Align questionnaire recall periods with study visits

Treat the recall period as part of the measure

A questionnaire asking about today, the past week or the past month collects different information. The recall period affects what a respondent must remember and which events can contribute to an answer. It is part of the instrument definition, not a minor wording detail that can be changed for convenience.

The FDA guidance on fit-for-purpose clinical outcome assessments emphasises selecting or modifying measures for their intended use. Before data collection, match the measure's concept, recall period and administration schedule to the study question and available validation evidence.

Draw the reporting windows

Place each planned visit on a timeline and mark the period covered by its questionnaire. Check for gaps, overlapping windows and windows that include time before treatment begins. Decide how delayed or early visits will be handled and whether the instrument asks about a fixed calendar period or time relative to completion.

Preserve the actual completion date and the exact questionnaire version. Do not relabel a response as a seven-day endpoint if the instrument asked about the previous month. If the recall period is modified, document the rationale, permissions and evidence needed to support the changed interpretation.

Hypothetical example

A fictional study administers a past-seven-days symptom questionnaire every seven days. A participant completes the second assessment four days late. Its recall window now overlaps the intended third-week period and no longer matches the planned second-week window exactly.

The analyst records the actual date and applies the prespecified visit-window rule. Moving the response date back to the scheduled date would hide the mismatch; asking the participant to reconstruct an earlier answer after the fact would collect different information.

Check what the endpoint can mean

Consider whether the concept is stable enough to recall over the chosen period and whether rare events require a different collection approach. Shorter recall is not automatically superior if it misses the period relevant to the question, and longer recall is not automatically acceptable because it reduces visit burden.

Use the date-duration audit to test window boundaries and the instrument-scoring contract to preserve the wording. Report planned and actual timing, recall-period changes and handling of out-of-window responses. The final comparison should describe the period participants actually reported, with uncertainty about timing made explicit rather than concealed by a scheduled-visit label.

Source verification: 6 November 2026. Examples are hypothetical.