, pmq4) %>%
dplyr::filter(!(event %in%
c("enrollment", "final_visit", "deadline_to_return", "deadline_to_opt_ou")
)) %>%
tidyr::pivot_wider(
id_cols = c(study_id, arm),
names_from = event,
values_from = c(pmq1, pmq2, pmq3, pmq4)
)
ds_wide#> # A tibble: 3 × 18
#> study_id arm pmq1_dose_1 pmq1_visit_1 pmq1_dose_2 pmq1_visit_2 pmq2_dose_1
#> <dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 100 1 2 1 3 0 2
#> 2 220 1 0 0 1 3 1
#> 3 304 2 0 2 NA NA 1
#> # ℹ 11 more variables: pmq2_visit_1 <dbl>, pmq2_dose_2 <dbl>,
#> # pmq2_visit_2 <dbl>, pmq3_dose_1 <dbl>, pmq3_visit_1 <dbl>,
#> # pmq3_dose_2 <dbl>, pmq3_visit_2 <dbl>, pmq4_dose_1 <dbl>,
#> # pmq4_visit_1 <dbl>, pmq4_dose_2 <dbl>, pmq4_visit_2 <dbl>However, in other widening scenarios, it can be easier to go even
longer/taller (e.g., ds_eav) before reversing
direction and going wide.
ds_eav <-
ds_long %>%
dplyr::select(study_id, redcap_event_name, pmq1, pmq2, pmq3, pmq4) %>%
dplyr::mutate(
event = sub(pattern, "\\1", redcap_event_name),
event = dplyr::recode(event, "first_dose" = "dose_1", "first_visit" = "visit_1"),
arm = as.integer(sub(pattern, "\\2", redcap_event_name))
) %>%
dplyr::select(study_id, event, arm, pmq1, pmq2, pmq3, pmq4) %>%
tidyr::pivot_longer(
cols = c(pmq1, pmq2, pmq3, pmq4),
names_to = "key",
values_to = "value"
) %>%
# For old versions of tidyr that predate `pivot_wider()`:
# tidyr::gather(key=key, value=value, pmq1, pmq2, pmq3, pmq4) %>%
dplyr::filter(!(event %in% c(
"enrollment", "final_visit", "deadline_to_return", "deadline_to_opt_ou")
)) %>%
dplyr::mutate( # Simulate correcting for mismatched names across arms:
key = paste0(key, "_", event)
) %>%
dplyr::select(-event)
# Show the first 10 rows of the EAV table.
ds_eav %>%
head(10)#> # A tibble: 10 × 4
#> study_id arm key value
#> <dbl> <int> <chr> <dbl>
#> 1 100 1 pmq1_dose_1 2
#> 2 100 1 pmq2_dose_1 2
#> 3 100 1 pmq3_dose_1 1
#> 4 100 1 pmq4_dose_1 1
#> 5 100 1 pmq1_visit_1 1
#> 6 100 1 pmq2_visit_1 0
#> 7 100 1 pmq3_visit_1 0
#> 8 100 1 pmq4_visit_1 0
#> 9 100 1 pmq1_dose_2 3
#> 10 100 1 pmq2_dose_2 1# Spread the EAV to wide.
ds_wide_2 <-
ds_eav %>%
tidyr::pivot_wider(
id_cols = c(study_id, arm),
names_from = key,
values_from = value
)
# For old versions of tidyr that predate `pivot_wider()`:
# tidyr::spread(key=key, value=value)
ds_wide_2#> # A tibble: 3 × 18
#> study_id arm pmq1_dose_1 pmq2_dose_1 pmq3_dose_1 pmq4_dose_1 pmq1_visit_1
#> <dbl> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 100 1 2 2 1 1 1
#> 2 220 1 0 1 0 2 0
#> 3 304 2 0 1 0 0 2
#> # ℹ 11 more variables: pmq2_visit_1 <dbl>, pmq3_visit_1 <dbl>,
#> # pmq4_visit_1 <dbl>, pmq1_dose_2 <dbl>, pmq2_dose_2 <dbl>,
#> # pmq3_dose_2 <dbl>, pmq4_dose_2 <dbl>, pmq1_visit_2 <dbl>,
#> # pmq2_visit_2 <dbl>, pmq3_visit_2 <dbl>, pmq4_visit_2 <dbl>Lots of packages and documentation exist. Our current preference is the tidyverse approach to pivoting, but the data.table approach is worth considering if you’re comfortable with that package. This Stack Overflow post describes several ways. We recommend against the reshape and reshape2 packages, because their developers have replaced them with the tidyr functions described above.
