Identification of behaviour change techniques and engagement strategies to design a smartphone app to reduce alcohol consumption using a formal consensus method

Background: Digital interventions to reduce excessive alcohol consumption have the potential to have a broader reach and be more cost-effective than traditional brief interventions. However, there is not yet a strong evidence base on their ability to engage users or on their effectiveness. Objective: This study aimed to identify the behaviour change techniques (BCTs) and engagement strategies most worthy of further study by inclusion in a smartphone application (app) to reduce alcohol consumption, using formal expert consensus methods. Methods: The first phase of the study consisted of a Delphi exercise with three rounds. It was conducted with seven international experts in the field of alcohol and/or behaviour change. In the first round, experts identified BCTs most likely to be effective at reducing alcohol consumption and strategies most likely to engage users with an app; these were rated in the second round; and those rated as effective by at least four out of seven participants were ranked in the third round. The rankings were analysed using Kendall’s W coefficient of concordance, which indicates consensus between participants. The second phase consisted of a new, independent group of experts (n=43) ranking the BCTs that were identified in the first phase. The correlation between the rankings of the two groups was assessed using Spearman’s rank correlation coefficient. Results: Twelve BCTs were identified as likely to be effective. There was moderate agreement among the experts over their ranking (W=.465, χ2(11)=35.77, P<.001) and the BCTs receiving the highest mean rankings were self-monitoring, goal-setting, action planning, and feedback in relation to goals. There was a significant correlation between the ranking of the BCTs by the group of experts who identified them and a second independent group of experts (Spearman’s rho=.690, P=.01). Seventeen responses were generated for strategies likely to engage users. There was moderate agreement among experts on the ranking of these engagement strategies ...