Open Access Policy Deposits

This series is automatically populated with publications deposited by UC Irvine Donald Bren School of Information and Computer Sciences Department of Informatics researchers in accordance with the University of California’s open access policies. For more information see Open Access Policy Deposits and the UC Publication Management System.

Sort By:

Show:

ChatGPT in education: global reactions to AI innovations.

(2023)

The release and rapid diffusion of ChatGPT have caught the attention of educators worldwide. Some educators are enthusiastic about its potential to support learning. Others are concerned about how it might circumvent learning opportunities or contribute to misinformation. To better understand reactions about ChatGPT concerning education, we analyzed Twitter data (16,830,997 tweets from 5,541,457 users). Based on topic modeling and sentiment analysis, we provide an overview of global perceptions and reactions to ChatGPT regarding education. ChatGPT triggered a massive response on Twitter, with education being the most tweeted content topic. Topics ranged from specific (e.g., cheating) to broad (e.g., opportunities), which were discussed with mixed sentiment. We traced that authority decisions may influence public opinions. We discussed that the average reaction on Twitter (e.g., using ChatGPT to cheat in exams) differs from discussions in which education and teaching-learning researchers are likely to be more interested (e.g., ChatGPT as an intelligent learning partner). This study provides insights into peoples reactions when new groundbreaking technology is released and implications for scientific and policy communication in rapidly changing circumstances.

Pregnant in a Pandemic: Connecting Perceptions of Uplifts and Hassles to Mental Health.

(2023)

How women experience pregnancy as uplifting or a hassle is related to their mental and physical health and birth outcomes. Pregnancy during a pandemic introduces new hassles, but may offer benefits that could affect how women perceive their pregnancy. Surveying 118 ethnically and racially diverse pregnant women, we explore (1) women's traditional and pandemic-related pregnancy uplifts and hassles and (2) how these experiences of pregnancy relate to their feelings of loneliness, positivity, depression, and anxiety. Regressions show that women who experience more intense feelings of uplifts than hassles also feel more positive, less lonely, and have better mental health. Findings suggest that focusing on positive aspects of being pregnant, in general and during a pandemic, might be beneficial for pregnant women's mental health.

Cover page of Socioenvironmental Adversity and Adolescent Psychotic Experiences: Exploring Potential Mechanisms in a UK Longitudinal Cohort.

Socioenvironmental Adversity and Adolescent Psychotic Experiences: Exploring Potential Mechanisms in a UK Longitudinal Cohort.

(2023)

Background and hypothesis

Children exposed to socioenvironmental adversities (eg, urbanicity, pollution, neighborhood deprivation, crime, and family disadvantage) are more likely to subsequently develop subclinical psychotic experiences during adolescence (eg, hearing voices, paranoia). However, the pathways through which this occurs have not been previously investigated. We hypothesized that cognitive ability and inflammation would partly explain this association.

Study design

Data were utilized from the Environmental-Risk Longitudinal Twin Study, a cohort of 2232 children born in 1994-1995 in England and Wales and followed to age 18. Socioenvironmental adversities were measured from birth to age 10 and classified into physical risk (defined by high urbanicity and air pollution) and socioeconomic risk (defined by high neighborhood deprivation, neighborhood disorder, and family disadvantage). Cognitive abilities (overall, crystallized, fluid, and working memory) were assessed at age 12; and inflammatory markers (C-reactive protein, interleukin-6, soluble urokinase plasminogen activator receptor) were measured at age 18 from blood samples. Participants were interviewed at age 18 regarding psychotic experiences.

Study results

Higher physical risk and socioeconomic risk were associated with increased odds of psychotic experiences in adolescence. The largest mediation pathways were from socioeconomic risk via overall cognitive ability and crystallized ability, which accounted for ~11% and ~19% of the association with psychotic experiences, respectively. No statistically significant pathways were found via inflammatory markers in exploratory (partially cross-sectional) analyses.

Conclusions

Cognitive ability, especially crystallized ability, may partly explain the association between childhood socioenvironmental adversity and adolescent psychotic experiences. Interventions to support cognitive development among children living in disadvantaged settings could buffer them against developing subclinical psychotic phenomena.

Cover page of Understanding Users Experiences of a Novel Web-Based Cognitive Behavioral Therapy Platform for Depression and Anxiety: Qualitative Interviews From Pilot Trial Participants.

Understanding Users Experiences of a Novel Web-Based Cognitive Behavioral Therapy Platform for Depression and Anxiety: Qualitative Interviews From Pilot Trial Participants.

