Patterns Of User Engagement Research Articles

Ensemble graph neural networks for fake news detection using user engagement and text features

In this digital world with massive information sharing, spreading misinformation is relatively easy, but its consequences are really enormous. Detecting fake news is a critical challenge in today's digital age, where misinformation can spread rapidly and influence public opinion. This paper proposes an ensemble model (EGNN) that leverages Graph Neural Networks (GNNs) and text features to enhance the accuracy of fake news detection. The model is designed to operate effectively even when labeled data is scarce by utilizing a substantial amount of unlabeled data. We employ multiple GNNs, including a standard GNN, a Graph Attention Network (GAT), and a Bidirectional GCN (BiGCN), to capture user engagement patterns and social context. Additionally, text embeddings are extracted using spaCy, BERT, and a combination of spaCy with user profile information. These embeddings are then classified using a neural network, and the final detection is performed using ensemble techniques such as Majority Voting, Weighted Average, and Stacking. To address class imbalance, we incorporate focal loss alongside the traditional binary cross-entropy loss. Experimental results demonstrate that our proposed EGNN model significantly improves the detection of fake news, providing a robust solution for this pervasive problem.

Public Participation Monitoring: Social Media Data Mining and Analysis of User Engagement Patterns

Social Media Platforms (SMP) like Twitter and Facebook have become public influencing medium in today’s digital age which facilitate community interactions. These necessitates the monitoring of public participation in these SMP. This work performs such an study by examining the user engagement patterns in social media in Indonesia during the national elections (#Pemilu2024), Ramadan celebrations (#Ramadan2024), and climate change discussions (#ClimateChange). The data for the study was collected during a period of six months from (January to June 2024) using event-specific hashtags and keywords. To identify the user engagement patterns datamining and Natural Language Processing (NLP) tools were utilized. The findings show that the Twitter platform has higher user engagement in morning with news updates and visual updates at evening. The Facebook show user engagement in afternoon with videos and evening engagement with shared articles. The Sentiment Analysis (SA) and network analysis was performed over the dataset and the findings have shown that higher positive sentiments towards elections and Ramadan but argumentative towards climate change discussions. The Twitter show rapid communication effectiveness compared to Facebook. Further the youth prefer faster update and older expect detailed content sharing.

Engagement analysis of a persuasive-design-optimized eHealth intervention through machine learning

The challenge of sustaining user engagement in eHealth interventions is a pressing issue with significant implications for the effectiveness of these digital health tools. This study investigates user engagement in a cognitive-behavioral therapy-based eHealth intervention for procrastination, using a dataset from a randomized controlled trial of 233 university students. Various machine learning models, including Decision Tree, Gradient Boosting, Logistic Regression, Random Forest, and Support Vector Machines, were employed to predict patterns of user engagement. The study adopted a two-phase analytical approach. In the first phase, all features of the dataset were included, revealing ‘total_minutes’—the total time participants spent on the intervention and the eHealth platform—as the most significant predictor of engagement. This finding emphasizes the intuitive notion that early time spent on the platform and the intervention is a strong indicator of later user engagement. However, to gain a deeper understanding of engagement beyond this predominant metric, the second phase of the analysis excluded ‘total_minutes’. This approach allowed for the exploration of the roles and interdependencies of other engagement indicators, such as ‘number_intervention_answersheets’—the number of completed lessons, ‘logins_first_4_weeks’—login frequency, and ‘number_diary_answersheets’—the number of completed diaries. The results from this phase highlighted the multifaceted nature of engagement, showing that while ‘total_minutes’ is strongly correlated with engagement, indicating that more engaged participants tend to spend more time on the intervention, the comprehensive engagement profile also depends on additional aspects like lesson completions and frequency of platform interactions.

