Optimizing engagement with Internet-based health behaviour change interventions: Comparison of self-assessment with and without tailored feedback using a mixed methods approach

OBJECTIVES Internet-based health behaviour interventions have variable effects on health-related outcomes. Effectiveness may be improved by optimizing the design of interventions. This study examined the specific effect on engagement of providing two different design features - tailoring and self-assessment. DESIGN Three versions of an Internet-delivered intervention to support the self-care of mild bowel problems were developed that provided (1) self-assessment without tailored feedback, (2) self-assessment with tailored feedback, and (3) generic information only. METHODS A qualitative study explored participants' engagement with each version of the intervention (N = 24). A larger quantitative study systematically compared participants' use of the intervention and self-reported engagement using a partial factorial design (n = 178). RESULTS Findings from the qualitative study suggested that self-assessment without tailored feedback appeared to be less acceptable to participants because it was viewed as offering no personal benefit in the absence of personalized advice. In the quantitative study, self-assessment without tailored feedback was associated with greater dropout than when provided in conjunction with tailored feedback. There were significant group differences in participants' engagement with the intervention and perceptions of the intervention. Self-assessment without tailored feedback was rated as marginally less engaging and was associated with fewer positive perceptions than the generic information condition. CONCLUSIONS The acceptability of self-assessment or monitoring components may be optimized by also providing tailored feedback. Without tailored feedback, these components do not appear to be any more engaging than generic information provision. Statement of contribution What is already known on this subject? Digital interventions can be effective for improving a range of health outcomes and behaviours. There is huge variation in the success of different interventions using different combinations of design features. What does this study add? This study used a systematic experimental design to identify the effect on engagement of providing interactive intervention design features alone and in combination. Intervention engagement was better when self-assessment was provided in conjunction with tailored feedback.