Classification of Health-Related Social Media Posts: Evaluation of Post Content-Classifier Models and Analysis of User Demographics

HPV vaccine misinformation on social media: A multi-method qualitative analysis of comments across three platforms

Objective The purpose of this study was to characterize similarities and differences in HPV vaccine misinformation narratives present in the comment sections of top-performing initial creator posts across three social media platforms. Methods A qualitative multi-method design was used to analyze comments collected from social media posts. A sample of 2996 comments were used for thematic analysis (identifying similar themes) and content analysis (identifying differences in comment type, opinion, and misinformation status). Results Misinformation was pervasive in comment sections. Cross-cutting misinformation themes included adverse reactions, unnecessary vaccine, conspiracy theories, and mistrust of authority. The proportion of comments related to these themes varied by platform. Initial creator posts crafted to be perceived as educational or with an anti-vaccine opinion had a higher proportion of misinformation in the comment sections. Facebook had the highest proportion of misinformation comments. Conclusion Differences in the proportion of cross-cutting themes in the comment sections across platforms suggests the need for targeted communication strategies to counter misinformation narratives and support vaccine uptake. Innovation This study is innovative due to its characterization of misinformation themes across three social media platforms using multiple qualitative methods to assess similarities and differences and focusing on conversations occurring within the comment sections.

Social Media Use for Cancer Support Among Young Adults with Cancer

Purpose: Social media can facilitate peer support among young adults with cancer; however, information is needed about what social media are used, by whom, and how to inform resource and intervention recommendations. Methods: In December 2021, we conducted an online survey with 396 young adults with cancer, ages 18-39, with any diagnosis ages 15-39. Participants reported their social media use to connect with other young adults with cancer, including frequency of use, type of support, and affect (positive to negative) when using to connect with cancer peers. Results: Participants were on average 31 years old (SD = 5.2), with an average age of 27 at diagnosis (63.4% male, 62.1% non-Hispanic White). Almost all (97.5%) reported using social media to connect with other young adults with cancer. Many (48.0%) used three or more social media platforms for cancer support, including Facebook (44.4%), YouTube (43.6%), Instagram (43.4%), Snapchat (36.9%), and Twitter (36.9%). Daily use for cancer support was common (32.9%-60.9%) among those who used social media, particularly among those who were younger; are not transgender; live in urban areas; or had brain, gynecologic, or testicular cancers. Across social media platforms, young adults with cancer reported seeking and sharing emotional support (88.9%), informational support (84.1%), and making connections (81.3%). Conclusion: Young adults with cancer use social media to connect with cancer peers for support. Commonly used existing social media (e.g., Facebook, YouTube, Instagram) should be prioritized in interventions to reach young adults who desire more age-appropriate resources to improve their psychosocial health.

Motivational Influences Affecting Middle-Aged and Elderly Users' Participation Intention in Health-Related Social Media

Social media provide users with multi-directional dialogue for creating and sharing health information that can effectively promote the self-management of health. In regard to the 'greying' trend in social media, most researchers have studied the health-related social media (HRSM) acceptance status and use behavior of middle-aged and elderly people, and have explored the role of HRSM in this group. However, the continuous participation of users is the key to the successful operation of HRSM, and is an essential prerequisite for the subsequent HRSM behavior habits of middle-aged and elderly people. Therefore, we aimed to explore what motivations drive the first use of HRSM among middle-aged and older adults, and the impact of their perception of HRSM, after personal use, on their intention to use it continually. In the study, we used the partial least squares structural equation model (PLS-SEM) to analyze data collected from online questionnaires. The results showed that a self-protection motivation and a social motivation promoted the initial participation of middle-aged and elderly individuals. In addition, these people experienced deeper levels of perceived usefulness and perceived entertainment after their initial participation. The results also revealed that these two perceptions could positively influence middle-aged and elderly individuals' intention to continue with their participation. Our findings should help service platforms to better understand the needs of middle-aged and elderly users. This would help researchers and practitioners to gain a more complete understanding of the motivation of middle-aged and elderly people for participating in HRSM, and the related impacts this may have.

An Exploration of e-Cigarette-Related Search Items on YouTube: Network Analysis

Background

e-Cigarette use among youth is high, which may be due in part to pro–e-cigarette content on social media such as YouTube. YouTube is also a valuable resource for learning about e-cigarette use, trends, marketing, and e-cigarette user perceptions. However, there is a lack of understanding on how similar e-cigarette–related search items result in similar or relatively mutually exclusive search results. This study uses novel methods to evaluate the relationship between e-cigarette–related search items and results.