If you require a feature that is not available from your instance’s API, first upgrade your institution’s REDCap instance and see if the feature has been added recently. Second, check if someone has released the desired API-like features as an REDCap External Module.
Third, you may need to query the database underneath REDCap’s web server. The Transfer Credentials section of the Security Database Vignette provides a complete example of using R to query the MySQL database through odbc.
We find it’s best to develop the query in MySQL Workbench,
then copy the code to R (or alternatively, use OuhscMunge::execute_sql_file()).
Here is an example that retrieves the first_submit_time,
which is helpful if you need a timestamp from surveys that were not
marked as completed. Replace ‘444’ with your pid, and 1001 through 1003
with the desired events.
SELECT
p.participant_id as participant_survey_id
,r.record as record_id
,p.event_id
,e.descrip as event_name
,r.first_submit_time
,r.completion_time
-- ,p.*
-- ,r.*
FROM redcapv3.redcap_surveys_participants as p
left join redcapv3.redcap_surveys_response as r on p.participant_id = r.participant_id
left join redcapv3.redcap_events_metadata as e on p.event_id = e.event_id
WHERE
p.survey_id = 444
and
p.event_id in (
1001, -- start of the year
1002, -- mid term
1003 -- end of year
)The official cURL site discusses the process of using SSL to verify the server being connected to.
Use the SSL cert file that come with the openssl
package.
cert_location <- system.file("cacert.pem", package = "openssl")
if (file.exists(cert_location)) {
config_options <- list(cainfo = cert_location)
ds_different_cert_file <- redcap_read_oneshot(
redcap_uri = uri,
token = token_simple,
config_options = config_options
)$data
}#> 5 records and 25 columns were read from REDCap in 0.2 seconds. The http status code was 200.Force the connection to use SSL=3 (which is not preferred, and possibly insecure).
config_options <- list(sslversion = 3)
ds_ssl_3 <- redcap_read_oneshot(
redcap_uri = uri,
token = token_simple,
config_options = config_options
)$data#> 5 records and 25 columns were read from REDCap in 0.2 seconds. The http status code was 200.config_options <- list(ssl.verifypeer = FALSE)
ds_no_ssl <- redcap_read_oneshot(
redcap_uri = uri,
token = token_simple,
config_options = config_options
)$data#> 5 records and 25 columns were read from REDCap in 0.2 seconds. The http status code was 200.The solution https://stackoverflow.com/a/51013678/1082435 converts levels specified in SPSS output like
SEX 0 Male
1 Female
LANGUAGE 1 English
2 Spanish
3 Other
6 Unknownto a dropdown choices in a REDCap data dictionary like
Variable Values
SEX 0, Male | 1, Female
LANGUAGE 1, English | 2, Spanish | 3, Other | 6, UnknownFor the sake of documentation and reproducibility, the current report was rendered in the following environment. Click the line below to expand.
#> ─ Session info ───────────────────────────────────────────────────────────────
#> setting value
#> version R version 4.4.1 Patched (2024-07-03 r86870 ucrt)
#> os Windows 11 x64 (build 26100)
#> system x86_64, mingw32
#> ui RTerm
#> language (EN)
#> collate C
#> ctype English_United States.utf8
#> tz America/Chicago
#> date 2024-10-22
#> pandoc 3.2 @ C:/Program Files/RStudio/resources/app/bin/quarto/bin/tools/ (via rmarkdown)
