Evaluation of the Current State of Chatbots for Digital Health: Scoping Review

Xue, Jia; Zhang, Bolun; Zhao, Yaxi; Zhang, Qiaoru; Zheng, Chengda; Jiang, Jielin; Li, Hanjia; Liu, Nian; Li, Ziqian; Fu, Weiying; Peng, Yingdong; Logan, Judith; Zhang, Jingwen; Xiang, Xiaoling

doi:10.2196/47217

Cited by 11 publications

(2 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To date, there have been no reviews examining the integration of LLMs in research on suicide prevention and self-harm. Prior reviews have focused predominantly on research undertaken in adjacent fields, such as applications of machine learning or chatbots to mental health more broadly [19][20][21][22]. Although several reviews have explored the use of machine learning [23][24][25][26] or AI-based strategies [27][28][29][30] in suicide prevention contexts, these have not focused on applications of LLMs specifically.…”

Section: Discussionmentioning

confidence: 99%

Applications of Large Language Models in the Field of Suicide Prevention: A Scoping Review (Preprint)

Holmes,

Tang,

Gupta

et al. 2024

Preprint

View full text Add to dashboard Cite

BACKGROUND Prevention of suicide is a global health priority. Around 800,000 individuals die by suicide yearly, and for every death, there are another 20 estimated suicide attempts. Large language models (LLMs) hold the potential to enhance scalable, accessible, and affordable digital services for suicide prevention and self-harm interventions. However, their use also raises clinical and ethical questions that require careful consideration. OBJECTIVE This scoping review aimed to identify emergent trends in applications of LLMs within the field of suicide and self-harm research. Additionally, it summarizes key clinical and ethical considerations relevant to this nascent area of research. METHODS Searches were conducted in four databases. Eligible studies described the application of LLMs for suicide or self-harm prevention, detection, or management. English-language peer-reviewed articles and conference proceedings were included, with no date restrictions. This review adhered to PRISMA-ScR standards. RESULTS Of the 533 studies identified, 36 met inclusion criteria, and an additional 7 more were identified through citation chaining, resulting in a total of 43 studies for review. A narrative synthesis approach was used to synthesize study characteristics, objectives, models, data sources, proposed clinical applications, and ethical considerations. Studies showed a bifurcation of publication fields with varying publication norms between computer science and mental health. While most studies (77%) focused on identifying suicide risk, newer applications leveraging generative functions (e.g., support, education, and training) are emerging. Social media was the most common source of LLM training data. BERT (Bidirectional Encoder Representation Transformer) was the predominant model used, although GPT (Generative Pre-trained Transformer) featured prominently in generative applications. Clinical applications of LLMs were reported in 60% of studies, often for suicide risk detection or as clinical assistance tools. Ethical considerations were reported in 33% of studies, with privacy, confidentiality, and consent strongly represented. CONCLUSIONS This evolving research area, bridging computer science and mental health, demands a multi-disciplinary approach. While open access models and datasets will likely shape this field, documenting their limitations and potential biases is crucial. High-quality training data is essential for refining these models and mitigating unwanted biases. Policies that address ethical concerns – particularly related to privacy and security when using social media data – are imperative. The emergence of generative AI signals a shift in approach, particularly in applications related to care, support, and education. Ongoing human oversight, whether through human-in-the-loop testing or expert external validation, is essential for responsible development and use.

show abstract

Section: Discussionmentioning

confidence: 99%

Applications of Large Language Models in the Field of Suicide Prevention: A Scoping Review (Preprint)

Holmes,

Tang,

Gupta

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…However, to our knowledge, no study has yet attempted to comprehensively map out the heterogeneous landscape of digital public health (DiPH) and digital public health interventions (DiPHIs). Despite existing reviews focusing on selected intervention categories [ 17 ], specific target groups such as adolescents [ 18 , 19 ] or patients who are chronically ill with specific medical conditions [ 20 ], a comprehensive mapping of interventions in digital health is still missing. In addition, in all existing reviews, the studies analyzed digital health interventions but not DiPHIs, thereby limiting these interventions to their medical and clinical usability.…”

