Diabetes topics associated with engagement on Twitter

Examining the Effectiveness of Social Media for the Dissemination of Research Evidence for Health and Social Care Practitioners: Systematic Review and Meta-Analysis

BACKGROUND

Social media use has potential to facilitate the rapid dissemination of research evidence to busy health and social care practitioners.

OBJECTIVE

This study aims to quantitatively synthesize evidence of the between- and within-group effectiveness of social media for dissemination of research evidence to health and social care practitioners. It also compared effectiveness between different social media platforms, formats, and strategies.

METHODS

We searched electronic databases for articles in English that were published between January 1, 2010, and January 10, 2023, and that evaluated social media interventions for disseminating research evidence to qualified, postregistration health and social care practitioners in measures of reach, engagement, direct dissemination, or impact. Screening, data extraction, and risk of bias assessments were carried out by at least 2 independent reviewers. Meta-analyses of standardized pooled effects were carried out for between- and within-group effectiveness of social media and comparisons between platforms, formats, and strategies. Certainty of evidence for outcomes was assessed using the GRADE (Grading of Recommendations, Assessment, Development, and Evaluations) framework.

RESULTS

In total, 50 mixed-quality articles that were heterogeneous in design and outcome were included (n=9, 18% were randomized controlled trials [RCTs]). Reach (measured in number of practitioners, impressions, or post views) was reported in 26 studies. Engagement (measured in likes or post interactions) was evaluated in 21 studies. Direct dissemination (measured in link clicks, article views, downloads, or altmetric attention score) was analyzed in 23 studies (8 RCTs). Impact (measured in citations or measures of thinking and practice) was reported in 13 studies. Included studies almost universally indicated effects in favor of social media interventions, although effect sizes varied. Cumulative evidence indicated moderate certainty of large and moderate between-group effects of social media interventions on direct dissemination (standardized mean difference [SMD] 0.88; P=.02) and impact (SMD 0.76; P<.001). After social media interventions, cumulative evidence showed moderate certainty of large within-group effects on reach (SMD 1.99; P<.001), engagement (SMD 3.74; P<.001), and direct dissemination (SMD 0.82; P=.004) and low certainty of a small within-group effect on impacting thinking or practice (SMD 0.45; P=.02). There was also evidence for the effectiveness of using multiple social media platforms (including Twitter, subsequently rebranded X; and Facebook), images (particularly infographics), and intensive social media strategies with frequent, daily posts and involving influential others. No included studies tested the dissemination of research evidence to social care practitioners.

CONCLUSIONS

Social media was effective for disseminating research evidence to health care practitioners. More intense social media campaigns using specific platforms, formats, and strategies may be more effective than less intense interventions. Implications include recommendations for effective dissemination of research evidence to health care practitioners and further RCTs in this field, particularly investigating the dissemination of social care research.

TRIAL REGISTRATION

PROSPERO International Prospective Register of Systematic Reviews CRD42022378793; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=378793.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID)

RR2-10.2196/45684.

Applying machine-learning to rapidly analyze large qualitative text datasets to inform the COVID-19 pandemic response: comparing human and machine-assisted topic analysis techniques

Introduction

Machine-assisted topic analysis (MATA) uses artificial intelligence methods to help qualitative researchers analyze large datasets. This is useful for researchers to rapidly update healthcare interventions during changing healthcare contexts, such as a pandemic. We examined the potential to support healthcare interventions by comparing MATA with "human-only" thematic analysis techniques on the same dataset (1,472 user responses from a COVID-19 behavioral intervention).

Methods

In MATA, an unsupervised topic-modeling approach identified latent topics in the text, from which researchers identified broad themes. In human-only codebook analysis, researchers developed an initial codebook based on previous research that was applied to the dataset by the team, who met regularly to discuss and refine the codes. Formal triangulation using a "convergence coding matrix" compared findings between methods, categorizing them as "agreement", "complementary", "dissonant", or "silent".

Results

Human analysis took much longer than MATA (147.5 vs. 40 h). Both methods identified key themes about what users found helpful and unhelpful. Formal triangulation showed both sets of findings were highly similar. The formal triangulation showed high similarity between the findings. All MATA codes were classified as in agreement or complementary to the human themes. When findings differed slightly, this was due to human researcher interpretations or nuance from human-only analysis.