#>
#> ─ Packages ───────────────────────────────────────────────────────────────────
#> package * version date (UTC) lib source
#> archive 1.1.9 2024-09-12 [3] CRAN (R 4.4.1)
#> backports 1.5.0 2024-05-23 [3] CRAN (R 4.4.0)
#> bit 4.5.0 2024-09-20 [3] CRAN (R 4.4.1)
#> bit64 4.5.2 2024-09-22 [3] CRAN (R 4.4.1)
#> bslib 0.8.0 2024-07-29 [3] CRAN (R 4.4.1)
#> cachem 1.1.0 2024-05-16 [3] CRAN (R 4.4.0)
#> checkmate 2.3.2 2024-07-29 [3] CRAN (R 4.4.1)
#> cli 3.6.3 2024-06-21 [3] CRAN (R 4.4.1)
#> colorspace 2.1-1 2024-07-26 [3] CRAN (R 4.4.1)
#> crayon 1.5.3 2024-06-20 [3] CRAN (R 4.4.1)
#> curl 5.2.3 2024-09-20 [3] CRAN (R 4.4.1)
#> digest 0.6.37 2024-08-19 [3] CRAN (R 4.4.1)
#> dplyr 1.1.4 2023-11-17 [3] CRAN (R 4.4.0)
#> evaluate 1.0.1 2024-10-10 [3] CRAN (R 4.4.1)
#> fansi 1.0.6 2023-12-08 [3] CRAN (R 4.4.0)
#> fastmap 1.2.0 2024-05-15 [3] CRAN (R 4.4.0)
#> generics 0.1.3 2022-07-05 [3] CRAN (R 4.4.0)
#> glue 1.8.0 2024-09-30 [3] CRAN (R 4.4.1)
#> highr 0.11 2024-05-26 [3] CRAN (R 4.4.0)
#> hms 1.1.3 2023-03-21 [3] CRAN (R 4.4.0)
#> htmltools 0.5.8.1 2024-04-04 [3] CRAN (R 4.4.0)
#> httr 1.4.7 2023-08-15 [3] CRAN (R 4.4.0)
#> jquerylib 0.1.4 2021-04-26 [3] CRAN (R 4.4.0)
#> jsonlite 1.8.9 2024-09-20 [3] CRAN (R 4.4.1)
#> kableExtra 1.4.0 2024-01-24 [3] CRAN (R 4.4.0)
#> knitr * 1.48 2024-07-07 [3] CRAN (R 4.4.1)
#> lifecycle 1.0.4 2023-11-07 [3] CRAN (R 4.4.0)
#> magrittr * 2.0.3 2022-03-30 [3] CRAN (R 4.4.0)
#> munsell 0.5.1 2024-04-01 [3] CRAN (R 4.4.0)
#> pillar 1.9.0 2023-03-22 [3] CRAN (R 4.4.0)
#> pkgconfig 2.0.3 2019-09-22 [3] CRAN (R 4.4.0)
#> png 0.1-8 2022-11-29 [3] CRAN (R 4.4.0)
#> purrr 1.0.2 2023-08-10 [3] CRAN (R 4.4.0)
#> R6 2.5.1 2021-08-19 [3] CRAN (R 4.4.0)
#> readr 2.1.5 2024-01-10 [3] CRAN (R 4.4.0)
#> REDCapR * 1.3.0 2024-10-22 [1] local
#> rlang 1.1.4 2024-06-04 [3] CRAN (R 4.4.0)
#> rmarkdown 2.28 2024-08-17 [3] CRAN (R 4.4.1)
#> rstudioapi 0.17.0 2024-10-16 [3] CRAN (R 4.4.1)
#> sass 0.4.9 2024-03-15 [3] CRAN (R 4.4.0)
#> scales 1.3.0 2023-11-28 [3] CRAN (R 4.4.0)
#> sessioninfo 1.2.2 2021-12-06 [3] CRAN (R 4.4.0)
#> stringi 1.8.4 2024-05-06 [3] CRAN (R 4.4.0)
#> stringr 1.5.1 2023-11-14 [3] CRAN (R 4.4.0)
#> svglite 2.1.3 2023-12-08 [3] CRAN (R 4.4.0)
#> systemfonts 1.1.0 2024-05-15 [3] CRAN (R 4.4.0)
#> tibble 3.2.1 2023-03-20 [3] CRAN (R 4.4.0)
#> tidyr 1.3.1 2024-01-24 [3] CRAN (R 4.4.0)
#> tidyselect 1.2.1 2024-03-11 [3] CRAN (R 4.4.0)
#> tzdb 0.4.0 2023-05-12 [3] CRAN (R 4.4.0)
#> utf8 1.2.4 2023-10-22 [3] CRAN (R 4.4.0)
#> vctrs 0.6.5 2023-12-01 [3] CRAN (R 4.4.0)
#> viridisLite 0.4.2 2023-05-02 [3] CRAN (R 4.4.0)
#> vroom 1.6.5 2023-12-05 [3] CRAN (R 4.4.0)
#> withr 3.0.1 2024-07-31 [3] CRAN (R 4.4.1)
#> xfun 0.48 2024-10-03 [3] CRAN (R 4.4.1)
#> xml2 1.3.6 2023-12-04 [3] CRAN (R 4.4.0)
#> yaml 2.3.10 2024-07-26 [3] CRAN (R 4.4.1)
#>
#> [1] C:/Users/wibea/AppData/Local/Temp/Rtmpe4oEow/Rinstbcbcf4f69bf
#> [2] C:/Users/wibea/AppData/Local/Temp/RtmpQ39SYc/temp_libpath4ef035ed24cb
#> [3] D:/projects/r-libraries
#> [4] C:/Users/wibea/AppData/Local/R/win-library/4.4
#> [5] C:/Program Files/R/R-4.4.1patched/library
#>
#> ──────────────────────────────────────────────────────────────────────────────Report rendered by wibea at 2024-10-22, 18:47 -0500 in 1 seconds.