Article

Characteristics of Twitter Use by State Medicaid Programs in the United States (Preprint)

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Background Twitter is a potentially valuable tool for public health officials and state Medicaid programs in the United States, which provide public health insurance to 72 million Americans. Objective We aim to characterize how Medicaid agencies and managed care organization (MCO) health plans are using Twitter to communicate with the public. Methods Using Twitter’s public application programming interface, we collected 158,714 public posts (“tweets”) from active Twitter profiles of state Medicaid agencies and MCOs, spanning March 2014 through June 2019. Manual content analyses identified 5 broad categories of content, and these coded tweets were used to train supervised machine learning algorithms to classify all collected posts. Results We identified 15 state Medicaid agencies and 81 Medicaid MCOs on Twitter. The mean number of followers was 1784, the mean number of those followed was 542, and the mean number of posts was 2476. Approximately 39% of tweets came from just 10 accounts. Of all posts, 39.8% (63,168/158,714) were classified as general public health education and outreach; 23.5% (n=37,298) were about specific Medicaid policies, programs, services, or events; 18.4% (n=29,203) were organizational promotion of staff and activities; and 11.6% (n=18,411) contained general news and news links. Only 4.5% (n=7142) of posts were responses to specific questions, concerns, or complaints from the public. Conclusions Twitter has the potential to enhance community building, beneficiary engagement, and public health outreach, but appears to be underutilized by the Medicaid program.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Although most of the chatter regarding Medicaid posted by consumers only included the term "medicaid" (or its variants), some directly tagged or mentioned relevant Twitter handles associated with MAs or the MCOs (eg, "@organization_name"). Corpus 2 is composed of such tweets, and the MA and MCO Twitter handles were identified in a previous study [25]. The full list of the handles used in data collection is presented in Table S2 in Multimedia Appendix 1. ...
Article
Full-text available
Background: The wide adoption of social media in daily life renders it a rich and effective resource for conducting near real-time assessments of consumers' perceptions of health services. However, its use in these assessments can be challenging because of the vast amount of data and the diversity of content in social media chatter. Objective: This study aims to develop and evaluate an automatic system involving natural language processing and machine learning to automatically characterize user-posted Twitter data about health services using Medicaid, the single largest source of health coverage in the United States, as an example. Methods: We collected data from Twitter in two ways: via the public streaming application programming interface using Medicaid-related keywords (Corpus 1) and by using the website's search option for tweets mentioning agency-specific handles (Corpus 2). We manually labeled a sample of tweets in 5 predetermined categories or other and artificially increased the number of training posts from specific low-frequency categories. Using the manually labeled data, we trained and evaluated several supervised learning algorithms, including support vector machine, random forest (RF), naïve Bayes, shallow neural network (NN), k-nearest neighbor, bidirectional long short-term memory, and bidirectional encoder representations from transformers (BERT). We then applied the best-performing classifier to the collected tweets for postclassification analyses to assess the utility of our methods. Results: We manually annotated 11,379 tweets (Corpus 1: 9179; Corpus 2: 2200) and used 7930 (69.7%) for training, 1449 (12.7%) for validation, and 2000 (17.6%) for testing. A classifier based on BERT obtained the highest accuracies (81.7%, Corpus 1; 80.7%, Corpus 2) and F1 scores on consumer feedback (0.58, Corpus 1; 0.90, Corpus 2), outperforming the second best classifiers in terms of accuracy (74.6%, RF on Corpus 1; 69.4%, RF on Corpus 2) and F1 score on consumer feedback (0.44, NN on Corpus 1; 0.82, RF on Corpus 2). Postclassification analyses revealed differing intercorpora distributions of tweet categories, with political (400778/628411, 63.78%) and consumer feedback (15073/27337, 55.14%) tweets being the most frequent for Corpus 1 and Corpus 2, respectively. Conclusions: The broad and variable content of Medicaid-related tweets necessitates automatic categorization to identify topic-relevant posts. Our proposed system presents a feasible solution for automatic categorization and can be deployed and generalized for health service programs other than Medicaid. Annotated data and methods are available for future studies.
... Using the combined data set of 10,945 users, we divided our data set into training versus validation sets using an 80:20 split, and we used these sets to train and test our classifier model, respectively. The naïve Bayes classifier yielded an accuracy of 83.2% with a precision of 0.82, a recall of 0.83, and an F1 score of 0.81; these values were considered satisfactory and are comparable to those in other studies [27][28][29]. Multimedia Appendix 1 presents the confusion matrix. Our classifier performance was also robust across multiple split strategies for dividing the data set for training and validation. ...