Discussion

Results produced by MATA were similar to human-only thematic analysis, with substantial time savings. For simple analyses that do not require an in-depth or subtle understanding of the data, MATA is a useful tool that can support qualitative researchers to interpret and analyze large datasets quickly. This approach can support intervention development and implementation, such as enabling rapid optimization during public health emergencies.

Response Consistency of Crowdsourced Web-Based Surveys on Type 1 Diabetes

Although Amazon Mechanical Turk facilitates the quick surveying of a large sample from various demographic and socioeconomic backgrounds, it may not be an optimal platform for obtaining reliable diabetes-related information from the online type 1 diabetes population.

Twitter-based crowdsourcing: What kind of measures can help to end the COVID-19 pandemic faster?

Background

Crowdsourcing is a low-cost, adaptable, and innovative method to collect ideas from numerous contributors with diverse backgrounds. Crowdsourcing from social media like Twitter can be used for generating ideas in a noticeably brief time based on contributions from globally distributed users. The world has been challenged by the COVID-19 pandemic in the last several years. Measures to combat the pandemic continue to evolve worldwide, and ideas and opinions on optimal counteraction strategies are of high interest.

Objective

This study aimed to validate the use of Twitter as a crowdsourcing platform in order to gain an understanding of public opinion on what measures can help to end the COVID-19 pandemic faster.

Methods

This cross-sectional study was conducted during the period from December 22, 2021, to February 4, 2022. Tweets were posted by accounts operated by the authors, asking “How to faster end the COVID-19 pandemic?” and encouraging the viewers to comment on measures that they perceive would be effective to achieve this goal. The ideas from the users' comments were collected and categorized into two major themes – personal and institutional measures. In the final stage of the campaign, a Twitter poll was conducted to get additional comments and to estimate which of the two groups of measures were perceived to be important amongst Twitter users.

Results

The crowdsourcing campaign generated seventeen suggested measures categorized into two major themes (personal and institutional) that received a total of 1,727 endorsements (supporting comments, retweets, and likes). The poll received a total of 325 votes with 58% of votes underscoring the importance of both personal and institutional measures, 20% favoring personal measures, 11% favoring institutional measures, and 11% of the votes given just out of curiosity to see the vote results.

Conclusions

Twitter was utilized successfully for crowdsourcing ideas on strategies how to end the COVID-19 pandemic faster. The results indicate that the Twitter community highly values the significance of both personal responsibility and institutional measures to counteract the pandemic. This study validates the use of Twitter as a primary tool that could be used for crowdsourcing ideas with healthcare significance.

Identifying Patients With Inflammatory Bowel Disease on Twitter and Learning From Their Personal Experience: Retrospective Cohort Study

Background

Patients use social media as an alternative information source, where they share information and provide social support. Although large amounts of health-related data are posted on Twitter and other social networking platforms each day, research using social media data to understand chronic conditions and patients’ lifestyles is limited.

Objective

In this study, we contributed to closing this gap by providing a framework for identifying patients with inflammatory bowel disease (IBD) on Twitter and learning from their personal experiences. We enabled the analysis of patients’ tweets by building a classifier of Twitter users that distinguishes patients from other entities. This study aimed to uncover the potential of using Twitter data to promote the well-being of patients with IBD by relying on the wisdom of the crowd to identify healthy lifestyles. We sought to leverage posts describing patients’ daily activities and their influence on their well-being to characterize lifestyle-related treatments.

Methods

In the first stage of the study, a machine learning method combining social network analysis and natural language processing was used to automatically classify users as patients or not. We considered 3 types of features: the user’s behavior on Twitter, the content of the user’s tweets, and the social structure of the user’s network. We compared the performances of several classification algorithms within 2 classification approaches. One classified each tweet and deduced the user’s class from their tweet-level classification. The other aggregated tweet-level features to user-level features and classified the users themselves. Different classification algorithms were examined and compared using 4 measures: precision, recall, F1 score, and the area under the receiver operating characteristic curve. In the second stage, a classifier from the first stage was used to collect patients' tweets describing the different lifestyles patients adopt to deal with their disease. Using IBM Watson Service for entity sentiment analysis, we calculated the average sentiment of 420 lifestyle-related words that patients with IBD use when describing their daily routine.

