Using recorded mental health recovery narratives as a resource for others: Narrative Experiences Online (NEON) intervention development

Background: Social media platforms have enabled the sharing of digital narratives about health concerns on a substantial scale. Some health interventions have integrated digital health narratives, for example to give voice to under-represented populations. Narratives describing recovery from health problems have been a specific focus of research, including those presented in “recorded” (e.g. invariant) form. No clinical trial has been conducted of an online intervention providing access to recorded recovery narratives. Objective: To present knowledge produced through the development of the NEON Intervention, an online intervention which recommends items from the NEON Collection of recovery narratives using an automated algorithm. Methods: Knowledge underpinning the NEON Intervention was developed through three methodical knowledge integration activities. Activity 1 synthesised three previous studies to produce the NEON Impact Model characterising the causal chain from accessing recorded recovery narratives to health-related outcomes. Activity 2 developed and implemented curation principles for the NEON Collection, through iterative consultation with the NEON Lived Experience Advisory Panel (LEAP) (n=11) and the curation of a preliminary collection of 100 recovery narratives. Activity 3 identified and implemented harm minimisation strategies for the NEON Intervention through four rounds of expert consultation with the NEON International Advisory Board (n=7) and the NEON LEAP. The NEON Intervention was finalised through two studies with research participants. In research study 1 current mental health service users (n=40) rated the immediate impact of randomly-presented narratives, allowing for the validation of three narrative feedback questions designed to collect information to inform the recommendation algorithm. In research study 2, current mental health service users (n=25) were interviewed about their immediate response to a prototype of the NEON Intervention and trial procedures, used the prototype for 1 month generating usage logs, and were interviewed about usage. This allowed the usability and acceptability of the prototype and associated trial procedures to be evaluated, and refinements to be made. Results: Knowledge integration activity 1 produced the NEON Impact Model which identifies moderators (recipient, context), mechanisms of connection (reflection, comparison, learning, empathy), processes (identification of change from narrative structure/content, internalisation of observed change) and outcomes (helpful, unhelpful). Knowledge integration activity 2 identified 22 curation principles, including a mission to build a large and heterogeneous collection to maximise opportunities for connection. Knowledge integration activity 3 identified seven harm minimisation strategies, including content warnings, proactive and reactive blocking of narratives, and providing resources for self-management of emotional distress. Research study 1 found variation in the impact of narratives on different participants indicating that participant-level feedback on individual narratives is needed to inform a recommender system, and that order of presentation did not predict narrative feedback. Research study 2 identified amendments to online trial procedures and to the NEON Intervention. Participants accessed some narratives multiple times, usage reduced over the four week period, and narrative feedback was provided for 105 (32%) of 330 narrative accesses. Conclusions: Recorded mental health recovery narratives can be integrated into online interventions, and it is feasible to conduct an evaluation of the NEON Intervention in a clinical trial. The mixed-methods design used to develop the NEON Intervention will also guide the extension of NEON to other clinical populations, the design of other online mental health interventions, and the development of narrative-based interventions in mental health. Trial registration: The NEON Intervention will be used in the NEON Trial (ISRCTN11152837), for people with experience of psychosis, the NEON-O Trial (ISRCTN63197153) for people with experience of non-psychosis mental health problems, and the NEON-C trial (ISRCTN76355273) for informal carers.