JMIR Publications

ABSTRACT

Background:

It seems unclear why people living with breast cancer choose to receive medical cannabis-related content from social media. Thus, the extent to which patients access cannabis-related information from social media and the quality of the content on the platforms needs evaluation.

Objective:

We analyzed the factors determining the quality of cannabis-related content for breast cancer care on social media platforms and the characteristics of bad quality cannabis-related content on social media platforms accessed and consumed by Thai breast cancer survivors following its legalization in 2019.

Methods:

A mixed methods study was conducted between November 2020 and May 2022. A breast cancer survivor support group was involved in this study throughout. The group identified content related to medical cannabis by using social media platforms. The content was categorized based on i) content creator(s), ii) platforms, iii) content category, and iv) upload dates. Four researchers used the Quality Evaluation Scoring Tool (QUEST) to assess content quality, with scores ranging from 0 to 28. Content was classified as ‘good’ or ‘bad’, based on expert ratings. The inter-observer reliability of the QUEST was analyzed. Youden index was used as the cutoff point for QUEST. Statistical significance was set at p < 0.05. Fairclough’s Critical Discourse Analysis was undertaken to examine the underlying discourses around bad quality content.

Results:

Sixty-two cannabis-related items were evaluated. Content sources were categorized as news channels (33.9%), government (25.8%), healthcare providers (19.4%), and alternative medicine providers (19.4%). Most content (48.4%) were uploaded to YouTube and other websites (29.7%). Of these, 36.5% (27 contents) were news or generic cannabis advertisements. Some (12%) content had no identifiable date. The inter-observer correlation of QUEST scores was 0.86 (p<0.05). The mean QUEST score was 12.1±7.6. Contents were considered ‘good’ when the experts’ rating was >3. With a QUEST score of 15 as the threshold, the sensitivity and specificity for differentiating between good and bad content quality were 81.2% and 97.5%, respectively. Content creator was the only significant factor between ‘good’ and ‘bad’ quality content. Bad quality content were primarily created by alternative medicine providers and news channels. Two discourses were identified: i) advocacy for the normalization of cannabis use and ii) romanticization of cannabis as a panacea. These discourses overly normalize and romanticize the use of cannabis, with attention on indications of cannabis use, instructions for cannabis use, and promotion of medical cannabis, without discussions on the contraindications and side-effects of cannabis.  

Conclusions:

The varying quality of medical cannabis information on social media accessed and shared by Thai breast cancer survivors is concerning. Given that content creators are the sole predictive factors of good content quality, the findings call the Thai government to consider placing restrictive regulations and control over cannabis content creators. Clinical Trial: NA