Results

Both classification approaches showed promising results. Although the precision rates were slightly higher for the tweet-level approach, the recall and area under the receiver operating characteristic curve of the user-level approach were significantly better. Sentiment analysis of tweets written by patients with IBD identified frequently mentioned lifestyles and their influence on patients’ well-being. The findings reinforced what is known about suitable nutrition for IBD as several foods known to cause inflammation were pointed out in negative sentiment, whereas relaxing activities and anti-inflammatory foods surfaced in a positive context.

Conclusions

This study suggests a pipeline for identifying patients with IBD on Twitter and collecting their tweets to analyze the experimental knowledge they share. These methods can be adapted to other diseases and enhance medical research on chronic conditions.

'Falsehood flies, and the truth comes limping after it': social media and public health

PURPOSE OF REVIEW

To highlight the various uses of social media by public health practitioners and organizations, with special emphasis on how social media has been successfully applied and where applications have struggled to achieve the desired effects.

RECENT FINDINGS

Social media has been used effectively in improving the timeliness and accuracy of public health surveillance. Social media has also been used to communicate information between public health organizations and reinforce consistent messaging about enduring threats to public health. It has been applied with some success to coordinate of disaster response and for keeping the public informed during other emergency situations. However, social media has also been weaponized against the public health community to spread disinformation and misinformation, and the public health community has yet to devise a successful strategy to mitigate this destructive use of social media.

SUMMARY

Social media can be an effective tool for public health practitioners and organizations who seek to disseminate information on a daily basis, rapidly convey information in emergent situations, and battle misinformation. Social media has been uniquely valuable and distinctly destructive when it comes to protecting and improving public health.

Use of social media data for disease based social network analysis and network modeling: A Systematic Review

Burden due to infectious and noncommunicable disease is increasing at an alarming rate. Social media usage is growing rapidly and has become the new norm of communication. It is imperative to examine what is being discussed in the social media about diseases or conditions and the characteristics of the network of people involved in discussion. The objective is to assess the tools and techniques used to study social media disease networks using network analysis and network modeling. PubMed and IEEEXplore were searched from 2009 to 2020 and included 30 studies after screening and analysis. Twitter, QuitNet, and disease-specific online forums were widely used to study communications on various health conditions. Most of the studies have performed content analysis and network analysis, whereas network modeling has been done in six studies. Posts on cancer, COVID-19, and smoking have been widely studied. Tools and techniques used for network analysis are listed. Health-related social media data can be leveraged for network analysis. Network modeling technique would help to identify the structural factors associated with the affiliation of the disease networks, which is scarcely utilized. This will help public health professionals to tailor targeted interventions.

Telemedicine for Acute Conditions During COVID-19: A Nationwide Survey Using Crowdsourcing

Background: COVID-19 has resulted in a rapid and significant adoption of telemedicine for acute conditions. Understanding whether patient demand will last after the pandemic helps providers and payers make informed decisions about whether to continue adopting telemedicine. Objective: We examine user experience as well as process and patient outcomes of using telemedicine for acute conditions during COVID-19 and assess how patient outcomes are affected by waiting times and demographics. Materials and Methods: A survey was conducted via Amazon Mechanical Turk during June 17-29, 2020. Inclusion criteria were: (1) ≥18 years old, (2) residing in the United States, (3) used telemedicine for acute conditions after January, and 4) a human intelligence task approval rate of >95%. Process outcomes included patient waiting time with patient outcomes being satisfaction and future use intention. Bivariate analysis and regressions of the data were performed. Results: On average, respondents reported appointment wait time of 2.76 days and virtual office wait time of 19.44 min. Overall, respondents reported moderate satisfaction (mean 5.08-5.35 of 7) and future use intention (mean 5.10-5.32 of 7). Over 72% of the respondents were satisfied and had future use intention. Females, heavier internet users, and those on the higher/lower ends of the education spectrum reported better patient outcomes. Patients "visiting" a doctor experiencing eye problems, vis-à-vis other ailments, reported lower satisfaction and intention. Waiting time negatively associates with satisfaction. Conclusions: Given the satisfactory outcomes, the high demand for telemedicine may continue after the COVID-19 pandemic. However, whether providers will continue to offer telemedicine visits may require more evidence.

