Skip to contents

Build a REDCap data quality report

Source: R/quality_report.R
build_quality_report.Rd

build_quality_report() pulls a REDCap project through the API, applies general data quality heuristics, and returns findings, summaries, and interpreted metadata.

Usage

build_quality_report(
  redcap_uri,
  token,
  checks = c("missingness", "metadata", "outliers", "operational", "consistency"),
  sparse_threshold = 0.95,
  outlier_iqr_multiplier = 3,
  progress = c("auto", "none", "show")
)

Arguments

redcap_uri

REDCap API URI.

token

REDCap API token.

checks

Character vector of check groups to run.

sparse_threshold

Number greater than 0 and less than or equal to 1. A field is flagged as unexpectedly sparse when its missing rate among applicable rows is greater than or equal to this value. The default 0.95 flags fields with at least 95 percent missing values.

outlier_iqr_multiplier

Positive number used to define IQR outlier bounds as Q1 minus this value times the IQR and Q3 plus this value times the IQR. Smaller values produce narrower bounds and more findings. The default is 3.

progress

Progress display mode. "auto" enables normally throttled progress updates only in interactive sessions. "none" suppresses progress. "show" enables progress in all sessions and forces every update to render.

Value

A list with findings, summaries, and metadata elements. findings includes record, form, event, repeat instrument, and repeat instance context when available from the REDCap export. summaries$project$raw_row_count equals nrow() for the records returned by the REDCap API. summaries$project$field_count is the count of distinct standardized metadata field names. summaries$fields$missing_rate is the field-level fraction missing among applicable rows. summaries$records contains one row per record; its missing-field count is summed across applicable event and repeat rows. summaries$forms$missing_rate and summaries$project$missing_rate are weighted fractions missing across applicable field-row cells. Applicability accounts for form availability, event-form mapping, repeating instrument structure, and branching logic. Missingness is only assessed where the raw REDCap form status column <form_name>_complete is 1/Unverified or 2/Complete. Status 0/Incomplete is handled by operational checks and does not contribute to missingness. Checkbox parent fields are observed when their choice exports contain explicit values, including an all-zero no-selection state; only an all-missing set of choice exports is missing. Branching expressions that use unsupported REDCap functions or smart variables are treated as applicable so required values remain reviewable.

Details

Current findings$issue values by findings$check:

  • missingness

    • required_field_missing

    • unexpected_sparse_field

  • metadata

    • duplicate_field_label

    • missing_field_label

    • missing_choice_definition

    • high_risk_free_text

  • outliers

    • outside_validation_range

    • numeric_iqr_outlier

    • future_date

  • operational

    • incomplete_form_status

  • consistency

    • invalid_choice_value

    • invalid_validation_format

    • checkbox_no_values_selected

    • checkbox_none_with_other

invalid_validation_format is evaluated only for text fields with a supported validation configured in text_validation_type_or_show_slider_number. Supported validations include REDCap integer and number formats, dates and datetimes, times, email addresses, and North American and Australian phone numbers. Number formats accept decimal values with or without a leading zero. Date and datetime values are validated in REDCap's canonical API export formats, regardless of their configured data-entry display order. outside_validation_range enforces the optional text_validation_min and text_validation_max bounds using the configured validation type. Unvalidated text fields and unsupported validation types are not assessed. invalid_choice_value applies to radio, dropdown, Yes/No, and True/False fields; checkbox fields retain their dedicated consistency checks. Choice and text validation checks use structurally applicable form, event, and repeat rows regardless of form completion status or branching logic.

Examples

if (FALSE) { # \dontrun{
report <- build_quality_report(
  redcap_uri = Sys.getenv("REDCAP_URI"),
  token = Sys.getenv("REDCAP_TOKEN")
)
} # }