(2023)

BACKGROUND: Digital mental health interventions (DMHIs) can help bridge the gap between the demand for mental health care and availability of treatment resources. The affordances of DMHIs have been proposed to overcome barriers to care such as accessibility, cost, and stigma. Despite these proposals, most evaluations of the DMHI focus on clinical effectiveness, with less consideration of users perspectives and experiences. OBJECTIVE: We conducted a pilot randomized controlled trial of Overcoming Thoughts, a web-based platform that uses cognitive and behavioral principles to address depression and anxiety. The Overcoming Thoughts platform included 2 brief interventions-cognitive restructuring and behavioral experimentation. Users accessed either a version that included asynchronous interactions with other users (crowdsourced platform) or a completely self-guided version (control condition). We aimed to understand the users perspectives and experiences by conducting a subset of interviews during the follow-up period of the trial. METHODS: We used purposive sampling to select a subset of trial participants based on group assignment (treatment and control) and symptom improvement (those who improved and those who did not on primary outcomes). We conducted semistructured interviews with 23 participants during the follow-up period that addressed acceptability, usability, and impact. We conducted a thematic analysis of the interviews until saturation was reached. RESULTS: A total of 8 major themes were identified: possible opportunities to expand the platform; improvements in mental health because of using the platform; increased self-reflection skills; platform being more helpful for certain situations or domains; implementation of skills into users lives, even without direct platform use; increased coping skills because of using the platform; repetitiveness of platform exercises; and use pattern. Although no differences in themes were found among groups based on improvement status (all P values >.05, ranging from .12 to .86), there were 4 themes that differed based on conditions (P values from .01 to .046): helpfulness of self-reflection supported by an exercise summary (greater in control); aiding in slowing thoughts and feeling calmer (greater in control); overcoming patterns of avoidance (greater in control); and repetitiveness of content (greater in the intervention). CONCLUSIONS: We identified the different benefits that users perceived from a novel DMHI and opportunities to improve the platform. Interestingly, we did not note any differences in themes between those who improved and those who did not, but we did find some differences between those who received the control and intervention versions of the platform. Future research should continue to investigate users experiences with DMHIs to better understand the complex dynamics of their use and outcomes.

Using machine learning to develop a clinical prediction model for SSRI-associated bleeding: a feasibility study.

(2023)

Introduction

Adverse drug events (ADEs) are associated with poor outcomes and increased costs but may be prevented with prediction tools. With the National Institute of Health All of Us (AoU) database, we employed machine learning (ML) to predict selective serotonin reuptake inhibitor (SSRI)-associated bleeding.

Methods

The AoU program, beginning in 05/2018, continues to recruit ≥ 18 years old individuals across the United States. Participants completed surveys and consented to contribute electronic health record (EHR) for research. Using the EHR, we determined participants who were exposed to SSRIs (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline, vortioxetine). Features (n = 88) were selected with clinicians' input and comprised sociodemographic, lifestyle, comorbidities, and medication use information. We identified bleeding events with validated EHR algorithms and applied logistic regression, decision tree, random forest, and extreme gradient boost to predict bleeding during SSRI exposure. We assessed model performance with area under the receiver operating characteristic curve statistic (AUC) and defined clinically significant features as resulting in > 0.01 decline in AUC after removal from the model, in three of four ML models.

Results

There were 10,362 participants exposed to SSRIs, with 9.6% experiencing a bleeding event during SSRI exposure. For each SSRI, performance across all four ML models was relatively consistent. AUCs from the best models ranged 0.632-0.698. Clinically significant features included health literacy for escitalopram, and bleeding history and socioeconomic status for all SSRIs.

Conclusions

We demonstrated feasibility of predicting ADEs using ML. Incorporating genomic features and drug interactions with deep learning models may improve ADE prediction.

Cover page of Facilitators of and Barriers to Integrating Digital Mental Health Into County Mental Health Services: Qualitative Interview Analyses.

Facilitators of and Barriers to Integrating Digital Mental Health Into County Mental Health Services: Qualitative Interview Analyses.

(2023)

Background

Digital mental health interventions (DMHIs) represent a promising solution to address the growing unmet mental health needs and increase access to care. Integrating DMHIs into clinical and community settings is challenging and complex. Frameworks that explore a wide range of factors, such as the Exploration, Preparation, Implementation, Sustainment (EPIS) framework, can be useful for examining multilevel factors related to DMHI implementation efforts.

Objective

This paper aimed to identify the barriers to, facilitators of, and best practice recommendations for implementing DMHIs across similar organizational settings, according to the EPIS domains of inner context, outer context, innovation factors, and bridging factors.

Methods

This study stems from a large state-funded project in which 6 county behavioral health departments in California explored the use of DMHIs as part of county mental health services. Our team conducted interviews with clinical staff, peer support specialists, county leaders, project leaders, and clinic leaders using a semistructured interview guide. The development of the semistructured interview guide was informed by expert input regarding relevant inner context, outer context, innovation factors, and bridging factors in the exploration, preparation, and implementation phases of the EPIS framework. We followed a recursive 6-step process to conduct qualitative analyses using inductive and deductive components guided by the EPIS framework.

Results

On the basis of 69 interviews, we identified 3 main themes that aligned with the EPIS framework: readiness of individuals, readiness of innovations, and readiness of organizations and systems. Individual-level readiness referred to the extent to which clients had the necessary technological tools (eg, smartphones) and knowledge (digital literacy) to support the DMHI. Innovation-level readiness pertained to the accessibility, usefulness, safety, and fit of the DMHI. Organization- and system-level readiness concerned the extent to which providers and leadership collectively held positive views about DMHIs as well as the extent to which infrastructure (eg, staffing and payment model) was appropriate.