Factors Engaging Users of Diabetes Social Media Channels on Facebook, Twitter, and Instagram: Observational Study

Background

Diabetes patient associations and diabetes-specific patient groups around the world are present on social media. Although active participation and engagement in these diabetes social media groups has been mostly linked to positive effects, very little is known about the content that is shared on these channels or the post features that engage their users the most.

Objective

The objective of this study was to analyze (1) the content and features of posts shared over a 3-year period on 3 diabetes social media channels (Facebook, Twitter, and Instagram) of a diabetes association, and (2) users’ engagement with these posts (likes, comments, and shares).

Methods

All social media posts published from the Norwegian Diabetes Association between January 1, 2017, and December 31, 2019, were extracted. Two independent reviewers classified the posts into 7 categories based on their content. The interrater reliability was calculated using Cohen kappa. Regression analyses were carried out to analyze the effects of content topic, social media channel, and post features on users’ engagement (likes, comments, and shares).

Results

A total of 1449 messages were posted. Posts of interviews and personal stories received 111% more likes, 106% more comments, and 112% more shares than miscellaneous posts (all P<.001). Messages posted about awareness days and other celebrations were 41% more likely to receive likes than miscellaneous posts (P<.001). Conversely, posts on research and innovation received 31% less likes (P<.001), 35% less comments (P=.02), and 25% less shares (P=.03) than miscellaneous posts. Health education posts received 38% less comments (P=.003) but were shared 39% more than miscellaneous posts (P=.007). With regard to social media channel, Facebook and Instagram posts were both 35 times more likely than Twitter posts to receive likes, and 60 times and almost 10 times more likely to receive comments, respectively (P<.001). Compared to text-only posts, those with videos had 3 times greater chance of receiving likes, almost 4 times greater chance of receiving comments, and 2.5 times greater chance of being shared (all P<.001). Including both videos and emoji in posts increased the chances of receiving likes by almost 7 times (P<.001). Adding an emoji to posts increased their chances of receiving likes and being shared by 71% and 144%, respectively (P<.001).

Conclusions

Diabetes social media users seem to be least engaged in posts with content topics that a priori could be linked to greater empowerment: research and innovation on diabetes, and health education. Diabetes social media groups, public health authorities, and other stakeholders interested in sharing research and innovation content and promoting health education on social media should consider including videos and emoji in their posts, and publish on popular and visual-based social media channels, such as Facebook and Instagram, to increase user engagement.

International Registered Report Identifier (IRRID)

RR2-10.1186/s12913-018-3178-7

Using photos for public health communication: A computational analysis of the Centers for Disease Control and Prevention Instagram photos and public responses

This study aims to explore the use of Instagram by the Centers for Disease Control and Prevention, one of the representative public health authorities in the United States. For this aim, all of the photos uploaded on the Centers for Disease Control and Prevention Instagram account were crawled and the content of them were analyzed using Microsoft Azure Cognitive Services. Also, engagement was measured by the sum of numbers of likes and comments to each photo, and sentiment analysis of comments was conducted. Results suggest that the photos that can be categorized into "text" and "people" took the largest share in the Centers for Disease Control and Prevention Instagram photos. And it was found that the Centers for Disease Control and Prevention's major way of delivering messages on Instagram was to imprint key messages that call for actions for better health on photos and to provide the source of complementary information on text component of each post. It was also found that photos with more and bigger human faces had lower level of engagement than the others, and happiness and neutral emotions expressed on the faces in photos were negatively associated with engagement. The features whose high value would make the photos look splendid and gaudy were negatively correlated with engagement, but sharpness was positively correlated.

Tweeting about public health policy: Social media response to the UK Government's announcement of a Parliamentary vote on draft standardised packaging regulations

BACKGROUND

Standardised tobacco packaging has been, and remains, a contentious policy globally, attracting corporate, public health, political, media and popular attention. In January 2015, the UK Government announced it would vote on draft regulations for the policy before the May 2015 General Election. We explored reactions to the announcement on Twitter, in comparison with an earlier period of little UK Government activity on standardised packaging.

METHODS

We obtained a random sample of 1038 tweets in two 4-week periods, before and after the UK Government's announcement. Content analysis was used to examine the following Tweet characteristics: support for the policy, purpose, Twitter-user's geographical location and affiliation, and evidence citation and quality. Chi-squared analyses were used to compare Tweet characteristics between the two periods.

