REDCap & Survey Export Cleanup
A registry of 500 patients built over 20 years in REDCap exports to Excel with coded fields, checkbox columns split across variables and blanks that have to be chased back to the notes or the chart. Isolating the ten variables a question needs takes 4 to 5 hours per project, before any analysis starts.
Teams need the export read the way REDCap and Qualtrics write it, the data dictionary used instead of retyped, each blank and code resolved by a recorded decision, and a dataset the analysis plan can start from the same day.
Done by hand, every project repeats the same 4 to 5 hours, the fixes live in one person's script, and the export is cleaned again by the next person who needs it.
4 to 5 hours of cleanup, done in minutes
The export is profiled, its codes and checkbox columns read, and a cleaning proposal produced in minutes; the person's time goes to the decisions, not the chasing.
18 identifier classes screened on upload
Every upload is screened for HIPAA-style identifier classes before anything else runs, and flagged columns are noted in the record.
The data dictionary read, not retyped
REDCap dictionaries and Qualtrics question-text rows become the codebook; checkbox fields, system columns and ImportId rows are recognised as such.
How it works
Connect or upload: a REDCap project through the connector, scoped to one project with the token stored encrypted and identifier fields excluded by default, or the export as a CSV, Excel, Qualtrics or REDCap file.
Audit and profile: export integrity, missing codes such as 999 and 888, duplicates, out-of-range values and label rows are reported with counts and locations.
Approve the cleaning proposal: each operation is listed with its count; one click records the decision, and nothing is dropped or filled without one.
Start the plan the same day: the verified dataset and its cleaning code hand straight to the analysis plan.