Introduction: Using medication and participating in psychosocial interventions are active processes that often involve complex decision-making. Shared decision making, SDM, provides a model for user and practitioner to cooperatively assess a treatment’s advantages and disadvantages. Decision aid tools adapted to the needs of users have the potential to restructure how people with mental illness and staff work together to arrive at shared decisions about the next steps in treatment or support.

Aims: The objective of this study is to investigate decisional and information needs among users with mental illness as a pre-requisite for the design and development of a decision aid aimed at supporting user participation in SDM.

Methods: Needs and preferences regarding information transfer in SDM were explored through semi-structured focus group interviews. Participants were adults with psychiatric diagnoses and experience of psychiatric services or close relatives to someone with a psychiatric illness.Qualitative content analysis was used to analyze the data. The transcriptions were initially analyzed in accordance with a conventional, inductive approach. A directed content analysis was then utilized, with Elwyn’s three step model for SDM as a theoretical framework from which to further operationalize the categories rendered in the first step of analysis.

Results: The majority of the findings were easily integrated within Elwyn ́s categories. However, some elements which emerged in the data and which are worth noting were not encompassed within Elwyn’s model, such as the wish for information prior to the meeting regarding the time frame and agenda. The importance of heeding the prior knowledge of the user as valuable for the decision process and not solely checking it in order to correct possible misinformation was frequently expressed, as was the value of follow-ups.

Conclusions: Even though Elwyn’s model is constructed as a pedagogic tool to be used by staff, while our focus is directed towards creating a tool for users, the application of the model upon our data was indeed helpful in rendering clearly defined and distinguishable codes from our categories. We therefore consider it to be a suitable model to continue to build upon in the development of a decision aid.