Conclusions

The successful implementation of DMHIs requires readiness at the individual, innovation, and organization and system levels. To improve individual-level readiness, we recommend equitable device distribution and digital literacy training. To improve innovation readiness, we recommend making DMHIs easier to use and introduce, clinically useful, and safe and adapting them to fit into the existing client needs and clinical workflow. To improve organization- and system-level readiness, we recommend supporting providers and local behavioral health departments with adequate technology and training and exploring potential system transformations (eg, integrated care model). Conceptualizing DMHIs as services allows the consideration of both the innovation characteristics of DMHIs (eg, efficacy, safety, and clinical usefulness) and the ecosystem around DMHIs, such as individual and organizational characteristics (inner context), purveyors and intermediaries (bridging factor), client characteristics (outer context), as well as the fit between the innovation and implementation settings (innovation factor).

Cover page of Blockchain-enabled immutable, distributed, and highly available clinical research activity logging system for federated COVID-19 data analysis from multiple institutions.

Blockchain-enabled immutable, distributed, and highly available clinical research activity logging system for federated COVID-19 data analysis from multiple institutions.

(2023)

Objective

We aimed to develop a distributed, immutable, and highly available cross-cloud blockchain system to facilitate federated data analysis activities among multiple institutions.

Materials and methods

We preprocessed 9166 COVID-19 Structured Query Language (SQL) code, summary statistics, and user activity logs, from the GitHub repository of the Reliable Response Data Discovery for COVID-19 (R2D2) Consortium. The repository collected local summary statistics from participating institutions and aggregated the global result to a COVID-19-related clinical query, previously posted by clinicians on a website. We developed both on-chain and off-chain components to store/query these activity logs and their associated queries/results on a blockchain for immutability, transparency, and high availability of research communication. We measured run-time efficiency of contract deployment, network transactions, and confirmed the accuracy of recorded logs compared to a centralized baseline solution.

Results

The smart contract deployment took 4.5 s on an average. The time to record an activity log on blockchain was slightly over 2 s, versus 5-9 s for baseline. For querying, each query took on an average less than 0.4 s on blockchain, versus around 2.1 s for baseline.

Discussion

The low deployment, recording, and querying times confirm the feasibility of our cross-cloud, blockchain-based federated data analysis system. We have yet to evaluate the system on a larger network with multiple nodes per cloud, to consider how to accommodate a surge in activities, and to investigate methods to lower querying time as the blockchain grows.

Conclusion

Blockchain technology can be used to support federated data analysis among multiple institutions.

Symposium: Workgroup on Interactive Systems in Healthcare (WISH)

(2023)

The Workgroup on Interactive Systems in Healthcare (WISH) connects academic and industry researchers across human-computer interaction, medical informatics, health informatics, digital health, and beyond to foster a community around innovations in consumer and medical health and wellbeing. The WISH Symposium at CHI 2023 will regather the HCI health and wellbeing research community for the first in-person community meeting in four years, allowing us to discuss and disseminate findings, methods, and approaches towards understanding and creating interactive health and wellbeing systems. We will continue the tradition of providing mentoring opportunities for early- and mid-career researchers, ranging from undergraduates to post-PhD, to establish future generations of scholars in the area. This will be the tenth WISH meeting, following a successful tradition of workshops at relevant venues including CHI over the past decade.

Cover page of Understanding the Benefits and Challenges of Deploying Conversational AI Leveraging Large Language Models for Public Health Intervention

Understanding the Benefits and Challenges of Deploying Conversational AI Leveraging Large Language Models for Public Health Intervention

(2023)

Recent large language models (LLMs) have advanced the quality of open-ended conversations with chatbots. Although LLM-driven chatbots have the potential to support public health interventions by monitoring populations at scale through empathetic interactions, their use in real-world settings is underexplored. We thus examine the case of CareCall, an open-domain chatbot that aims to support socially isolated individuals via check-up phone calls and monitoring by teleoperators. Through focus group observations and interviews with 34 people from three stakeholder groups, including the users, the teleoperators, and the developers, we found CareCall offered a holistic understanding of each individual while offloading the public health workload and helped mitigate loneliness and emotional burdens. However, our findings highlight that traits of LLM-driven chatbots led to challenges in supporting public and personal health needs. We discuss considerations of designing and deploying LLM-driven chatbots for public health intervention, including tensions among stakeholders around system expectations.

This Watchface Fits with my Tattoos: Investigating Customisation Needs and Preferences in Personal Tracking

(2023)

People engage in self-tracking with diverse data collection and visualisation needs and preferences. Customisable self-tracking tools offer the potential to support individualized preferences by letting people make changes to the aesthetics and functionality of tracker displays. In this paper, we use the customisation options offered by the displays of commercial fitness smartwatches as a lens to investigate when, why and how 386 self-trackers engage in customisations in their daily lives. We find that people largely customise their trackers' display frequently, multiple times a day, or not at all, with frequent customisations reflecting situational data, aesthetic and personal meaning needs. We discuss implications for the design of tracking tools aiming to support customisation and discuss the utility of customisations towards goal scaffolding and maintaining interest in tracking.