The Aesthetic and Narrative Strategies of Subcultures on TikTok: Analyzing the Presentation of Niche Identities

This paper explores the aesthetic and narrative strategies employed by subcultures on TikTok to present and cultivate niche identities in a digital age characterized by rapid content consumption and algorithm-driven engagement. TikTok’s unique format—short, looping videos coupled with a powerful recommendation algorithm—has facilitated the rise and proliferation of diverse subcultural communities, such as cottagecore, dark academia, and witchtok. These groups leverage distinct visual styles, soundscapes, and narrative techniques to construct cohesive subcultural identities that resonate with both creators and viewers. Aesthetics serve as visual markers that communicate values and philosophies, while soundscapes and music act as auditory signifiers that further define the emotional and cultural landscapes of these communities. The platform’s algorithmic curation significantly influences the visibility, evolution, and dynamics of subcultural communities by promoting content that aligns with user engagement patterns.

Design of A Property Product Recommendation System Using Association Rule Method Based on User Interaction Patterns

This work creates an association rule-based real estate product recommendation system. Personalizing property suggestions based on user behaviour optimises property searches. Data-driven insights enhance dynamic property market user experience. Association rules alter property advice. Data-driven insights and adaptability improve property search by proposing homes depending on user engagement patterns. Strong algorithms establish location, budget, and property associations, and association rule technology and user interaction patterns increase property recommendations. Personalized property discovery uses accurate and adaptive suggestions from continuous learning. Results reveal that user interaction pattern-based association rule techniques improve property suggestion accuracy and personalization. The system's tailored advice improves property market decisions, confirming its usefulness and adaptability. Insufficient user data might distort suggestions, especially for specific interests. Not enough user diversity can lower system accuracy. User data and privacy must be secured to optimize the recommendation system. Association rules and user engagement patterns can transform property recommendations. This innovative technique can improve property searches, provide personalized ideas, and help consumers make informed decisions in a competitive market.

Toward Inclusivity: Rethinking Islamophobic Content Classification in the Digital Age

In this paper, we implement a comprehensive three class system to categorize social media discussions about Islam and Muslims, enhancing the typical binary approach. These classes are: I) General Discourse About Islam and Muslims, II) Criticism of IslamicTeachings and Figures, and III) Comments Against Muslims. These categories are designed to balance the nuances of free speech while protecting diverse groups like Muslims, ex-Muslims, LGBTQ+ communities, and atheists. By utilizing machine learning and employing transformer-based models, we analyze the distribution and characteristics of these classes in social media content. Our findings reveal distinct patterns of user engagement with topics related to Islam, providing valuable insights into the complexities of digital discourse. This research contributes to the fields of quantitative social science by offering an improved method for understanding and moderating online discussions on sensitive religious and cultural subjects.

A Review of Machine Learning Approaches for Rumour Detection: Techniques, Challenges, and Future Directions (Machine Learning for Rumour Detection)

Rumours and misinformation propagate rapidly across online social networks, posing significant challenges to maintaining the integrity of information dissemination. In recent years, machine learning (ML) techniques have emerged as promising tools for automating the detection and mitigation of rumours. This review paper provides a comprehensive examination of the advancements in rumour detection using ML approaches. The paper begins by outlining the landscape of rumour dissemination in online social networks, highlighting the characteristics and challenges associated with rumour detection. Subsequently, it systematically categorizes and analyzes various ML methods employed for rumour detection, including supervised, unsupervised, and semi-supervised learning approaches. Furthermore, the review delves into the diverse features and representations utilized in ML models for rumour detection, such as textual content, user engagement patterns, network structures, and temporal dynamics. It discusses the strengths and limitations of different feature sets and their impact on the effectiveness of rumour detection systems. Moreover, the paper explores the intricacies of dataset construction and evaluation methodologies for training and testing rumour detection models. It examines commonly used benchmark datasets and evaluation metrics, emphasizing the importance of robust evaluation frameworks for assessing the performance of ML-based rumour detection systems accurately. Additionally, the review identifies key challenges and open research questions in the field of rumour detection using ML, including handling evolving rumour patterns, addressing adversarial attacks, and enhancing the interpretability and explain ability of ML models. It also discusses potential directions for future research aimed at advancing the state-of-the-art in rumour detection and mitigation.