RESULTS

Overall, significantly more sampled Tweets were in favour of the policy (49%) in comparison to those opposed (19%). Yet, at Time 2, following the announcement, a greater proportion of sampled tweets opposed standardised packaging compared to the period sampled at Time 1, prior to the announcement (p<0.001). The quality of evidence and research cited in URLs linked at Time 2 was significantly lower than at Time 1 (p<0.001), with peer-reviewed research more likely to be shared in positive Tweets (p<0.001) and in Tweets linking to URLs originating from the health sector (p<0.001). The decline in the proportion of positive Tweets was mirrored by a reduction in Tweets by health sector Twitter-users at Time 2 (p<0.001).

CONCLUSIONS

Microblogging sites can reflect offline policy debates and are used differently by policy proponents and opponents dependent on the policy context. Twitter-users opposed to standardised packaging increased their activity following the Government's announcement, while those in support broadly maintained their rate of Twitter engagement. The findings offer insight into the public health community's options for using Twitter to influence policy and disseminate research. In particular, proliferation of Twitter activity following pro-public health policy announcements could be considered to ensure pro-health messages are not overshadowed by anti-regulation voices.

Text–Image Relationships in Tweets: Shaping the Meanings of an Epidemic

1. Background: While many studies analyze the functions that images can fulfill during humanitarian crises or catastrophes, an understanding of how meaning is constructed in text–image relationships is lacking. This article explores how discourses are produced using different types of text–image interactions. It presents a case study focusing on a humanitarian crisis, more specifically the sexual transmission of Ebola. 2. Methods: Data were processed both quantitatively and qualitatively through a keyword-based selection. Tweets containing an image were retrieved from a database of 210,600 tweets containing the words “Ebola” and “semen”, in English and in French, over the course of 12 months. When this first selection was crossed with the imperative of focusing on a specific thematic (the sexual transmission of Ebola) and avoiding off-topic text–image relationships, it led to reducing the corpus to 182 tweets. 3. Results: The article proposes a four-category classification of text–image relationships. Theoretically, it provides original insights into how discourses are built in social media; it also highlights the semiotic significance of images in expressing an opinion or an emotion. 4. Conclusion: The results suggest that the process of signification needs to be rethought: Content enhancement and dialogism through images have a bearing on Twitter’s use as a public sphere, such as credibilization of discourses or politicization of events. This opens the way to a new, more comprehensive approach to the rhetorics of users on Twitter.

Information Circulation in times of Ebola: Twitter and the Sexual Transmission of Ebola by Survivors

INTRODUCTION

The 2013-2015 outbreak of Ebola was by far the largest to date, affecting Guinea, Liberia, Sierra Leone, and secondarily, Nigeria, Senegal and the United States. Such an event raises questions about the circulation of health information across social networks. This article presents an analysis of tweets concerning a specific theme: the sexual transmission of the virus by survivors, at a time when there was a great uncertainty about the duration and even the possibility of such transmission.

METHODS

This article combines quantitative and qualitative analysis. From a sample of 50,000 tweets containing the words "Ebola" in French and English, posted between March 15 and November 8, 2014, we created a graphic representation of the number of tweets over time, and identified two peaks: the first between July 27 and August 16, 2014 (633 tweets) and the second between September 28 and November 8, 2014 (2,577 tweets). This sample was divided into two parts, and every accessible publication was analyzed and coded according to the authors' objectives, feelings expressed and/or publication type.

RESULTS

While the results confirm the significant role played by mainstream media in disseminating information, media did not create the debate around the sexual transmission of Ebola and Twitter does not fully reflect mainstream media contents. Social media rather work like a "filter": in the case of Ebola, Twitter preceded and amplified the debate with focusing more than the mainstream media on the sexual transmission, as expressed in jokes, questions and criticism.

DISCUSSION

Online debates can of course feed on journalistic or official information, but they also show great autonomy, tinged with emotions or criticisms. Although numerous studies have shown how this can lead to rumors and disinformation, our research suggests that this relative autonomy makes it possible for Twitter users to bring into the public sphere some types of information that have not been widely addressed. Our results encourage further research to understand how this "filter" works during health crises, with the potential to help public health authorities to adjust official communications accordingly. Without a doubt, the health authorities would be well advised to put in place a special watch on the comments circulating on social media (in addition to that used by the health monitoring agencies).