Article
Full-text available
Background With restrictions on movement and stay-at-home orders in place due to the COVID-19 pandemic, social media platforms such as Twitter have become an outlet for users to express their concerns, opinions, and feelings about the pandemic. Individuals, health agencies, and governments are using Twitter to communicate about COVID-19. Objective The aims of this study were to examine key themes and topics of English-language COVID-19–related tweets posted by individuals and to explore the trends and variations in how the COVID-19–related tweets, key topics, and associated sentiments changed over a period of time from before to after the disease was declared a pandemic. Methods Building on the emergent stream of studies examining COVID-19–related tweets in English, we performed a temporal assessment covering the time period from January 1 to May 9, 2020, and examined variations in tweet topics and sentiment scores to uncover key trends. Combining data from two publicly available COVID-19 tweet data sets with those obtained in our own search, we compiled a data set of 13.9 million English-language COVID-19–related tweets posted by individuals. We use guided latent Dirichlet allocation (LDA) to infer themes and topics underlying the tweets, and we used VADER (Valence Aware Dictionary and sEntiment Reasoner) sentiment analysis to compute sentiment scores and examine weekly trends for 17 weeks. Results Topic modeling yielded 26 topics, which were grouped into 10 broader themes underlying the COVID-19–related tweets. Of the 13,937,906 examined tweets, 2,858,316 (20.51%) were about the impact of COVID-19 on the economy and markets, followed by spread and growth in cases (2,154,065, 15.45%), treatment and recovery (1,831,339, 13.14%), impact on the health care sector (1,588,499, 11.40%), and governments response (1,559,591, 11.19%). Average compound sentiment scores were found to be negative throughout the examined time period for the topics of spread and growth of cases, symptoms, racism, source of the outbreak, and political impact of COVID-19. In contrast, we saw a reversal of sentiments from negative to positive for prevention, impact on the economy and markets, government response, impact on the health care industry, and treatment and recovery. Conclusions Identification of dominant themes, topics, sentiments, and changing trends in tweets about the COVID-19 pandemic can help governments, health care agencies, and policy makers frame appropriate responses to prevent and control the spread of the pandemic.
Article
Full-text available
Spurred by accumulated evidence documenting how social determinants of health shape health outcomes as well as the push for better value, the healthcare sector is embracing interventions that address patients’ health-related social needs. An increasing number of healthcare organizations and payers are experimenting with strategies to identify needs and connect patients to resources that address identified needs with the goal of improving health outcomes, reducing avoidable utilization of costly health services, and improving health equity. Although many studies link social factors to health, relatively little published research exists about how the healthcare sector can effectively intervene to help identify and address social needs. This paper summarizes emerging evidence and identifies key areas where more research is needed to advance implementation and policy development. Although some healthcare-based social needs interventions have been shown to improve health and reduce avoidable utilization, important gaps remain in terms of comparative effectiveness and cost effectiveness of social needs intervention approaches. Additionally, the field would benefit from an increased understanding of mechanisms of action to maximize practitioners’ ability to tailor interventions. More research is also needed to guard against unintended consequences and ensure these interventions reduce health inequities. Finally, implementation science research should identify supports and incentives for adoption of effective interventions. Focusing both public and private research efforts on these evidence gaps can help advance identification of interventions that maximize both health equity and healthcare value. Supplement information This article is part of a supplement entitled Identifying and Intervening on Social Needs in Clinical Settings: Evidence and Evidence Gaps, which is sponsored by the Agency for Healthcare Research and Quality of the U.S. Department of Health and Human Services, Kaiser Permanente, and the Robert Wood Johnson Foundation.
Article
Full-text available
Objective To investigate factors associated with engagement of U.S. Federal Health Agencies via Twitter. Our specific goals are to study factors related to a) numbers of retweets, b) time between the agency tweet and first retweet and c) time between the agency tweet and last retweet. Methods We collect 164,104 tweets from 25 Federal Health Agencies and their 130 accounts. We use negative binomial hurdle regression models and Cox proportional hazards models to explore the influence of 26 factors on agency engagement. Account features include network centrality, tweet count, numbers of friends, followers, and favorites. Tweet features include age, the use of hashtags, user-mentions, URLs, sentiment measured using Sentistrength, and tweet content represented by fifteen semantic groups. Results A third of the tweets (53,556) had zero retweets. Less than 1% (613) had more than 100 retweets (mean = 284). The hurdle analysis shows that hashtags, URLs and user-mentions are positively associated with retweets; sentiment has no association with retweets; and tweet count has a negative association with retweets. Almost all semantic groups, except for geographic areas, occupations and organizations, are positively associated with retweeting. The survival analyses indicate that engagement is positively associated with tweet age and the follower count. Conclusions Some of the factors associated with higher levels of Twitter engagement cannot be changed by the agencies, but others can be modified (e.g., use of hashtags, URLs). Our findings provide the background for future controlled experiments to increase public health engagement via Twitter.