Computer-Aided Technology in Product Communication and Promotion Strategy

In today's competitive marketplace, effective in-product communication and promotion strategies are essential for capturing consumer attention and driving product adoption. These strategies involve utilizing various channels within the product itself to convey key messages, showcase features, and encourage user engagement. Computer-aided promotion strategies leverage digital tools and technologies to enhance marketing efforts and reach target audiences more effectively. By utilizing data analytics, automation, and digital advertising platforms, businesses can optimize promotional campaigns, personalize messaging, and target specific demographic segments with precision. This paper proposes an innovative approach to product communication and promotion strategy utilizing computer-aided technology, specifically tailored for the Weibo platform, named the Weibo Stacked Communication Classification Strategy (WeSCCS). The WeSCCS framework integrates advanced computational techniques with Weibo's communication dynamics to classify and categorize promotional content effectively. With machine learning algorithms and natural language processing, the system analyzes user engagement patterns, content preferences, and social interactions to identify optimal communication strategies. WeSCCS classifies user interactions with promotional content such as "positive engagement" (70% of interactions), "neutral engagement" (20%), and "negative engagement" (10%). Additionally, automated content generation and sentiment analysis enable marketers to generate relevant promotional content with a sentiment score of +0.8, indicating high positivity. Real-time monitoring facilitates dynamic adjustments to promotional strategies based on user feedback, ensuring responsiveness to evolving consumer preferences. By integrating computer-aided technology, WeSCCS empowers marketers to maximize promotional impact, increase audience engagement, and drive conversions effectively on Weibo and similar social media platforms.

Twitter’s pulse on hydrogen energy in 280 characters: a data perspective

Uncovering the public discourse on hydrogen energy is essential for understanding public behaviour and the evolving nature of conversations over time and across different regions. This paper presents a comprehensive analysis of a large multilingual dataset pertaining to hydrogen energy collected from Twitter spanning a decade (2013–2022) using selected keywords. The analysis aims to explore various aspects, including the temporal and spatial dimensions of the discourse, factors influencing Twitter engagement, user engagement patterns, and the interpretation of conversations through hashtags and ngrams. By delving into these aspects, this study offers valuable insights into the dynamics of public discourse surrounding hydrogen energy and the perceptions of social media users.

Health Information Sharing Pattern on Social Media and User Engagement: A Content Analysis on Facebook Public Posts in Bangladesh

In this age of social media, people are using various platforms for seeking and sharing information. This study aims at exploring the patterns of sharing health information on the widely used social media platform Facebook and the users' engagement with the information. For this purpose, the researchers used mixed-method content analysis. We analyzed 485 Facebook posts about 14 types of issues across nine categories. The study results showed that the highest number (27.84%) of posts were related to health awareness, and the lowest were personal opinions. This paper offers a basis for comprehending the processes of sharing health information and how it could affect user involvements. To clarify the structures that guide user interactions with health content on social media, the study also integrates components from communication and information-sharing concepts. The study thus offers empirical insights into the kinds of health information provided, the most popular health topics, and user engagement patterns through a thorough content analysis of public Facebook postings in Bangladesh. The analysis provides sophisticated knowledge of the issues impacting engagement, including language, kind of material, and multimedia use, by measuring the interactions of users. Scholars can understand from this study how users interact with health-related content in the digital era. The result of this study can be used by academics exploring public health, media studies, and health communication to improve theories and create focused interventions for health communication. Additionally, this research can provide useful information for health practitioners to improve their online communication tactics.

Language adaptations of mental health interventions: User interaction comparisons with an AI-enabled conversational agent (Wysa) in English and Spanish.

In recent times, digital mental health interventions (DMHIs) have been proven to be efficacious; however, most are available only for English speakers, leaving limited options for non-English languages like Spanish. Research shows that mental health services in one's dominant language show better outcomes. Conversational agents (CAs) offer promise in supporting mental health in non-English populations. This study compared a culturally adapted version of an artificial intelligence (AI)-led mental health app, called Wysa, in Spanish and English. To compare user engagement patterns on Wysa-Spanish and Wysa-English and to understand expressions of distress and preferred language in both versions of Wysa. We adopted a cross-sectional retrospective exploratory design with mixed methods, analyzing users from 10 Spanish-speaking countries between 1 February and 1 August 2022. A quantitative sample A (n = 2767) was used for descriptive statistics, including user engagement metrics with a Wilcoxon test. A subset qualitative sample B (n = 338) was examined for word count differences based on valence, and a content analysis was conducted to examine idioms of distress. Compared to Wysa-English, Wysa-Spanish had more sessions (P < .001, d = 0.18) and a greater volume of disclosure of distress. In Wysa-Spanish, the average length of a conversation was significantly longer than in Wysa-English (P < .001, d = 0.44). Users preferred interventions with free text responses ("Thought recording") in Spanish (P < .01, d = 0.41), and Spanish messages were significantly longer (P < .01, d = 0.24). Wysa-Spanish saw more frequent expressions of negative emotions and feelings of self-harm and suicide. Given the high engagement within the Spanish version of Wysa, the findings demonstrate the need for culturally adapted DMHIs among non-English populations, emphasizing the importance of considering linguistic and cultural differences in the development of DMHIs to improve accessibility for diverse populations.