Section: Introductionmentioning

confidence: 99%

Mapping Digital Public Health Interventions Among Existing Digital Technologies and Internet-Based Interventions to Maintain and Improve Population Health in Practice: Scoping Review

Maaß,

Angoumis,

Freye

et al. 2024

J Med Internet Res

View full text Add to dashboard Cite

Background:The rapid progression and integration of digital technologies into public health have reshaped the global landscape of healthcare delivery and disease prevention. In pursuit of better population health and healthcare accessibility, many countries have integrated digital interventions into their healthcare systems, such as online consultations, electronic health records, and telemedicine. Despite the increasing prevalence and relevance of digital technologies in public health and their varying definitions, there has been a shortage of studies examining whether these technologies align with the established definition and core characteristics of digital public health (DiPH) interventions. Hence, the imperative need for a scoping review emerges to explore the breadth of literature dedicated to this subject.Objective: This scoping review aims to outline DiPH interventions from different implementation stages for health promotion, primary to tertiary prevention, including healthcare and disease surveillance and monitoring. Additionally, we aim to map the reported intervention characteristics, including their technical features and non-technical elements.Methods: Original studies or reports of DiPH intervention focused on population health were eligible for this review. PubMed, Web of Science, the Cochrane Central Register of Controlled Trials, IEEE Xplore, and the ACM Full-text collection were searched for relevant literature (last updated October 5, 2022). Intervention characteristics of each identified DiPH intervention, such as target groups, level of prevention or healthcare, digital health functions, intervention types, and public health functions, were extracted and used to map DiPH interventions. MAXQDA 2022.7 was used for qualitative data analysis of such interventions' technical functions and non-technical characteristics. Results:In total, we identified and screened 15,701 records. Of these, 1,562 full-texts were considered relevant and were assessed for eligibility. Finally, we included 185 references, which reported 179 different DiPH interventions. Our analysis revealed a diverse landscape of interventions, with telemedical services, health apps, and electronic health records as dominant types. These interventions target a wide range of populations and settings, demonstrating their adaptability. The analysis highlighted the multifaceted nature of digital interventions, necessitating precise definitions and standardized terminologies for effective collaboration and evaluation.

show abstract

Standardizing and Scaffolding Healthcare AI-Chatbot Evaluation

Hua,

Xia,

Bates

et al. 2024

Preprint

View full text Add to dashboard Cite

The rapid rise of healthcare chatbots, valued at $787.1 million in 2022 and projected to grow at 23.9% annually through 2030, underscores the need for robust evaluation frameworks. Despite their potential, the absence of standardized evaluation criteria and rapid AI advancements complicate assessments. This study addresses these challenges by developing the first comprehensive evaluation framework inspired by health app regulations and integrating insights from diverse stakeholders. Following PRISMA guidelines, we reviewed 11 existing frameworks, refining 271 questions into a structured framework encompassing three priority constructs, 18 second-level constructs, and 60 third-level constructs. Our framework emphasizes safety, privacy, trustworthiness, and usefulness, aligning with recent concerns about AI in healthcare. This adaptable framework aims to serve as the initial step in facilitating the responsible integration of chatbots into healthcare settings.

show abstract

Evaluation of the Current State of Chatbots for Digital Health: Scoping Review

Cited by 11 publications

References 39 publications

Applications of Large Language Models in the Field of Suicide Prevention: A Scoping Review (Preprint)

Applications of Large Language Models in the Field of Suicide Prevention: A Scoping Review (Preprint)

Mapping Digital Public Health Interventions Among Existing Digital Technologies and Internet-Based Interventions to Maintain and Improve Population Health in Practice: Scoping Review

Standardizing and Scaffolding Healthcare AI-Chatbot Evaluation

Contact Info

Product

Resources

About