Characterizing the Representation of and Conversation about ASD Treatments on Twitter

Background: The process by which parents of young children with autism spectrum disorder (ASD) make treatment decisions is complex and nuanced, often rooted in information retrieved from the internet, and not necessarily with attention to scientific support for treatment options. Current research suggests that information on ASD that is available on websites lacks adequate reference to scientific support. While websites provide health information, social media platforms including Twitter are now additional forums for the exchange of content and opinions related to health. Research has not yet been conducted regarding the spread of information related to ASD treatment recommendations via Twitter.

Objectives: The objective of this study was to characterize the information networks, as well as the content and quality of such information shared on Twitter, specifically with respect to the dissemination of evidence-based recommendations for ASD treatment for young children. We aimed to characterize the content of linked sources from tweets related to ASD treatment recommendations and to identify opinion leaders in networks of users tweeting about ASD treatment recommendations.

Methods: A search of Twitter messages (tweets) was conducted using the social media analysis tool, NodeXL. All tweets shared during a 1-week period in October 2016, were filtered using keywords related to intervention for young children with ASD (AUTISM or ASD and TREATMENT or INTERVENTION or THERAPY or CURE or CARE or HELP). All sources linked to from the extracted tweets were coded with regard to type of source and content. Those Twitter accounts that were most prominent within this network of shared tweets were identified using NodeXL automated analytics of user networks.

Results: 894 tweets met the search criteria and included a URL. Tweets were excluded if their linked URL did not address an intervention, or if the intervention was specified to be inappropriate for children under age 3. Of the remaining 71 unique URLs, a breakdown of source type was as follows: 59% were an ‘unbiased’ presentation of information about a treatment; 21% provided a personalized testimonial; 9% advertised a therapy product; 4% linked directly to a current event regarding a treatment; 3% linked directly to a peer-reviewed journal article; 2% linked to a crowdsourcing advice page. In terms of the content, 54% of sources described or endorsed a complementary and alternative medicine (see table). Finally, of the top-ten users who were most frequently mentioned by other tweets, 2 user accounts belonged to advocacy organizations (e.g., Autism Speaks), while 4 of these top-ten accounts belonged to self-advocates.

Conclusions: The rationale for this study was to provide evidence of the social media influences that contribute to ASD treatment decisions. Results of this preliminary study indicate that information readily available to parents on Twitter regarding ASD treatments includes a widely varied set of options offered by both stakeholders and providers. However, the significant representation of complementary and alternative medicines that have been identified as lacking scientific evidence suggests that further research is needed to develop methods for using social media to make evidence-based recommendations as accessible as alternative recommendations.