Data Collection Standard Practice Guidelines


It is critical that data used for research purposes is collected in a manner which allows it to be traced back to its original source. Data collection instruments, which are often called Case Report Forms (CRFs), are frequently created in both paper and electronic format. In some cases, the data on the CRF is the original source of information; at other times the CRF is used to transcribe data collected from another source, such as a medical record or lab report.

Any revision to the data on a CRF, whether serving as a source document or not, should be clearly documented without destroying the underlying or original results. Case Report Forms must also be created in a manner that allows the data to be clearly linked to the appropriate research study, site, subject, and study event (time point) for successful management of the data.

Developing Paper CRFs

When developing paper CRFs, include consistent headers and footers for easy identification and tracking of subject CRFs. Headers typically include the study name, HUM number, CRF name, Subject ID number, visit, etc. Footers often include the CRF version number, creation date, and CRF page number(s) such as “page 1 of 2.”


ALCOA is a mnemonic for the fundamentally recommended elements of data quality These element are noted by the FDA in Guidance for Industry - Electronic Source Data in Clinical Investigations and by ICH GCP 4.9 Records and Reports.

Following the principles behind ALCOA will ensure the data are:

Attributable: It should be clear who created a record and when. It should be clear as to who amended a record, when, and the reason for the change. Electronic source records should have 21 CFR Part 11 compliant audit trails.

Legible: The research record should be easily read and recorded. If using written records, pencil should not be used.

Contemporaneous: Study evidence/results should be recorded as they are observed. All signatures/initials should be attached to a date indicated when the signature was added to the document.

Original: Research records should be original (where the data was initially recorded), not photocopies.

Accurate: Research records should have a high level of integrity and honesty to what was truly observed and should give a full accounting of the research process. Research records should be thorough and correct; double-check your work for unintentional errors.

Making Corrections to Data

Electronic CRFs are frequently linked to an Electronic Data Capture (EDC) system such as REDCap. Most electronic systems will maintain an audit trail that automatically tracks any modifications to the data. When making corrections to data on paper CRFs it is best to cross out the errant data with a single line. The person making the correction should also include his/her initials and date of change.

Unacceptable ways to make corrections include:

  • Use of white-out or pencil

  • Revisions made by unqualified or unauthorized staff


Data Collection
Appendix A Document Tracking Log


Contact us at