Study reports often describe blinding in broad terms, such as ‘double blind’. This term makes it impossible to know who was blinded (Schulz 2002a). Such terms are also used very inconsistently (Devereaux 2001, Boutron 2005, Haahr 2006), and the frequency of explicit reporting of the blinding status of study participants and personnel remains low even in trials published in top journals (Montori 2002), despite recommendations to be explicit in the CONSORT Statement (Moher 2001c). A review of methods used for blinding highlights the variety of methods used in practice (Boutron 2006). The following considerations may help review authors assess whether any blinding used in a study was likely to be sufficient to protect against bias, when using the Collaboration’s tool (Section 8.5).
When considering the risk of bias from lack of blinding it is important to consider specifically:
Who was and was not blinded;
Risk of bias in actual outcomes due to lack of blinding during the study (e.g. due to co-intervention or differential behaviour);
Risk of bias in outcome assessments (considering how subjective or objective an outcome is).
Assessors of some outcomes may be blinded while assessors of other outcomes are not. For example, in a surgical trial in which patients are aware of their own intervention, patient-reported outcomes (e.g. quality of life) would obviously be collected in knowledge of the intervention received, whereas other outcomes, measured by an independent clinician (e.g. physical ability), might be blinded. Furthermore, risk of bias may be high for some outcomes and low for others, even if the same people were unblinded in the study. For example, knowledge of the assigned intervention may impact on behavioural outcomes (such as number of clinic visits), while not impacting on physiological outcomes or mortality. In many circumstances assessment of total mortality might be considered to be unbiased, even if outcome assessors were aware of intervention assignments. Thus, assessments of risk of bias resulting from lack of blinding may need to be made separately for different outcomes.
Rather than assessing risk of bias for each outcome separately, it is often convenient to group outcomes with similar risks of bias (see Section 8.5). For example, there may be a common assessment of risk of bias for all subjective outcomes that is different from a common assessment of blinding for all objective outcomes.