Examination of Content Types and Social Media Engagement Indicators on Facebook: Case Analysis of a 5-Star Hotels of Visegrad Group Countries

Delving into the intricate world of social media engagement, this comprehensive study analyses the dynamics of user interaction with posts from 5-star hotels on Facebook across the Visegrad Group countries. It meticulously aims to shed light on the variances among engagement metrics—shares, comments, and reactions—across different types of posted content. Furthermore, it explores the complex interrelations among these metrics to provide a holistic understanding of user engagement patterns. In pursuit of this goal, we scrutinized an extensive dataset comprising 10,820 Facebook posts shared by selected 5-star hotels throughout 2019. The data were meticulously collected from the social network utilizing the accessible API, ensuring a robust and reliable foundation for analysis. The investigation employed advanced statistical tools, namely the Kruskal‒Wallis test and Spearman’s rho test, to thoroughly examine and interpret the complex data. The insights gleaned from this research are invaluable, painting a detailed picture of brand management strategies on social platforms. A significant finding of this study is the variation in user engagement levels in relation to the type of content disseminated. It highlights that visual content—specifically videos and photos—tends to dominate in terms of eliciting user responses, as compared to text statuses and links. This underlines the importance of leveraging visual media to captivate and engage the audience effectively. The study also reveals that engagement metrics are intricately linked, suggesting a synergistic effect rather than isolated impacts. This indicates that an integrated approach, considering these metrics as a cohesive unit, could be more beneficial in strategizing content for social media. Understanding these relationships and the dominant role of visual content can greatly inform and transform the way brands, especially in the hospitality industry, navigate their presence on social networks. These findings serve as a guiding framework for optimizing social media content strategies, aiming to maximize engagement and enhance the effectiveness of online brand management.

A Comparative Study of Content Originality and Plagiarism in Teaching Game Short Video: Based on Users Usage in Douyin and Kuaishou

In the contemporary era, with the growing popularity of short videos encompassing educational gaming content, individuals have exhibited an increasing inclination toward consuming creative and original video content. However, a concomitant rise in the prevalence of video plagiarism has hindered the effective resolution of this issue. The present study aims to investigate whether substantial disparities exist in user engagement patterns between original videos and copied videos. Employing a diverse selection of materials sourced from the two major short video platforms, Douyin and Kuaishou, this study meticulously analyzed various metrics including likes, comments, collection, and transmission volumes. Additionally, the study examined the influence of other pertinent factors. The research findings unequivocally indicate the presence of significant disparities between original videos and copied videos. This comprehensive analysis delves into the dissimilarities observed in the utilization patterns of basic users to offer valuable insights to researchers and practitioners involved in the study of originality and plagiarism within this domain.

