The impact of behavioural screening on intervention outcomes in a randomised, controlled multiple behaviour intervention trial

Abstract

Abstract Background With an increasing research focus on multiple health behaviour change interventions, a methodological issue requiring further investigation is whether or not to employ pre-trial behavioural screening to exclude participants who are achieving a pre-specified level of one or more behaviours. Behavioural screening can be used to direct limited resources to participants most in need of a behaviour change intervention; but may reduce the representativeness of the sample and limit comparability with trials that do not employ pre-trial behavioural screening. Furthermore, the impact of this type of screening on intervention participation and intervention effects is unknown. Methods Data for this study come from the Logan Healthy Living Program, a randomised, controlled telephone counselling lifestyle intervention trial which did not employ behavioural screening prior to randomisation. Screening for physical activity, diet or the combination was simulated using baseline trial data. To examine the impact of behavioural screening on intervention participation (in terms of participant characteristics, intervention dose received and retention), characteristics of participants included an excluded under the various screening scenarios were compared. To examine the impact of behavioural screening on intervention effects, results from the main trial analysis were compared with results obtained from the same analyses performed separately for each of the screened groups. Results Simulated pre-trial behavioural screening impacted minimally on intervention dose received and trial retention rate. Beyond the anticipated effect of reducing baseline levels of the behaviours being screened for, behavioural screening affected baseline levels of behaviours not targeted by screening, and participants' demographic and health-related characteristics. Behavioural screening impacted on intervention effects in ways that were anticipated and positive, but also unexpected and detrimental. Physical activity screening (alone or in combination with diet) resulted in improved intervention effects for physical activity, while fruit and vegetable screening had no impact on intervention effects for these outcomes. All three types of screening impacted detrimentally on intervention effects for behaviours not being targeted by screening. Conclusions Behavioural screening may have desirable and undesirable consequences in the context of multiple behaviour intervention trials, and thus its potential merits and pitfalls should be carefully considered.