Case report form: collect what the protocol will actually analyse

A case report form is the instrument used to record the data specified by a clinical trial protocol for each participant. In modern trials it is almost always electronic, completed at the site and held in a data capture system, with the paper form surviving as a design artefact and a fallback.

Its design is a cost decision disguised as a documentation task. Every field is completed at every visit for every participant, checked, and queried when inconsistent. A form carrying twenty fields that no analysis will use imposes that burden on sites for the life of the study, and the effect on data quality is negative rather than neutral — attention is finite, and it is spread across everything asked for.

Case report form design rules that hold up

• Every field traces to a protocol objective, an endpoint, a safety requirement or a regulatory obligation. If nothing traces, the field goes.
• Collect raw values rather than derived ones — record the measurement and the date, and let analysis compute the interval or the change.
• One question per field, with unambiguous units stated on the form.
• Use coded response options rather than free text wherever a coded answer is possible; free text costs coding effort and produces inconsistent data.
• Match the flow of the visit, so the person completing it moves through it in the order things happen.
• Make the required fields genuinely required and the optional ones genuinely optional, because a form that fights the user gets worked around.

Edit checks belong in the design, not bolted on

Range checks, cross-field consistency and required-field logic prevent errors at the point of entry, which is enormously cheaper than a query raised weeks later. What they should not do is block a site from recording what actually happened: a check that refuses an out-of-range value forces staff to leave the field empty or to enter something plausible instead, and both are worse than an accurate value with a comment. Allow the truth, flag it, and let a human adjudicate.

Data recorded in the form must be traceable to source. Where the form is the first place a value is written, it becomes the source record, and that decision should be deliberate and documented rather than discovered during monitoring.

Changing it mid-study is expensive

Amending a form after enrolment starts means version control across sites, potential re-training, migration of already-collected data, and questions about what earlier participants were asked. That is a strong argument for spending longer on the design — including a genuine pilot with the people who will complete it — and a strong argument against adding fields because they might be interesting later.

Where the design lives

Ettex Forms holds the form design and its versions during development so the structure can be piloted and reviewed before it reaches a validated system, and Ettex Records keeps the design rationale, the field-to-endpoint mapping and the approvals. The file the study’s documents belong in is covered in trial master file.

Being direct, and this matters: we are not an electronic data capture system and must not be used to collect trial data. Clinical data collection belongs in a validated system meeting the applicable electronic records requirements, with audit trails and controlled access. What this covers is the design work that precedes it.