The impact of a dedicated social media strategy on enhancing surgical education

PurposeSocial media (SoMe) is increasingly important in surgical education and may be necessary in the current learning environment. Whilst expanding in use and applications, few studies detail the impact of SoMe on measurable outcomes. The goal of this study was to quantify the impact of a dedicated SoMe strategy on engagement metrics for surgical research.MethodsA retrospective review of a peer-reviewed surgical journal’s Twitter microblog platform (@ColorectalDis) was performed from 6/2015 to 4/2021. A formal SoMe strategy was introduced in September 2018. Data were stratified into 2 time periods: pre-intervention (6-2015 to 9-2018) and post-SoMe intervention (9-2018 to 4-2021). The main outcome was the impact of the SoMe strategy on user engagement with the Twitter platform, journal, and traditional journal metrics. Twitter Analytics and Twitonomy were used to analyse engagement.ResultsFrom conception to analysis, the microblog published 1198 original tweets, generating 5 million impressions and 231,000 engagements. Increased account activity (increased tweets published per month—5.51 vs 28.79; p < 0.01) was associated with significant engagement growth, including new monthly followers (213 vs 38; p < 0.01) and interactions with posted articles (4,096,167 vs 269,152; p < 0.01). Article downloads increased twenty-fold post-SoMe intervention (210,449 vs 10,934; p < 0.01), with significant increases in traditional journal metrics of new subscribers (+11%), article submissions (+24%), and impact factor (+0.9) (all p < 0.01).ConclusionSoMe directly impacts traditional journal metrics in surgical research. By examining the patterns of user engagement between SoMe and journal sites, the growing beneficial impact of a structured social media strategy and SoMe as an educational tool is demonstrated.

From YouTube Comments to Insights: A Sentiment Analysis of Opinions on Productivity Tools

Abstract: YouTube, a diverse content hub, acts as a prominent platform for both viewing videos and engaging through comments that express opinions. By carefully analyzing these comments, sentiments associated with products or topics become evident. This research delves into sentiment analysis of YouTube comments, specifically focusing on sentiments towards productivity tools and technologies. Through this examination of conveyed sentiments, it deepens understanding of public perceptions and attitudes toward these tools. Utilizing YouTube's reach and engagement, the study systematically collects comment data using the YouTube API and employs the established VADER sentiment analysis framework. Employing a Bag-of-Words approach, the research incorporates machine learning algorithms like Naive Bayes and Random Forest, achieving accuracy rates of 84.01% and 91.34%, respectively. In-depth temporal analysis of user engagement patterns uncovers trends, with heightened engagement followed by decline, correlating with external events, notably the Covid-19 pandemic. These insights enhance sentiment analysis and illuminate dynamics between societal occurrences, user sentiments, and digital dialogues, offering perspectives on evolving opinions about productivity tools and technologies

AI-Led Mental Health Support (Wysa) for Health Care Workers During COVID-19: Service Evaluation.

The impact that the COVID-19 pandemic has had on health care workers' mental health, in particular, cannot be ignored. Not only did the pandemic exacerbate mental health challenges through elevated stress, anxiety, risk of infection, and social isolation, but regulations to minimize infection additionally hindered the conduct of traditional in-person mental health care. This study explores the feasibility of using Wysa, an artificial intelligence-led mental health app, among health care workers. A national tertiary health care cluster in Singapore piloted the use of Wysa among its own health care workers to support the management of their mental well-being during the pandemic (July 2020-June 2022). The adoption of this digital mental health intervention circumvented the limitations of in-person contact and enabled large-scale access to evidence-based care. Rates and patterns of user engagement were evaluated. Overall, the opportunity to use Wysa was well-received. Out of the 527 staff who were onboarded in the app, 80.1% (422/527) completed a minimum of 2 sessions. On average, users completed 10.9 sessions over 3.80 weeks. The interventions most used were for sleep and anxiety, with a strong repeat-use rate. In this sample, 46.2% (73/158) of health care workers reported symptoms of anxiety (Generalized Anxiety Disorder Assessment-7 [GAD-7]), and 15.2% (24/158) were likely to have symptoms of depression (Patient Health Questionnaire-2 [PHQ-2]). Based on the present findings, Wysa appears to strongly engage those with none to moderate symptoms of anxiety. This evaluation demonstrates the viability of implementing Wysa as a standard practice among this sample of health care workers, which may support the use of similar digital interventions across other communities.

