Understanding blinding in trials

In this episode, hear from Dr Kate Scrivener, Dr Aidan Cashin and Clinical Associate Professor Mark Elkins about understanding the importance of blinding in trials.

There are numerous stakeholders involved in any clinical trial. These include patients and participants, therapists, researchers, outcome assessors and statisticians.

Stakeholders are a source of bias in trials. This is because they can consciously or subconsciously influence procedures or results based on knowing whether a patient has been allocated to the intervention or control group. To minimise biases, a trial can ‘blind’ the stakeholders to which group participants are allocated. Blinding is considered successful if the stakeholder is unable to distinguish between the treatments applied to the groups.

Three important people or groups to be blinded in clinical trials include the:

1. Patient or participant: where the patient is unaware if they are receiving the intervention or control
2. Therapist: where the therapist is unaware if they are delivering the intervention or control
3. Assessor: where outcome assessor(s) are unaware if the participant being assessed had received the intervention or control

In most physiotherapy trials, it is very difficult to blind participants and therapists. For example, if interventions are physical or active (e.g. exercise), participants will know they are receiving the intervention and therapists will know if they are delivering it. In regard to blinding the assessor, blinding is successful if the assessor does not know which group the patient has been allocated to and outcome measures are objective (e.g. passive range of motion). However, when outcome measures are patient-reported or self-reported (e.g. pain), the assessor is considered blind if the patient was blinded.

Studies frequently report the occurrence of blinding in the title or abstract using terms such as ‘single blinded’ or ‘double blinded’. However, there is inconsistent use of these terms. For example, one ‘double blinded’ trial may have blinded the therapists and outcome assessors, while another may have blinded the patients and statisticians. Readers should investigate which elements of a clinical trial have been blinded and authors should avoid this ambiguous terminology and explicitly state who was blinded.

Some clinical trials attempt to blind patients to their allocated group by providing control interventions that are similar to active interventions. To assess the perceived similarity of the control and active intervention, some studies report the ‘treatment credibility’, where patients are asked ‘How convinced are you that you have received an active therapy?’. Similar treatment credibility between the active and control interventions usually indicates successful blinding.

Many people in a clinical trial can be blinded. Although blinding helps minimise biases, it is often difficult to blind every person. Readers need to assess how a lack of blinding could influence the conduct and reporting of a trial.