Mapping of Crowdsourcing in Health: Systematic Review

Background

Crowdsourcing involves obtaining ideas, needed services, or content by soliciting Web-based contributions from a crowd. The 4 types of crowdsourced tasks (problem solving, data processing, surveillance or monitoring, and surveying) can be applied in the 3 categories of health (promotion, research, and care).

Objective

This study aimed to map the different applications of crowdsourcing in health to assess the fields of health that are using crowdsourcing and the crowdsourced tasks used. We also describe the logistics of crowdsourcing and the characteristics of crowd workers.

Methods

MEDLINE, EMBASE, and ClinicalTrials.gov were searched for available reports from inception to March 30, 2016, with no restriction on language or publication status.

Results

We identified 202 relevant studies that used crowdsourcing, including 9 randomized controlled trials, of which only one had posted results at ClinicalTrials.gov. Crowdsourcing was used in health promotion (91/202, 45.0%), research (73/202, 36.1%), and care (38/202, 18.8%). The 4 most frequent areas of application were public health (67/202, 33.2%), psychiatry (32/202, 15.8%), surgery (22/202, 10.9%), and oncology (14/202, 6.9%). Half of the reports (99/202, 49.0%) referred to data processing, 34.6% (70/202) referred to surveying, 10.4% (21/202) referred to surveillance or monitoring, and 5.9% (12/202) referred to problem-solving. Labor market platforms (eg, Amazon Mechanical Turk) were used in most studies (190/202, 94%). The crowd workers’ characteristics were poorly reported, and crowdsourcing logistics were missing from two-thirds of the reports. When reported, the median size of the crowd was 424 (first and third quartiles: 167-802); crowd workers’ median age was 34 years (32-36). Crowd workers were mainly recruited nationally, particularly in the United States. For many studies (58.9%, 119/202), previous experience in crowdsourcing was required, and passing a qualification test or training was seldom needed (11.9% of studies; 24/202). For half of the studies, monetary incentives were mentioned, with mainly less than US $1 to perform the task. The time needed to perform the task was mostly less than 10 min (58.9% of studies; 119/202). Data quality validation was used in 54/202 studies (26.7%), mainly by attention check questions or by replicating the task with several crowd workers.

Conclusions

The use of crowdsourcing, which allows access to a large pool of participants as well as saving time in data collection, lowering costs, and speeding up innovations, is increasing in health promotion, research, and care. However, the description of crowdsourcing logistics and crowd workers’ characteristics is frequently missing in study reports and needs to be precisely reported to better interpret the study findings and replicate them.

Comparing Amazon's Mechanical Turk Platform to Conventional Data Collection Methods in the Health and Medical Research Literature

BACKGROUND

The goal of this article is to conduct an assessment of the peer-reviewed primary literature with study objectives to analyze Amazon.com 's Mechanical Turk (MTurk) as a research tool in a health services research and medical context.

METHODS

Searches of Google Scholar and PubMed databases were conducted in February 2017. We screened article titles and abstracts to identify relevant articles that compare data from MTurk samples in a health and medical context to another sample, expert opinion, or other gold standard. Full-text manuscript reviews were conducted for the 35 articles that met the study criteria.

RESULTS

The vast majority of the studies supported the use of MTurk for a variety of academic purposes.

DISCUSSION

The literature overwhelmingly concludes that MTurk is an efficient, reliable, cost-effective tool for generating sample responses that are largely comparable to those collected via more conventional means. Caveats include survey responses may not be generalizable to the US population.

Messengers and Messages for Tweets That Used #thinspo and #fitspo Hashtags in 2016

INTRODUCTION

Twitter is widely used by young adults and is popular for seeking and sharing health information. The hashtags #thinspo and #fitspo provide a way to identify tweets designed to inspire thinness (thinspiration, thinspo) or fitness (fitspiration, fitspo). However, despite having different purposes, both terms may be associated with content that promotes eating disorders. We sought to 1) examine and compare the characteristics of senders and the content of tweets using these hashtags and 2) identify characteristics associated with engagement with a #thinspo or #fitspo tweet.

METHODS

In May 2016 we collected 1,035 tweets with #thinspo and #fitspo hashtags by using a constructed week sampling procedure. Using consensus coding, pairs of raters assessed each tweet's topic and associated images and videos. We used descriptive statistics to examine topics and user characteristics and inferential models to determine topics and characteristics associated with retweets, likes, and replies to tweets.

