Report blinding procedures and breaches precisely

Replace the label with the actual procedure

Descriptions such as single-blind or double-blind can leave readers uncertain about who knew the assigned intervention. Start by naming the roles: participants, care providers, outcome assessors, data collectors and analysts. Explain how assignment information was concealed from each role after allocation and where blinding was not feasible.

The CONSORT 2025 explanation and elaboration asks for clear reporting of who was blinded and how. Allocation concealment before assignment and blinding after assignment are different safeguards; reporting one does not establish the other.

Record practical threats to masking

Describe intervention appearance, administration procedures, access to treatment records and emergency unblinding arrangements where relevant. Record actual breaches with timing, reason and affected role, using the study's authorised reporting process. Explain whether a breach could have affected care, outcome assessment or analysis decisions.

For interventions that cannot be masked, consider whether outcome assessment or analysis can still be protected. State those measures specifically rather than describing the entire study as blinded. Objective and subjective outcomes may have different vulnerability to knowledge of assignment.

Hypothetical example

In a fictional trial, participants and treating staff know the intervention, but centrally reviewed images are labelled with study codes and assessed without treatment information. The report should describe blinded image assessment and open treatment delivery. Calling the trial simply double-blind would misrepresent the procedure.

If an image reviewer accidentally receives an unmasked clinical note before assessment, the team records the event and follows the prespecified reassessment or sensitivity procedure. It should not silently delete the breach from the study narrative.

Interpret blinding assessments cautiously

If participants or staff are asked to guess assignment, report the timing, response options and reason for collecting that information. Correct guesses can reflect perceived effects or adverse effects as well as failure of the masking procedure. A guess distribution alone does not prove that blinding succeeded or failed in every relevant role.

Keep any assessment descriptive unless a justified analysis was planned. Use the protocol deviation classification worksheet to retain the event trail. The final methods and results should allow readers to judge which decisions were protected from assignment knowledge, which were exposed and how important deviations were handled.

Source verification: 6 November 2026. Examples are hypothetical.