Examining a brief web and longitudinal app-based intervention [Wysa] for mental health support in Singapore during the COVID-19 pandemic: mixed-methods retrospective observational study.

The COVID-19 pandemic in Singapore led to limited access to mental health services, resulting in increased distress among the population. This study explores the potential benefits of offering a digital mental health intervention (DMHI), Wysa, as a brief and longitudinal intervention as part of the mindline.sg initiative launched by the MOH Office for Healthcare Transformation in Singapore. The paper aims to (i) Evaluate the engagement and retention of Singaporean users across the brief intervention on the mindline.sg website and the longitudinal app version of Wysa; (ii) Examine the types of negative thoughts and challenges managed during the pandemic; and (iii) Assess the impact of the conversational agent (CA) in supporting cognitive restructuring across attributional styles and cognitive patterns. A retrospective observational design with a mixed-methods approach was utilized. Website users ( = 69,055) and app users ( = 4,103) from September 1, 2020, to July 25, 2022, were included in the study. Engagement and retention were evaluated through usage data, and T-tests were used to compare engagement and retention between the app and website. A thematic analysis assessed the types of negative thoughts and the success of cognitive restructuring. Logistic regression was used to estimate the likelihood of restructuring based on negative thought type and attributional style. Users who used the longitudinal intervention after first using a brief intervention demonstrated significantly higher engagement and retention ( < 0.001). In user ratings received for mindline.sg ( = 8,959), 83.03% rated the app 3 or higher (out of 5) on helpfulness. 91.6% of the users ( = 862) who attempted cognitive restructuring ( = 790) on the app successfully reframed a thought. A single conversation with Wysa was also significantly associated with the ability to restructure future-oriented negative thoughts ( < 0.001) and internal, stable and global ( < 0.001) negative thoughts, while other attributional styles required more intervention. Psychosocial challenges managed by users during COVID-19 were also documented through negative thoughts mentioned within the CA. The findings demonstrate that brief interventions can facilitate enhanced engagement with DMHIs and that digital interventions can successfully facilitate cognitive restructuring and improve mental health outcomes. The study provides useful inputs to guide the development of DMHIs and improve their effectiveness.