From smoke to sweat: developing and evaluating an accessible virtual coach to support smoking cessation and physical activity

Abstract

Smoking and insufficient physical activity increase the risk of chronic diseases. These behaviors are common, particularly among people in vulnerable positions, and contribute to health inequalities. Supporting smoking cessation and physical activity is therefore important. Moreover, physical activity can help people refrain from smoking, while quitting smoking can facilitate physical activity.

This dissertation focuses on the development and evaluation of “Perfect Fit”, a smartphone-based intervention with a virtual coach (chatbot) that supports smoking cessation and physical activity. The coach provides personalized advice at moments when support is most needed, such as during strong cravings.

A practical approach is presented for actively involving end-users in developing digital interventions. Additionally, a literature review and a virtual reality experiment illustrate how behavior change techniques can be applied in virtual coach interventions. For instance, identity-based approaches appear promising, but their optimal design and application remain unclear. Finally, a mixed-methods evaluation of Perfect Fit (N=99) showed that 36% of participants had quit smoking after the intervention, and average physical activity increased compared to baseline. The intervention was considered usable, though improvements, for example in personalization, are needed. Overall, these findings suggest that virtual coaches could be a scalable and promising addition to public health interventions.