Who speaks and who is heard on Facebook? Political mobilization and engagement patterns of partisanship and gender in Switzerland’s direct democracy

This descriptive study investigates political mobilization and user engagement patterns on Facebook and associated partisan and gender discrepancies. It focuses on Switzerland, where political actors frequently seek to mobilize and shape citizens’ opinions before direct-democratic voting on wide-ranging policy issues. Using digital trace data from CrowdTangle, the analysis focuses on the posting frequency and received user interactions of 770 Swiss political actors’ Facebook pages. The analysis period covers 20 months, from November 2019 to July 2021, during which five popular votes occurred, and the sampled FB pages published more than 226,000 posts and received more than 18,000,000 user interactions. A descriptive quantitative analysis and a multiple regression analysis revealed an overall skewed pattern: Mobilization and user engagement are driven by small subsets of highly active and interacted-with FB pages. A few FB pages of the right-wing Swiss People’s Party and male politicians receive highly disproportionate user engagement – relative to more centrist parties and female politicians – but also relative to their electoral share in the national parliament. The results show that only a few dominant political voices are widely heard on Facebook, even if many speak. These insights are of interest beyond Switzerland, as Facebook and other social media platforms shape political discourse across liberal democracies.

Analyzing User Engagement Within a Patient-Reported Outcomes Texting Tool for Diabetes Management: Engagement Phenotype Study.

Patient-reported outcomes (PROs) capture patients' views on their health conditions and its management, and are increasingly used in clinical trials, including those targeting type 2 diabetes (T2D). Mobile health (mHealth) tools offer novel solutions for collecting PRO data in real time. Although patients are at the center of any PRO-based intervention, few studies have examined user engagement with PRO mHealth tools. This study aimed to evaluate user engagement with a PRO mHealth tool for T2D management, identify patterns of user engagement and similarities and differences between the patients, and identify the characteristics of patients who are likely to drop out or be less engaged with a PRO mHealth tool. We extracted user engagement data from an ongoing clinical trial that tested the efficacy of a PRO mHealth tool designed to improve hemoglobin A1c levels in patients with uncontrolled T2D. To date, 61 patients have been randomized to the intervention, where they are sent 6 PRO text messages a day that are relevant to T2D self-management (healthy eating and medication adherence) over the 12-month study. To analyze user engagement, we first compared the response rate (RR) and response time between patients who completed the 12-month intervention and those who dropped out early (noncompleters). Next, we leveraged latent class trajectory modeling to classify patients from the completer group into 3 subgroups based on similarity in the longitudinal engagement data. Finally, we investigated the differences between the subgroups of completers from various cross-sections (time of the day and day of the week) and PRO types. We also explored the patient demographics and their distribution among the subgroups. Overall, 19 noncompleters had a lower RR to PRO questions and took longer to respond to PRO questions than 42 completers. Among completers, the longitudinal RRs demonstrated differences in engagement patterns over time. The completers with the lowest engagement showed peak engagement during month 5, almost at the midstage of the program. The remaining subgroups showed peak engagement at the beginning of the intervention, followed by either a steady decline or sustained high engagement. Comparisons of the demographic characteristics showed significant differences between the high engaged and low engaged subgroups. The high engaged completers were predominantly older, of Hispanic descent, bilingual, and had a graduate degree. In comparison, the low engaged subgroup was composed mostly of African American patients who reported the lowest annual income, with one of every 3 patients earning less than US $20,000 annually. There are discernible engagement phenotypes based on individual PRO responses, and their patterns vary in the timing of peak engagement and demographics. Future studies could use these findings to predict engagement categories and tailor interventions to promote longitudinal engagement. Clinicaltrials.gov NCT03652389; https://clinicaltrials.gov/ct2/show/NCT03652389. RR2-10.2196/18554.

Engagement of individuals aged 14–25 years with a digital mental health service during the COVID-19 pandemic in the UK: a cohort study of service usage data