Article
Full-text available
One of the essential services provided by the US local health departments is informing and educating constituents about health. Communication with constituents about public health issues and health risks is among the standards required of local health departments for accreditation. Past research found that only 61% of local health departments met standards for informing and educating constituents, suggesting a considerable gap between current practices and best practice. Social media platforms, such as Twitter, may aid local health departments in informing and educating their constituents by reaching large numbers of people with real-time messages at relatively low cost. Little is known about the followers of local health departments on Twitter. The aim of this study was to examine characteristics of local health department Twitter followers and the relationship between local health department characteristics and follower characteristics. In 2013, we collected (using NodeXL) and analyzed a sample of 4779 Twitter followers from 59 randomly selected local health departments in the United States with Twitter accounts. We coded each Twitter follower for type (individual, organization), location, health focus, and industry (eg, media, government). Local health department characteristics were adopted from the 2010 National Association of City and County Health Officials Profile Study data. Local health department Twitter accounts were followed by more organizations than individual users. Organizations tended to be health-focused, located outside the state from the local health department being followed, and from the education, government, and non-profit sectors. Individuals were likely to be local and not health-focused. Having a public information officer on staff, serving a larger population, and "tweeting" more frequently were associated with having a higher percentage of local followers. Social media has the potential to reach a wide and diverse audience. Understanding audience characteristics can help public health organizations use this new tool more effectively by tailoring tweet content and dissemination strategies for their audience.
Article
Full-text available
Public health agencies are actively using social media, including Twitter. In the public health and nonprofit sectors, Twitter has been limited to one-way communication. Two-way, interactive communication on Twitter has the potential to enhance organizational relationships with followers and help organizations achieve their goals by increasing communication and dialog between the organization and its followers. Research shows that nonprofit organizations use Twitter for three main functions: information sharing, community building, and action. It is not known whether state health departments are using Twitter primarily for one-way information sharing or if they are trying to engage followers to build relationships and promote action. The purpose of this research was to discover what the primary function of Twitter use is among state health departments in the United States and whether this is similar to or different from nonprofit organizations. A complete list of "tweets" made by each state health department account was obtained using the Twitter application programming interface. We randomly sampled 10% of each state health department's tweets. Four research assistants hand-coded the tweets' primary focus (organization centric or personal health information centric) and then the subcategories of information dissemination, engagement, or action. Research assistants coded each tweet for interactivity, sophistication, and redirects to another website. Data were analyzed using SPSS version 20. There were 4221 tweets from 39 state health departments. There was no statistically significant difference in the number of tweets made by a state health department and the state population density (P=.25). The majority of tweets focused on personal health topics (69.37%, 2928/4221) while one-third were tweets about the organization (29.14% , 1230/4221). The main function of organization-based tweets was engagement through conversations to build community (65.77%, 809/1236). These engagement-related tweets were primarily recognition of other organizations' events (43.6%, 353/809) and giving thanks and recognition (21.4%, 173/809). Nearly all of the personal health information-centric tweets involved general public health information (92.10%, 1399/1519) and 79.03% (3336/4221) of tweets directed followers to another link for more information. This is the first study to assess the purpose of public health tweets among state health departments. State health departments are using Twitter as a one-way communication tool, with tweets focused primarily on personal health. A state health department Twitter account may not be the primary health information source for individuals. Therefore, state health departments should reconsider their focus on personal health tweets and envision how they can use Twitter to develop relationships with community agencies and partners. In order to realize the potential of Twitter to establish relationships and develop connections, more two-way communication and interaction are essential.
Article
Full-text available
Use of data generated through social media for health studies is gradually increasing. Twitter is a short-text message system developed 6 years ago, now with more than 100 million users generating over 300 million Tweets every day. Twitter may be used to gain real-world insights to promote healthy behaviors. The purposes of this paper are to describe a practical approach to analyzing Tweet contents and to illustrate an application of the approach to the topic of physical activity. The approach includes five steps: (1) selecting keywords to gather an initial set of Tweets to analyze; (2) importing data; (3) preparing data; (4) analyzing data (topic, sentiment, and ecologic context); and (5) interpreting data. The steps are implemented using tools that are publically available and free of charge and designed for use by researchers with limited programming skills. Content mining of Tweets can contribute to addressing challenges in health behavior research.