RESULTS

Of the 1,035 tweets, 696 (67.2%) were relevant to body image, fitness, food, dieting, or eating disorders. Fitspo tweets came from organizations or businesses, were promotional, and focused on nutrition and exercise, whereas #thinspo tweets came from individuals, focused on thinness and disordered eating behaviors, and contained images of extremely thin women. Rates of retweeting and liking were significantly higher for #thinspo than for #fitspo.

CONCLUSION

Characteristics of messages and messengers differed between #thinspo and #fitspo tweets; #thinspo tweets were used for messages about disordered eating. Public health professionals should consider using the #thinspo hashtag to reach the #thinspo group.

Implementing 360° Quantified Self for childhood obesity: feasibility study and experiences from a weight loss camp in Qatar

Background

The explosion of consumer electronics and social media are facilitating the rise of the Quantified Self (QS) movement where millions of users are tracking various aspects of their daily life using social media, mobile technology, and wearable devices. Data from mobile phones, wearables and social media can facilitate a better understanding of the health behaviors of individuals. At the same time, there is an unprecedented increase in childhood obesity rates worldwide. This is a cause for grave concern due to its potential long-term health consequences (e.g., diabetes or cardiovascular diseases). Childhood obesity is highly prevalent in Qatar and the Gulf Region. In this study we examine the feasibility of capturing quantified-self data from social media, wearables and mobiles within a weight lost camp for overweight children in Qatar.

Methods

Over 50 children (9–12 years old) and parents used a wide range of technologies, including wearable sensors (actigraphy), mobile and social media (WhatsApp and Instagram) to collect data related to physical activity and food, that was then integrated with physiological data to gain insights about their health habits.

In this paper, we report about the acquired data and visualization techniques following the 360° Quantified Self (360QS) methodology (Haddadi et al., ICHI 587–92, 2015).

Results

360QS allows for capturing insights on the behavioral patterns of children and serves as a mechanism to reinforce education of their mothers via social media. We also identified human factors, such as gender and cultural acceptability aspects that can affect the implementation of this technology beyond a feasibility study. Furthermore, technical challenges regarding the visualization and integration of heterogeneous and sparse data sets are described in the paper.

Conclusions

We proved the feasibility of using 360QS in childhood obesity through this pilot study. However, in order to fully implement the 360QS technology careful planning and integration in the health professionals’ workflow is needed.

Trial Registration

The trial where this study took place is registered at ClinicalTrials.gov on 14 November 2016 (NCT02972164).

Use of Social Media in the Diabetes Community: An Exploratory Analysis of Diabetes-Related Tweets

Background

Use of social media is becoming ubiquitous, and disease-related communities are forming online, including communities of interest around diabetes.

Objective

Our objective was to examine diabetes-related participation on Twitter by describing the frequency and timing of diabetes-related tweets, the geography of tweets, and the types of participants over a 2-year sample of 10% of all tweets.

Methods

We identified tweets with diabetes-related search terms and hashtags in a dataset of 29.6 billion tweets for the years 2013 and 2014 and extracted the text, time, location, retweet, and user information. We assessed the frequencies of tweets used across different search terms and hashtags by month and day of week and, for tweets that provided location information, by country. We also performed these analyses for a subset of tweets that used the hashtag #dsma, a social media advocacy community focused on diabetes. Random samples of user profiles in the 2 groups were also drawn and reviewed to understand the types of stakeholders participating online.

Results

We found 1,368,575 diabetes-related tweets based on diabetes-related terms and hashtags. There was a seasonality to tweets; a higher proportion occurred during the month of November, which is when World Diabetes Day occurs. The subset of tweets with the #dsma were most frequent on Thursdays (coordinated universal time), which is consistent with the timing of a weekly chat organized by this online community. Approximately 2% of tweets carried geolocation information and were most prominent in the United States (on the east and west coasts), followed by Indonesia and the United Kingdom. For the user profiles randomly selected among overall tweets, we could not identify a relationship to diabetes for the majority of users; for the profiles using the #dsma hashtag, we found that patients with type 1 diabetes and their caregivers represented the largest proportion of individuals.

Conclusions

Twitter is increasingly becoming a space for online conversations about diabetes. Further qualitative and quantitative content analysis is needed to understand the nature and purpose of these conversations.

