Background: In behavioural trials, experimental interventions are usually compared against active control groups. Systematic reviews ignoring variability in comparator interventions may draw inaccurate conclusions about the effectiveness of interventions. Poor reporting of comparator interventions may obscure this problem and impede development of methods to account for this variability.
Objectives: 1) Develop a reliable method for retrospectively assessing the active ingredients (behaviour change techniques; BCTs) of comparator interventions in smoking cessation trials; 2) Examine the variability and effectiveness of comparator interventions.
Methods: Trial authors (first, second, last) were emailed and asked for materials describing the comparator intervention in their trial. They were also asked to complete a ‘comparator intervention checklist’, which we developed based on guidelines for smoking cessation services, advisory board and PPI input, and scientific literature. Descriptive and mixed-effects meta-regression analyses were used to examine variability and impact of comparator interventions on control group smoking cessation rates.
Results: 67% (96/143) of trial authors sent information on their comparator interventions: 83/96 (86%) completed the checklist (Cronbach’s alpha’s .94) and 55/96 (57%) provided existing materials. In total, 1955 BCTs were identified, of which 75% (1469/1955) could only be identified in the checklist and other materials sent by study authors. There was substantial variability in the comparator intervention dose (the sum of BCTs) between trials (range 0-45). Comparator intervention dose was a strong predictor of smoking cessation rates in the control groups (p