Efficacy of relational agents for loneliness across age groups: a systematic review and meta-analysis.

BACKGROUND: Loneliness is a serious public health concern. Although previous interventions have had some success in mitigating loneliness, the field is in search of novel, more effective, and more scalable solutions. Here, we focus on "relational agents", a form of software agents that are increasingly powered by artificial intelligence and large language models (LLMs). We report on a systematic review and meta-analysis to investigate the impact of relational agents on loneliness across age groups. METHODS: In this systematic review and meta-analysis, we searched 11 databases including Ovid MEDLINE and Embase from inception to Sep 16, 2022. We included randomised controlled trials and non-randomised studies of interventions published in English across all age groups. These loneliness interventions, typically attempt to improve social skills, social support, social interaction, and maladaptive cognitions. Peer-reviewed journal articles, books, book chapters, Master's and PhD theses, or conference papers were eligible for inclusion. Two reviewers independently screened studies, extracted data, and assessed risk of bias via the RoB 2 and ROBINS-I tools. We calculated pooled estimates of Hedge's g in a random-effects meta-analysis and conducted sensitivity and sub-group analyses. We evaluated publication bias via funnel plots, Egger's test, and a trim-and-fill algorithm. FINDINGS: Our search identified 3,935 records of which 14 met eligibility criteria and were included in our meta-analysis. Included studies comprised 286 participants with individual study sample sizes ranging from 4 to 42 participants (x̄ = 20.43, s = 11.58, x̃ = 20). We used a Bonferroni correction with αBonferroni = 0.05 / 4 = 0.0125 and applied Knapp-Hartung adjustments. Relational agents reduced loneliness significantly at an adjusted αBonferroni (g = -0.552; 95% Knapp-Hartung CI, -0.877 to -0.226; P = 0.003), which corresponds to a moderate reduction in loneliness. CONCLUSION: Our results are currently the most comprehensive of their kind and provide promising evidence for the efficacy of relational agents. Relational agents are a promising technology that can alleviate loneliness in a scalable way and that can be a meaningful complement to other approaches. The advent of LLMs should boost their efficacy, and further research is needed to explore the optimal design and use of relational agents. Future research could also address shortcomings of current results, such as small sample sizes and high risk of bias. Particularly young audiences have been overlooked in past research.

Development of a Virtual Human for Supporting Tobacco Cessation During the COVID-19 Pandemic.

People who consume tobacco are at greater risk of developing severe COVID-19. Unfortunately, the COVID-19 pandemic reduced the accessibility of tobacco cessation services as a result of necessary social restrictions. Innovations were urgently needed to support tobacco cessation during the pandemic. Virtual humans are artificially intelligent computer agents with a realistic, humanlike appearance. Virtual humans could be a scalable and engaging way to deliver tobacco cessation information and support. Florence, a virtual human health worker, was developed in collaboration with the World Health Organization to remotely support people toward tobacco cessation during the COVID-19 pandemic. Florence delivers evidence-based information, assists with making quit plans, and directs people to World Health Organization-recommended cessation services in their country. In this viewpoint, we describe the process of developing Florence. The development was influenced by a formative evaluation of data from 115 early users of Florence from 49 countries. In general, Florence was positively perceived; however, changes were requested to aspects of her design and content. In addition, areas for new content were identified (eg, for nonsmoker support persons). Virtual health workers could expand the reach of evidence-based tobacco cessation information and personalized support. However, as they are a new innovation in tobacco cessation, their efficacy, feasibility, and acceptability in this application needs to be evaluated, including in diverse populations.

The Impact of Virtual Humans on Psychosomatic Medicine.

