Aim:

To improve uptake, safety and usefulness of online mental health communities

Background:

Online mental health communities are dedicated support platforms aimed at helping individuals to share, discuss and solicit information and support related to mental health. They offer 24- hour opportunities for members to feel understood, access important information, make friends, use their own experiences to help others, whilst managing how they present themselves online. Online communities can be an important gateway to offline support as people test out sharing, challenge stigma, and are signposted elsewhere. They can increase access to support where face-to-face support is limited

However, online communities have potential to cause harm. Information shared can be misleading, inaccurate, harmful, or triggering. Hearing about the experiences of others can exacerbate low mood and risk of harms, create dependency and generate unrealistic expectations and greater anxiety about one’s own condition. Normative and reductionist ideas about health may leave some users feeling misunderstood or even bullied, and people in mental distress may be particularly vulnerable to the negative impacts of this behaviour. Members may be worried about how their data may be used by hosts relying on advertising or data exploitation as a funding source. There is an urgent need to understand how online communities impact on mental health outcomes, for whom, and in what context. This will inform how online communities are commissioned and designed to ensure they are safe and useful.

Methods:

Realist informed mixed methods evaluation, with multiple case studies to co-design theory informed best practice guidance tools for online mental health communities. Cases are theoretically- sampled communities. Mixed-methods include: a realist informed review of existing literature and stakeholder theories; qualitative interviews with community members, moderators and hosts; quantitative surveys of community members; corpus-based discourse analysis and natural language processing of “real time” contextualised posts; co-design of best practice tools and implementation strategy.