Coding Psychological Constructs in Text Using Mechanical Turk: A Reliable, Accurate, and Efficient Alternative

In this paper we evaluate how to effectively use the crowdsourcing service, Amazon's Mechanical Turk (MTurk), to content analyze textual data for use in psychological research. MTurk is a marketplace for discrete tasks completed by workers, typically for small amounts of money. MTurk has been used to aid psychological research in general, and content analysis in particular. In the current study, MTurk workers content analyzed personally-written textual data using coding categories previously developed and validated in psychological research. These codes were evaluated for reliability, accuracy, completion time, and cost. Results indicate that MTurk workers categorized textual data with comparable reliability and accuracy to both previously published studies and expert raters. Further, the coding tasks were performed quickly and cheaply. These data suggest that crowdsourced content analysis can help advance psychological research.

Crowdsourcing and the Accuracy of Online Information Regarding Weight Gain in Pregnancy: A Descriptive Study

Background

Excess weight gain affects nearly half of all pregnancies in the United States and is a strong risk factor for adverse maternal and fetal outcomes, including long-term obesity. The Internet is a prominent source of information during pregnancy; however, the accuracy of this online information is unknown.

Objective

To identify, characterize, and assess the accuracy of frequently accessed webpages containing information about weight gain during pregnancy.

Methods

A descriptive study was used to identify and search frequently used phrases related to weight gain during pregnancy on the Google search engine. The first 10 webpages of each query were characterized by type and then assessed for accuracy and completeness, as compared to Institute of Medicine guidelines, using crowdsourcing.

Results

A total of 114 queries were searched, yielding 305 unique webpages. Of these webpages, 181 (59.3%) included information regarding weight gain during pregnancy. Out of 181 webpages, 62 (34.3%) contained no specific recommendations, 48 (26.5%) contained accurate but incomplete recommendations, 41 (22.7%) contained complete and accurate recommendations, and 22 (12.2%) were inaccurate. Webpages were most commonly from for-profit websites (112/181, 61.9%), followed by government (19/181, 10.5%), medical organizations or associations (13/181, 7.2%), and news sites (12/181, 6.6%). The largest proportion of for-profit sites contained no specific recommendations (44/112, 39.3%). Among pages that provided inaccurate information (22/181, 12.2%), 68% (15/22) were from for-profit sites.

Conclusions

For-profit websites dominate the online space with regard to weight gain during pregnancy and largely contain incomplete, inaccurate, or no specific recommendations. This represents a significant information gap regarding an important risk factor for obesity among mothers and infants. Our findings suggest that greater clinical and public health efforts to disseminate accurate information regarding healthy weight gain during pregnancy may help prevent significant morbidity and may support healthier pregnancies among at-risk women and children.

The content of social media's shared images about Ebola: a retrospective study

OBJECTIVE

Social media have strongly influenced awareness and perceptions of public health emergencies, but a considerable amount of social media content is now carried through images, rather than just text. This study's objective is to explore how image-sharing platforms are used for information dissemination in public health emergencies.

STUDY DESIGN

Retrospective review of images posted on two popular image-sharing platforms to characterize public discourse about Ebola.

METHODS

Using the keyword '#ebola' we identified a 1% sample of images posted on Instagram and Flickr across two sequential weeks in November 2014. Images from both platforms were independently coded by two reviewers and characterized by themes. We reviewed 1217 images posted on Instagram and Flickr and identified themes.

RESULTS

Nine distinct themes were identified. These included: images of health care workers and professionals [308 (25%)], West Africa [75 (6%)], the Ebola virus [59 (5%)], and artistic renderings of Ebola [64 (5%)]. Also identified were images with accompanying embedded text related to Ebola and associated: facts [68 (6%)], fears [40 (3%)], politics [46 (4%)], and jokes [284 (23%)]. Several [273 (22%)] images were unrelated to Ebola or its sequelae. Instagram images were primarily coded as jokes [255 (42%)] or unrelated [219 (36%)], while Flickr images primarily depicted health care workers and other professionals [281 (46%)] providing care or other services for prevention or treatment.

CONCLUSION

Image sharing platforms are being used for information exchange about public health crises, like Ebola. Use differs by platform and discerning these differences can help inform future uses for health care professionals and researchers seeking to assess public fears and misinformation or provide targeted education/awareness interventions.