BackgroundPublic use of digital mental health technologies has informed several studies focusing on patterns of engagement within user-led digital support systems. General engagement with these services has increased since the start of the COVID-19 pandemic. However, to the best of our knowledge, this is the first study to explore how user engagement patterns with these platforms changed during the pandemic. MethodsIn this cohort study, we included individuals aged 14–25 years who signed up to the online mental health platform Kooth in the UK before the pandemic (May 1, 2019 to March 10, 2020) and during the pandemic (March 11, 2020 to Oct 31, 2021). Routinely collected usage data were assessed for consenting users who returned to Kooth at least once after sign-up. We used logistic regression to compare service component usage and overall engagement across cohorts. Findings5048 individuals (mean sign-ups per day 9·0 [12·0]) were included in the pre-pandemic cohort and 13 841 (mean sign-ups per day 11·5 [13·9]) in the pandemic cohort, representing a significant increase in sign-ups per day during the pandemic compared with the pre-pandemic period (Welch's t-test p=0·0001). Compared with users pre-pandemic, users during the pandemic were less likely to have a drop-in chat with a practitioner (odds ratio [OR] 0·50 [95% CI 0·47–0·54], p<0·0001) but were more likely to journal (1·92 [1·77–2·07], p<0·0001) and comment on existing content (1·10 [1·03–1·18], p=0·008). Users who signed up to Kooth during the pandemic and requested a chat with a practitioner (7816 [56·5%] of 13 841 users) were also less likely to receive a chat than users who requested a chat pre-pandemic (3791 [75·1%] of 5048 users; OR 0·68 [0·63–0·74], p<0·0001). InterpretationDuring the COVID-19 pandemic, people with the ability and resources to access digital mental health support were diverted from face-to-face services to digital mental health services. This change might have resulted in an unmet need for drop-in practitioner chats, with users probably resorting to asynchronous support through community interaction or journaling. FundingKooth Digital Health.

Characterizing User Engagement With a Digital Intervention for Pain Self-management Among Youth With Sickle Cell Disease and Their Caregivers: Subanalysis of a Randomized Controlled Trial.

BackgroundSickle cell disease (SCD) is characterized by severe acute pain episodes as well as risk for chronic pain. Digital delivery of SCD pain self-management support may enhance pain self-management skills and accessibility for youth. However, little is known about how youth with SCD and their caregivers engage with digital health programs. iCanCope with pain is a digital pain self-management platform adapted for youth with SCD and caregivers through a user-centered design approach. The program was delivered via a website (separate versions for youth and caregiver) and mobile app (youth only).ObjectiveWe aimed to characterize patterns of user engagement with the iCanCope with SCD program among youth with SCD and their caregivers.MethodsA randomized controlled trial was completed across multiple North American SCD clinics. Eligible youth were aged 12-18 years, diagnosed with SCD, English-speaking, and experiencing moderate-to-severe pain interference. Eligible caregivers were English-speaking with a child enrolled in the study. Dyads were randomized to receive the iCanCope intervention or attention-control education for 8-12 weeks. This report focused on engagement among dyads who received the intervention. User-level analytics were captured. Individual interviews were conducted with 20% of dyads. Descriptive statistics characterized quantitative engagement. Content analysis summarized qualitative interview data. Exploratory analysis tested the hypothesis that caregiver engagement would be positively associated with child engagement.ResultsThe cohort included primarily female (60% [34/57] of youth; 91% [49/56] of caregivers) and Black (>90% of youth [53/57] and caregivers [50/56]) participants. Among 56 dyads given program access, differential usage patterns were observed: both the youth and caregiver engaged (16/56, 29%), only the youth engaged (24/56, 43%), only the caregiver engaged (1/56, 2%), and neither individual engaged (16/56, 29%). While most youth engaged with the program (40/57, 70%), most caregivers did not (39/56, 70%). Youth were more likely to engage with the app than the website (85% [34/57] versus 68% [23/57]), and the most popular content categories were goal setting, program introduction, and symptom history. Among caregivers, program introduction, behavioral plans, and goal setting were the most popular content areas. As hypothesized, there was a moderate positive association between caregiver and child engagement (χ21=6.6; P=.01; ϕ=0.34). Interviews revealed that most dyads would continue to use the program (11/12, 92%) and recommend it to others (10/12, 83%). The reasons for app versus website preference among youth were ease of use, acceptable time commitment, and interactivity. Barriers to caregiver engagement included high time burden and limited perceived relevance of content.ConclusionsThis is one of the first studies to apply digital health analytics to characterize patterns of engagement with SCD self-management among youth and caregivers. The findings will be used to optimize the iCanCope with SCD program prior to release.Trial RegistrationClinicalTrials.gov NCT03201874; https://clinicaltrials.gov/ct2/show/NCT03201874