OBJECTIVE: Virtual humans are likely to enhance the delivery of health care over the next decade. Virtual humans are artificially intelligent computer agents with hyperrealistic, autonomously animated embodiments based on affective computing techniques. Virtual humans could be programmed to screen for health conditions, triage patients, and deliver health interventions, with appropriate facial expressions and body gestures, functioning as a supplement to human care. This article provides a perspective on the implications of virtual humans for behavioral and psychosomatic medicine, and health psychology. METHODS: A narrative review was conducted to integrate observations and findings from research on virtual humans from 91 articles in this multidisciplinary area. RESULTS: Virtual humans can be used for multimodal behavior analysis of patients, individualized tailoring of interventions, and detection of changes of psychological and behavioral measures over time. Virtual humans can also pair the scalability of a website with the interactivity and relational skills of a human tele-therapist. Research is beginning to show the acceptability, feasibility, and preliminary effectiveness of virtual humans in a range of populations. Virtual humans can be easily tailored in terms of their appearance, voice, and language, and may be adapted to fit the characteristics of a patient population or hard-to-reach groups. If co-designed with these communities, virtual humans may help to promote health care engagement and improve outcomes. CONCLUSIONS: Virtual humans can engage and motivate patients, and deliver personalized psychological and behavioral health care. This article provides an overview of the potential impact of virtual humans on psychosomatic medicine and discusses ethical implications.

Comparing the Feasibility and Acceptability of a Virtual Human, Teletherapy, and an e-Manual in Delivering a Stress Management Intervention to Distressed Adult Women: Pilot Study.

BACKGROUND: Virtual humans (VHs), teletherapy, and self-guided e-manuals may increase the accessibility of psychological interventions. However, there is limited research on how these technologies compare in terms of their feasibility and acceptability in delivering stress management interventions. OBJECTIVE: We conducted a preliminary comparison of the feasibility and acceptability of a VH, teletherapy, and an e-manual at delivering 1 module of cognitive behavioral stress management (CBSM) to evaluate the feasibility of the trial methodology in preparation for a future randomized controlled trial (RCT). METHODS: A pilot RCT was conducted with a parallel, mixed design. A community sample of distressed adult women were randomly allocated to receive 1 session of CBSM involving training in cognitive and behavioral techniques by a VH, teletherapy, or an e-manual plus homework over 2 weeks. Data were collected on the feasibility of the intervention technologies (technical support and homework access), trial methods (recruitment methods, questionnaire completion, and methodological difficulty observations), intervention acceptability (intervention completion, self-report ratings, therapist rapport, and trust), and acceptability of the trial methods (self-report ratings and observations). Qualitative data in the form of written responses to open-ended questions were collected to enrich and clarify the findings on intervention acceptability. RESULTS: Overall, 38 participants' data were analyzed. A VH (n=12), teletherapy (n=12), and an e-manual (n=14) were found to be feasible and acceptable for delivering 1 session of CBSM to distressed adult women based on the overall quantitative and qualitative findings. Technical difficulties were minimal and did not affect intervention completion, and no significant differences were found between the conditions (P=.31). The methodology was feasible, although improvements were identified for a future trial. All conditions achieved good satisfaction and perceived engagement ratings, and no significant group differences were found (P>.40). Participants had similar willingness to recommend each technology (P=.64). There was a nonsignificant trend toward participants feeling more open to using the VH and e-manual from home than teletherapy (P=.10). Rapport (P<.001) and trust (P=.048) were greater with the human teletherapist than with the VH. The qualitative findings enriched the quantitative results by revealing the unique strengths and limitations of each technology that may have influenced acceptability. CONCLUSIONS: A VH, teletherapy, and a self-guided e-manual were found to be feasible and acceptable methods of delivering 1 session of a stress management intervention to a community sample of adult women. The technologies were found to have unique strengths and limitations that may affect which works best for whom and in what circumstances. Future research should test additional CBSM modules for delivery by these technologies and conduct a larger RCT to compare their feasibility, acceptability, and effectiveness when delivering a longer home-based stress management program. TRIAL REGISTRATION: Australian New Zealand Clinical Trials Registry ACTRN12620000859987; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=380114&isReview=true.

Artificial intelligence for older people receiving long-term care: a systematic review of acceptability and effectiveness studies.

