Single data entry occurs when one person enters data into the research database. It is efficient and common, especially when electronic validation and routine quality checks are available. However, single entry is vulnerable to transcription errors, misread handwriting, skipped fields, and accidental entry into the wrong record. The risk can be reduced through validation rules, clear forms, user training, query review, and targeted source verification.

Double data entry involves two independent entries of the same data, followed by comparison and reconciliation. Traditionally, two clerks enter the same paper CRF into separate systems or entry screens. Differences are identified and resolved against the source document.  Double entry can improve accuracy for paper-based studies, especially where transcription error is a major concern, but it requires more staff time and may delay data availability.
Direct electronic capture records data directly into the electronic system at the point of collection. This may occur when a study nurse enters vital signs into REDCap during a visit, when a participant completes an electronic survey, or when a laboratory system sends results electronically. Direct capture reduces transcription and can apply validation immediately. However, it requires reliable procedures for authentication, device management, source definition, and user training. It also requires clear procedures for correcting errors.
The choice between single entry, double entry, and direct capture should be risk-based. A low-risk operational database may use single entry with routine validation. A pivotal clinical trial using paper CRFs for primary outcomes may require double entry or intensive verification. A study using tablets in the field may use direct capture with synchronization checks. The data manager should consider the criticality of variables, study design, regulatory expectations, staff capacity, infrastructure, and acceptable timelines.
In some studies, different workflows may apply to different data domains. Participant
reported questionnaires may be directly captured electronically, laboratory values may be imported from a spreadsheet, and historical clinical data may be abstracted from paper records.
Each workflow should be documented, and each should have quality controls appropriate to its risk.