Objective

The aim of this study is to apply network modeling and rule-based classification to characterize the relationships between e-cigarette–related search items on YouTube and gauge the level of importance of each search item as part of an e-cigarette information network on YouTube.

Methods

We used 16 fictitious YouTube profiles to retrieve 4201 distinct videos from 18 keywords related to e-cigarettes. We used network modeling to represent the relationships between the search items. Moreover, we developed a rule-based classification approach to classify videos. We used betweenness centrality (BC) and correlations between nodes (ie, search items) to help us gain knowledge of the underlying structure of the information network.

Results

By modeling search items and videos as a network, we observed that broad search items such as e-cig had the most connections to other search items, and specific search items such as cigalike had the least connections. Search items with similar words (eg, vape and vaping) and search items with similar meaning (eg, e-liquid and e-juice) yielded a high degree of connectedness. We also found that each node had 18 (SD 34.8) connections (common videos) on average. BC indicated that general search items such as electronic cigarette and vaping had high importance in the network (BC=0.00836). Our rule-based classification sorted videos into four categories: e-cigarette devices (34%-57%), cannabis vaping (16%-28%), e-liquid (14%-37%), and other (8%-22%).

Conclusions

Our findings indicate that search items on YouTube have unique relationships that vary in strength and importance. Our methods can not only be used to successfully identify the important, overlapping, and unique e-cigarette–related search items but also help determine which search items are more likely to act as a gateway to e-cigarette–related content.

Human Papillomavirus Vaccine Messaging on TikTok: Social Media Content Analysis

BACKGROUND

The human papillomavirus (HPV) vaccine is viewed as a critical tool to protect against six HPV-related cancers. Vaccination is recommended from early adolescence through age 26 years. As young people have become increasingly involved in personal health-related decisions, there is a need to tailor HPV vaccine messaging and reach this priority population on social media and digital outlets. TikTok is a growing social media platform with approximately 70% of its users between the ages of 13 and 24 years.

PURPOSE

The aim of this study was to understand HPV vaccine messaging and interactions on TikTok as a needed first step to identifying effective strategies to reach young people with important health messaging.

METHODS

Content analysis was performed on 170 top TikToks focused on the HPV vaccine. TikToks were assessed for content, classification type, and number of interactions.

RESULTS

Most TikToks were provaccine, while antivaccine TikToks had more user interactions. Cancer and prevention were the main content areas of the analyzed provaccine TikToks, while the side effects were the primary focus of antivaccine messages. Approximately 30% of all top TikToks analyzed were developed by health professionals. TikToks without an explicit vaccine opinion primarily described personal experiences and mentioned side effects most often.

IMPLICATIONS

TikTok is a growing social media platform that can be used to reach young people and encourage HPV vaccine uptake. Health professionals need to consider the interest that users have in personal experiences and address antivaccine narratives related to side effects.

Recommendations From the Twitter Hashtag #DoctorsAreDickheads: Qualitative Analysis

BACKGROUND

The social media site Twitter has 145 million daily active users worldwide and has become a popular forum for users to communicate their health care concerns and experiences as patients. In the fall of 2018, a hashtag titled #DoctorsAreDickheads emerged, with almost 40,000 posts calling attention to health care experiences.

OBJECTIVE

This study aims to identify common health care conditions and conceptual themes represented within the phenomenon of this viral Twitter hashtag.

METHODS

We analyzed a random sample of 5.67% (500/8818) available tweets for qualitative analysis between October 15 and December 31, 2018, when the hashtag was the most active. Team coders reviewed the same 20.0% (100/500) tweets and the remainder individually. We abstracted the user's health care role and clinical conditions from the tweet and user profile, and used phenomenological content analysis to identify prevalent conceptual themes through sequential open coding, memoing, and discussion of concepts until an agreement was reached.

RESULTS

Our final sample comprised 491 tweets and unique Twitter users. Of this sample, 50.5% (248/491) were from patients or patient advocates, 9.6% (47/491) from health care professionals, 4.3% (21/491) from caregivers, 3.7% (18/491) from academics or researchers, 1.0% (5/491) from journalists or media, and 31.6% (155/491) from non-health care individuals or other. The most commonly mentioned clinical conditions were chronic pain, mental health, and musculoskeletal conditions (mainly Ehlers-Danlos syndrome). We identified 3 major themes: disbelief in patients' experience and knowledge that contributes to medical errors and harm, the power inequity between patients and providers, and metacommentary on the meaning and impact of the #DoctorsAreDickheads hashtag.