Article
Full-text available
Effective communication is a critical function within any public health system. Social media has enhanced communication between individuals and organizations and has the potential to augment public health communication. However, there is a lack of reported data on social media adoption within public health settings. The purposes of this study were to assess: 1) the extent to which state public health departments (SHDs) are using social media; 2) which social media applications are used most often; and 3) how often social media is used interactively to engage audiences. This was a non-experimental, cross sectional study of SHD social media sites. Screen capture software Snag-It® was used to obtain screenshots of SHD social media sites across five applications. These sites were coded for social media presence, interactivity, reach, and topic. Sixty percent of SHDs reported using at least one social media application. Of these, 86.7% had a Twitter account, 56% a Facebook account, and 43% a YouTube channel. There was a statistically significant difference between average population density and use of social media (p = .01). On average, SHDs made one post per day on social media sites, and this was primarily to distribute information; there was very little interaction with audiences. SHDs have few followers or friends on their social media sites. The most common topics for posts and tweets related to staying healthy and diseases and conditions. Limitations include the absence of a standard by which social media metrics measure presence, reach, or interactivity; SHDs were only included if they had an institutionally maintained account; and the study was cross sectional. Social media use by public health agencies is in the early adoption stage. However, the reach of social media is limited. SHDs are using social media as a channel to distribute information rather than capitalizing on the interactivity available to create conversations and engage with the audience. If public health agencies are to effectively use social media then they must develop a strategic communication plan that incorporates best practices for expanding reach and fostering interactivity and engagement.
Article
Full-text available
Given the rapid changes in the communication landscape brought about by participative Internet use and social media, it is important to develop a better understanding of these technologies and their impact on health communication. The first step in this effort is to identify the characteristics of current social media users. Up-to-date reporting of current social media use will help monitor the growth of social media and inform health promotion/communication efforts aiming to effectively utilize social media. The purpose of the study is to identify the sociodemographic and health-related factors associated with current adult social media users in the United States. Data came from the 2007 iteration of the Health Information National Trends Study (HINTS, N = 7674). HINTS is a nationally representative cross-sectional survey on health-related communication trends and practices. Survey respondents who reported having accessed the Internet (N = 5078) were asked whether, over the past year, they had (1) participated in an online support group, (2) written in a blog, (3) visited a social networking site. Bivariate and multivariate logistic regression analyses were conducted to identify predictors of each type of social media use. Approximately 69% of US adults reported having access to the Internet in 2007. Among Internet users, 5% participated in an online support group, 7% reported blogging, and 23% used a social networking site. Multivariate analysis found that younger age was the only significant predictor of blogging and social networking site participation; a statistically significant linear relationship was observed, with younger categories reporting more frequent use. Younger age, poorer subjective health, and a personal cancer experience predicted support group participation. In general, social media are penetrating the US population independent of education, race/ethnicity, or health care access. Recent growth of social media is not uniformly distributed across age groups; therefore, health communication programs utilizing social media must first consider the age of the targeted population to help ensure that messages reach the intended audience. While racial/ethnic and health status-related disparities exist in Internet access, among those with Internet access, these characteristics do not affect social media use. This finding suggests that the new technologies, represented by social media, may be changing the communication pattern throughout the United States.
Conference Paper
Ensemble methods are learning algorithms that construct a set of classifiers and then classify new data points by taking a (weighted) vote of their predictions. The original ensemble method is Bayesian averaging, but more recent algorithms include error-correcting output coding, Bagging, and boosting. This paper reviews these methods and explains why ensembles can often perform better than any single classifier. Some previous studies comparing ensemble methods are reviewed, and some new experiments are presented to uncover the reasons that Adaboost does not overfit rapidly.
Article
Background Social media provide new channels for hospitals to engage with communities, a goal of increasing importance as non-profit hospitals face stricter definitions of community benefit under the Affordable Care Act. We describe the variability in social media presence among US children's hospitals and the distribution of their Facebook content curation. Methods Social media data from freestanding children's hospitals were extracted from September–November 2013. Social media adoption was reviewed for each hospital-generated Facebook, Twitter, YouTube, Google+ and Pinterest platform. Facebook page (number of Likes) and Twitter account (number of followers) engagement were examined by hospital characteristics. Facebook posts from each hospital over a 6-week period were thematically characterized. Results We reviewed 5 social media platforms attributed to 45 children's hospitals and 2004 associated Facebook posts. All hospitals maintained Facebook and Twitter accounts and most used YouTube (82%), Google+ (53%) and Pinterest (69%). Larger hospitals were more often high performers for Facebook (67% versus 10%, p<0.01) and Twitter (75% versus 17%, p<0.05) engagement than small hospitals. The most common Facebook post-themes were hospital promotion 35% (706), education and information 35% (694), community partnership or benefit 24% (474), fundraising 21% (426), and narratives 12% (241). Of health education posts, 73% (509) provided pediatric health supervision and anticipatory guidance. Conclusions Social media adoption by US children's hospitals was widespread. Implications Beyond its traditional marketing role, social media can serve as a conduit for health education, engagement with communities, including community benefit.
