OBJECTIVES: Mining of social media data is increasingly popular, and current evolving analytical approaches often include largely quantitative analyses (e.g., natural language processing) of the substantial amount of available information. This study aimed to explore the use of methodical and structured qualitative open-coding techniques to analyze social media data. The case example was designed to understand patients’ perceptions and preferences for biologic therapies for severe asthma.

METHODS: Two publicly accessible drug review sites and one forum with asthma-specific posts were mined, extracting 517 posts. Two researchers independently analyzed the posts using open-coding to inductively categorize key concepts related to biologic therapies for asthma, and deductively code information on key attributes of biologic therapies. Concepts associated with biologic therapy for asthma were coded thematically, as well as positively or negatively, when appropriate.

RESULTS: The qualitatively coded data could be easily summarized quantitatively, including the most commonly mentioned positive impacts of biologic therapy (efficacy-related impacts: n=169; reduced need for other treatments: n=105) and negative impacts (access difficulty: n=60; financial difficulty: n=60). The most commonly mentioned biologic therapy attributes were medication efficacy (n=212), dosing frequency (n=77), route of administration (n=73) and side effects (n=70). The qualitative open-coding approach resulted efficient organization and synthesis of these data to provide greater depth of understanding (e.g., reasons for perceived infusion burdens/benefits) and assessment of variability within concepts (e.g., perceptions of time to onset of action of biologic therapy).

CONCLUSIONS: The qualitative approach to analysis of social media data is a rich, innovative and powerful means for rapidly and cost-effectively yielding insights from a large number of patients. Specifically, open-coding of qualitative social media data yields a more precise understanding of social media data that can complement current quantitative approaches.