collaborator: Scalable multi-centre research using R and REDCap

Collaborator: REDCap summary data

Evaluating data uploaded to a REDCap project in the context of multi-centre research projects is an important task for ensuring the highest quality dataset for subsequent analyses, and for sharing progress (whether internally or externally via social media). However, summarising this data, particularly at a DAG-level, can be a challenging and time-consuming task to produce on a regular basis.

The redcap_sum() function is designed to easily produce high quality and informative summary data on a REDCap project (overall and at a data_access_group level). This can be used for the purposes of sharing progress (including identifying top performing DAGs), and identifying individual DAGs which have not yet uploaded data or have not completed data upload.

The redcap_sum() function is designed to be simple from the point of use - the only requirements are a valid URI (Uniform Resource Identifier) and API (Application Programming Interface) for the REDCap project.

However, this is intended to have a high degree of customisability to fit the needs of a variety of projects. For example, being able to easily:

Select variables (using var_include or var_exclude) to be included in the assessment of completeness. For example, to focus on only essential variables to determine level of completeness.
Select individual records (using record_exclude or record_include) or whole data access groups (using dag_exclude or dag_include) to be assessed. For example to remove records or DAGs that were found to be ineligible.
Select individual users (using user_exclude or user_include) or whole data access groups (using dag_exclude or dag_include) to be assessed. For example to remove users (e.g. administrator user accounts) or DAGs that were found to be ineligible from the total REDCap user count.
Generation of summary data by DAG unless centre_sum is specified as FALSE (default centre_sum=T)

Limitations:

This function has not yet been tested on REDCap projects with multiple events.

At it’s most basic, redcap_sum() can produce an overall summary of current data on the REDCap project: - n_record_all is the number of all records currently on the REDCap project (minus any records removed using record_exclude or in DAGs removed using dag_exclude).

n_record_com is the number of all complete records currently on the REDCap project (minus any records removed using record_exclude or in DAGs removed using dag_exclude), with no missing data across the record (unless certain data fields are either excluded (var_exclude) or specified (var_complete)).
prop_com /pct_com is the respective proportion and percentage of complete records in the project.
n_dag is the number of data access groups (DAGs) for all records currently on the REDCap project (minus any records removed using record_exclude or in DAGs removed using dag_exclude).
n_users is the number of users on the REDCap project (minus any users in DAGs removed using dag_exclude). Note all users not assigned to a DAG will automatically be excluded.
last_update is the date which the summary data was generated on.

collaborator::redcap_sum(redcap_project_uri = Sys.getenv("collaborator_test_uri"),
                         redcap_project_token = Sys.getenv("collaborator_test_token"),
                         centre_sum = F)  %>%
  knitr::kable()

| n_record_all | n_record_com | prop_com | pct_com | n_dag | n_users | last_update | | -------------: | -------------: | --------: | :------- | -----: | -------: | :----------- | | 50 | 0 | 0 | 0% | 8 | 30 | 21-Apr-2020 |

However, more granular summary data on each DAG can also be obtained using the same function. This centre summary data will automatically be included within the output from redcap_sum() unless centre_sum is specified as FALSE.

1. `$dag_all` Output

This will produce a dataframe of the same summary data as outlined above grouped by each DAG instead (minus any DAGs removed using dag_exclude).

output <- collaborator::redcap_sum(redcap_project_uri = Sys.getenv("collaborator_test_uri"),
                                   redcap_project_token = Sys.getenv("collaborator_test_token"),
                                   centre_sum = T)

  knitr::kable(output$dag_all)

| redcap_data_access_group | record_all | record_com | prop_com | pct_com | user_all | last_update | | :-------------------------- | ----------: | ----------: | --------: | :------- | --------: | :----------- | | hospital_a | 10 | 0 | 0 | 0% | 3 | 21-Apr-2020 | | hospital_e | 9 | 0 | 0 | 0% | 4 | 21-Apr-2020 | | hospital_g | 7 | 0 | 0 | 0% | 2 | 21-Apr-2020 | | hospital_b | 6 | 0 | 0 | 0% | 3 | 21-Apr-2020 | | hospital_f | 6 | 0 | 0 | 0% | 2 | 21-Apr-2020 | | hospital_h | 6 | 0 | 0 | 0% | 4 | 21-Apr-2020 | | hospital_d | 4 | 0 | 0 | 0% | 3 | 21-Apr-2020 | | hospital_c | 2 | 0 | 0 | 0% | 1 | 21-Apr-2020 | | hospital_i | 0 | NA | NA | NA | 4 | 21-Apr-2020 | | hospital_j | 0 | NA | NA | NA | 4 | 21-Apr-2020 |

2. `$dag_nodata` Output

This will produce a dataframe of all DAG with users assigned on the REDCap project, but no data uploaded to REDCap. This may be useful for the purposes of targeting encouragement to upload data, or establishing authorship on any research output.

knitr::kable(output$dag_nodata)

3. `$dag_incom` Output

This will produce a dataframe of all DAG with incomplete records (the definition of completeness customisable as discussed above). This may be useful for the purposes of follow up regarding (essential) missing data at each of these DAGs.

knitr::kable(output$dag_incom)

4. `$dag_top_n` Output

This will produce a dataframe of the DAGs with the most records uploaded overall (the number of DAGs listed is defined by top with top 10 DAG being default). This may be useful for the purposes of publicity surrounding the project.

collaborator::redcap_sum(redcap_project_uri = Sys.getenv("collaborator_test_uri"),
                                   redcap_project_token = Sys.getenv("collaborator_test_token"),
                                   centre_sum = T, top = 5)$dag_top %>% 
  knitr::kable()

| redcap_data_access_group | record_all | | :-------------------------- | ----------: | | hospital_a | 10 | | hospital_e | 9 | | hospital_g | 7 | | hospital_b | 6 | | hospital_f | 6 |

kamclean/collaborator documentation built on Nov. 17, 2023, 3:52 a.m.

rdrr.io home R language documentation Run R code online

CRAN packages Bioconductor packages R-Forge packages GitHub packages

Note that we can't provide technical support on individual packages. You should contact the package authors for that.

Tweet to @rdrrHQ

GitHub issue tracker

ian@mutexlabs.com