Artificial intelligence (AI)-enhanced interventions show promise for improving the delivery of long-term care (LTC) services for older people. However, the research field is developmental and has yet to be systematically synthesised. This systematic review aimed to synthesise the literature on the acceptability and effectiveness of AI-enhanced interventions for older people receiving LTC services. We conducted a systematic search that identified 2720 records from Embase, Ovid, Global Health, PsycINFO, and Web of Science. 31 articles were included in the review that evaluated AI-enhanced social robots (n=22), environmental sensors (n=6), and wearable sensors (n=5) with older people receiving LTC services across 15 controlled and 14 non-controlled trials in high-income countries. Risk of bias was evaluated using the RoB 2, RoB 2 CRT, and ROBINS-I tools. Overall, AI-enhanced interventions were found to be somewhat acceptable to users with mixed evidence for their effectiveness across different health outcomes. The included studies were found to have high risk of bias which reduced confidence in the results. AI-enhanced interventions are promising innovations that could reshape the landscape of LTC globally. However, more trials are required to support their widespread implementation. Pathways are needed to support more high-quality trials, including in low-income and middle-income countries.

Effects of Emotional Expressiveness of a Female Digital Human on Loneliness, Stress, Perceived Support, and Closeness Across Genders: Randomized Controlled Trial.

BACKGROUND: Loneliness is a growing public health problem that has been exacerbated in vulnerable groups during the COVID-19 pandemic. Social support interventions have been shown to reduce loneliness, including when delivered through technology. Digital humans are a new type of computer agent that show promise as supportive peers in health care. For digital humans to be effective and engaging support persons, it is important that they develop closeness with people. Closeness can be increased by emotional expressiveness, particularly in female relationships. However, it is unknown whether emotional expressiveness improves relationships with digital humans and affects physiological responses. OBJECTIVE: The aim of this study is to investigate whether emotional expression by a digital human can affect psychological and physiological outcomes and whether the effects are moderated by the user's gender. METHODS: A community sample of 198 adults (101 women, 95 men, and 2 gender-diverse individuals) was block-randomized by gender to complete a 15-minute self-disclosure conversation with a female digital human in 1 of 6 conditions. In these conditions, the digital human varied in modality richness and emotional expression on the face and in the voice (emotional, neutral, or no face; emotional or neutral voice). Perceived loneliness, closeness, social support, caring perceptions, and stress were measured after each interaction. Heart rate, skin temperature, and electrodermal activity were assessed during each interaction. 3-way factorial analyses of variance with post hoc tests were conducted. RESULTS: Emotional expression in the voice was associated with greater perceptions of caring and physiological arousal during the interaction, and unexpectedly, with lower feelings of support. User gender moderated the effect of emotional expressiveness on several outcomes. For women, an emotional voice was associated with increased closeness, social support, and caring perceptions, whereas for men, a neutral voice increased these outcomes. For women, interacting with a neutral face was associated with lower loneliness and subjective stress compared with no face. Interacting with no face (ie, a voice-only black screen) resulted in lower loneliness and subjective stress for men, compared with a neutral or emotional face. No significant results were found for heart rate or skin temperature. However, average electrodermal activity was significantly higher for men while interacting with an emotional voice. CONCLUSIONS: Emotional expressiveness in a female digital human has different effects on loneliness, social, and physiological outcomes for men and women. The results inform the design of digital human support persons and have theoretical implications. Further research is needed to evaluate how more pronounced emotional facial expressions in a digital human might affect the results. TRIAL REGISTRATION: Australia New Zealand Clinical Trials Registry (ANZCTR) ACTRN12621000865819; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=381816&isReview.

A Digital Human for Delivering a Remote Loneliness and Stress Intervention to At-Risk Younger and Older Adults During the COVID-19 Pandemic: Randomized Pilot Trial.

