Bias due to MEasurement Reactions In Trials to improve health (MERIT): protocol for research to develop MRC guidance.

BACKGROUND: There is now clear systematic review evidence that measurement can affect the people being measured; much of this evidence focusses on how asking people to complete a questionnaire can result in changes in behaviour. Changes in measured behaviour and other outcomes due to this reactivity may introduce bias in otherwise well-conducted randomised controlled trials (RCTs), yielding incorrect estimates of intervention effects. Despite this, measurement reactivity is not currently adequately considered in risk of bias frameworks. The present research aims to produce a set of guidance statements on how best to avoid or minimise bias due to measurement reactivity in studies of interventions to improve health, with a particular focus on bias in RCTs. METHODS: The MERIT study consists of a series of systematic and rapid reviews, a Delphi study and an expert workshop to develop guidance on how to minimise bias in trials due to measurement reactivity. An existing systematic review on question-behaviour effects on health-related behaviours will be updated and three new rapid reviews will be conducted to identify (1) existing guidance on measurement reactivity; (2) systematic reviews of studies that have quantified the effects of measurement on outcomes relating to behaviour and affective outcomes in health and non-health contexts and (3) trials that have investigated the effects of objective measurements of behaviour on concurrent or subsequent behaviour itself. A Delphi procedure will be used to combine the views of experts with a view to reaching agreement on the scope of the guidance statements. Finally, a workshop will be held in autumn 2018, with the aim of producing a set of guidance statements that will form the central part of new MRC guidance on how best to avoid bias due to measurement reactivity in studies of interventions to improve health. DISCUSSION: Our ambition is to produce MRC guidance on measurement reactions in trials which will be used by future trial researchers, leading to the development of trials that are less likely to be at risk of bias.