CONCLUSIONS

People publicly disclose personal and often troubling health care experiences on Twitter. This adds new accountability for the patient-provider interaction, highlights how harmful communication affects diagnostic safety, and shapes the public's viewpoint of how clinicians behave. Hashtags such as this offer valuable opportunities to learn from patient experiences. Recommendations include developing best practices for providers to improve communication, supporting patients through challenging diagnoses, and promoting patient engagement.

YouTube Video Comments on Healthy Eating: Descriptive and Predictive Analysis

Background

Poor nutrition and food selection lead to health issues such as obesity, cardiovascular disease, diabetes, and cancer. This study of YouTube comments aims to uncover patterns of food choices and the factors driving them, in addition to exploring the sentiments of healthy eating in networked communities.

Objective

The objectives of the study are to explore the determinants, motives, and barriers to healthy eating behaviors in online communities and provide insight into YouTube video commenters’ perceptions and sentiments of healthy eating through text mining techniques.

Methods

This paper applied text mining techniques to identify and categorize meaningful healthy eating determinants. These determinants were then incorporated into hypothetically defined constructs that reflect their thematic and sentimental nature in order to test our proposed model using a variance-based structural equation modeling procedure.

Results

With a dataset of 4654 comments extracted from YouTube videos in the context of Malaysia, we apply a text mining method to analyze the perceptions and behavior of healthy eating. There were 10 clusters identified with regard to food ingredients, food price, food choice, food portion, well-being, cooking, and culture in the concept of healthy eating. The structural equation modeling results show that clusters are positively associated with healthy eating with all P values less than .001, indicating a statistical significance of the study results. People hold complex and multifaceted beliefs about healthy eating in the context of YouTube videos. Fruits and vegetables are the epitome of healthy foods. Despite having a favorable perception of healthy eating, people may not purchase commonly recognized healthy food if it has a premium price. People associate healthy eating with weight concerns. Food taste, variety, and availability are identified as reasons why Malaysians cannot act on eating healthily.

Conclusions

This study offers significant value to the existing literature of health-related studies by investigating the rich and diverse social media data gleaned from YouTube. This research integrated text mining analytics with predictive modeling techniques to identify thematic constructs and analyze the sentiments of healthy eating.

Using Twitter to Surveil the Opioid Epidemic in North Carolina: An Exploratory Study

BACKGROUND

Over the last two decades, deaths associated with opioids have escalated in number and geographic spread, impacting more and more individuals, families, and communities. Reflecting on the shifting nature of the opioid overdose crisis, Dasgupta, Beletsky, and Ciccarone offer a triphasic framework to explain that opioid overdose deaths (OODs) shifted from prescription opioids for pain (beginning in 2000), to heroin (2010 to 2015), and then to synthetic opioids (beginning in 2013). Given the rapidly shifting nature of OODs, timelier surveillance data are critical to inform strategies that combat the opioid crisis. Using easily accessible and near real-time social media data to improve public health surveillance efforts related to the opioid crisis is a promising area of research.

OBJECTIVE

This study explored the potential of using Twitter data to monitor the opioid epidemic. Specifically, this study investigated the extent to which the content of opioid-related tweets corresponds with the triphasic nature of the opioid crisis and correlates with OODs in North Carolina between 2009 and 2017.

METHODS

Opioid-related Twitter posts were obtained using Crimson Hexagon, and were classified as relating to prescription opioids, heroin, and synthetic opioids using natural language processing. This process resulted in a corpus of 100,777 posts consisting of tweets, retweets, mentions, and replies. Using a random sample of 10,000 posts from the corpus, we identified opioid-related terms by analyzing word frequency for each year. OODs were obtained from the Multiple Cause of Death database from the Centers for Disease Control and Prevention Wide-ranging Online Data for Epidemiologic Research (CDC WONDER). Least squares regression and Granger tests compared patterns of opioid-related posts with OODs.

RESULTS

The pattern of tweets related to prescription opioids, heroin, and synthetic opioids resembled the triphasic nature of OODs. For prescription opioids, tweet counts and OODs were statistically unrelated. Tweets mentioning heroin and synthetic opioids were significantly associated with heroin OODs and synthetic OODs in the same year (P=.01 and P<.001, respectively), as well as in the following year (P=.03 and P=.01, respectively). Moreover, heroin tweets in a given year predicted heroin deaths better than lagged heroin OODs alone (P=.03).

CONCLUSIONS

Findings support using Twitter data as a timely indicator of opioid overdose mortality, especially for heroin.