Article
Social media offers unprecedented opportunities for public health to engage audiences in conversations and collaboration that could potentially lead to improved health conditions. While there is some evidence that local health departments (LHDs) are using social media and Twitter in particular, little is known about how Twitter is used by LHDs and how they use it to engage followers versus disseminating one-way information. To examine how LHDs use Twitter to share information, engage with followers, and promote action, as well as to discover differences in Twitter use among LHDs by size of population served. The Twitter accounts for 210 LHDs were stratified into three groups based on size of population served (n=69 for less than 100,000; n=89 for 100,000-499,999; n=52 for 500,000 or greater). A sample of 1000 tweets was obtained for each stratum and coded as being either about the organization or about personal-health topics. Subcategories for organization included information, engagement, and action. Subcategories for personal health included information and action. Of all LHD tweets (n=3000), 56.1% (1682/3000) related to personal health compared with 39.5% (1186/3000) that were about the organization. Of the personal-health tweets, 58.5% (984/1682) involved factual information and 41.4% (697/1682) encouraged action. Of the organization-related tweets, 51.9% (615/1186) represented one-way communication about the organization and its events and services, 35.0% (416/1186) tried to engage followers in conversation, and 13.3% (158/1186) encouraged action to benefit the organization (eg, attend events). Compared with large LHDs, small LHDs were more likely to post tweets about their organization (Cramer's V=0.06) but were less likely to acknowledge events and accomplishments of other organizations (χ(2)=12.83, P=.02, Cramer's V=0.18). Small LHDs were also less likely to post personal health-related tweets (Cramer's V=0.08) and were less likely to post tweets containing suggestions to take action to modify their lifestyle. While large LHDs were more likely to post engagement-related tweets about the organization (Cramer's V=0.12), they were less likely to ask followers to take action that would benefit the organization (χ(2)=7.59, P=.02. Cramer's V=0.08). While certain associations were statistically significant, the Cramer's V statistic revealed weak associations. Twitter is being adopted by LHDs, but its primary use involves one-way communication on personal-health topics as well as organization-related information. There is also evidence that LHDs are starting to use Twitter to engage their audiences in conversations. As public health transitions to more dialogic conversation and engagement, Twitter's potential to help form partnerships with audiences and involve them as program participants may lead to action for improved health.
Conference Paper
. Ensemble methods are learning algorithms that construct a set of classifiers and then classify new data points by taking a (weighted) vote of their predictions. The original ensemble method is Bayesian averaging, but more recent algorithms include error-correcting output coding, Bagging, and boosting. This paper reviews these methods and explains why ensembles can often perform better than any single classifier. Some previous studies comparing ensemble methods are reviewed, and some new experiments are presented to uncover the reasons that Adaboost does not overfit rapidly. 1 Introduction Consider the standard supervised learning problem. A learning program is given training examples of the form f(x 1
Share of US adults using social media, including Facebook, is mostly unchanged since 2018. FactTank.: Pew Research Center
  • A Perrin
  • M Anderson
Perrin A, Anderson M. Share of US adults using social media, including Facebook, is mostly unchanged since 2018. FactTank.: Pew Research Center; 2019 Apr 10. URL: https://www.pewresearch.org/fact-tank/2019/04/10/ share-of-u-s-adults-using-social-media-including-facebook-is-mostly-unchanged-since-2018/ [accessed 2020-04-22]
Beneficiaries at a Glance. Health Care Quality Measures.: Centers for Medicare and Medicaid Services
  • Chip Medicaid
Medicaid and CHIP Beneficiaries at a Glance. Health Care Quality Measures.: Centers for Medicare and Medicaid Services; 2020 Feb. URL: https://www.medicaid.gov/medicaid/quality-of-care/downloads/beneficiary-ataglance.pdf [accessed 2020-03-16]
), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information
  • Abeed Zhu
  • Sarah Sarker
  • Raina Gollust
  • David Merchant
  • Grande
©Jane M Zhu, Abeed Sarker, Sarah Gollust, Raina Merchant, David Grande. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 17.08.2020. This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.