BACKGROUND: Loneliness is a growing public health issue that has been exacerbated in vulnerable groups during the COVID-19 pandemic. Computer agents are capable of delivering psychological therapies through the internet; however, there is limited research on their acceptability to date. OBJECTIVE: The objectives of this study were to evaluate (1) the feasibility and acceptability of a remote loneliness and stress intervention with digital human delivery to at-risk adults and (2) the feasibility of the study methods in preparation for a randomized controlled trial. METHODS: A parallel randomized pilot trial with a mixed design was conducted. Participants were adults aged 18 to 69 years with an underlying medical condition or aged 70 years or older with a Mini-Mental State Examination score of >24 (ie, at greater risk of developing severe COVID-19). Participants took part from their place of residence (independent living retirement village, 20; community dwelling, 7; nursing home, 3). Participants were randomly allocated to the intervention or waitlist control group that received the intervention 1 week later. The intervention involved completing cognitive behavioral and positive psychology exercises with a digital human facilitator on a website for at least 15 minutes per day over 1 week. The exercises targeted loneliness, stress, and psychological well-being. Feasibility was evaluated using dropout rates and behavioral observation data. Acceptability was evaluated from behavioral engagement data, the Friendship Questionnaire (adapted), self-report items, and qualitative questions. Psychological measures were administered to evaluate the feasibility of the trial methods and included the UCLA Loneliness Scale, the 4-item Perceived Stress Scale, a 1-item COVID-19 distress measure, the Flourishing Scale, and the Scale of Positive and Negative Experiences. RESULTS: The study recruited 30 participants (15 per group). Participants were 22 older adults and 8 younger adults with a health condition. Six participants dropped out of the study. Thus, the data of 24 participants were analyzed (intervention group, 12; waitlist group, 12). The digital human intervention and trial methods were generally found to be feasible and acceptable in younger and older adults living independently, based on intervention completion, and behavioral, qualitative, and some self-report data. The intervention and trial methods were less feasible to nursing home residents who required caregiver assistance. Acceptability could be improved with additional content, tailoring to the population, and changes to the digital human's design. CONCLUSIONS: Digital humans are a promising and novel technological solution for providing at-risk adults with access to remote psychological support during the COVID-19 pandemic. Research should further examine design techniques to improve their acceptability in this application and investigate intervention effectiveness in a randomized controlled trial. TRIAL REGISTRATION: Australia New Zealand Clinical Trials Registry ACTRN12620000786998; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=380113.

Friends from the Future: A Scoping Review of Research into Robots and Computer Agents to Combat Loneliness in Older People.

BACKGROUND AND AIM: Loneliness is a common problem in older adults and contributes to poor health. This scoping review aimed to synthesize and report evidence on the effectiveness of interventions using social robots or computer agents to reduce loneliness in older adults and to explore intervention strategies. METHODS: The review adhered to the Arksey and O'Malley process for conducting scoping reviews. The SCOPUS, PUBMED, Web of Science, EMBASE, CINAHL, PsycINFO, ACM Digital Library and IEEE Xplore databases were searched in November, 2020. A two-step selection process identified eligible research. Information was extracted from papers and entered into an Excel coding sheet and summarised. Quality assessments were conducted using the Mixed Methods Appraisal Tool. RESULTS: Twenty-nine studies were included, of which most were of moderate to high quality. Eighteen were observational and 11 were experimental. Twenty-four used robots, four used computer agents and one study used both. The majority of results showed that robots or computer agents positively impacted at least one loneliness outcome measure. Some unintended negative consequences on social outcomes were reported, such as sadness when the robot was removed. Overall, the interventions helped to combat loneliness by acting as a direct companion (69%), a catalyst for social interaction (41%), facilitating remote communication with others (10%) and reminding users of upcoming social engagements (3%). CONCLUSION: Evidence to date suggests that robots can help combat loneliness in older adults, but there is insufficient research on computer agents. Common strategies for reducing loneliness include direct companionship and enabling social interactions. Future research could investigate other strategies used in human interventions (eg, addressing maladaptive social cognition and improving social skills), and the effects of design features on efficacy. It is recommended that more robust experimental and mixed methods research be conducted, using a combination of validated self-report, observational, and interview measures of loneliness.

The Effect of Multimodal Emotional Expression on Responses to a Digital Human during a Self-Disclosure Conversation: a Computational Analysis of User Language.

Digital humans show promise for use in healthcare as virtual therapists to deliver psychotherapy or companions for social support. For digital humans to be effective and engaging in these roles, it is important they can build close relationships with people. Emotional expressiveness can improve social closeness in human relationships, especially for females. However, it is unknown whether multimodal emotional expression improves relationships with digital humans. Participants were 185 adults aged 18 years or older with English fluency. Participants were block-randomized by gender to complete the Relationship Closeness Induction Task with one of six versions of a digital human. The digital humans varied in modality richness (face, no face) and emotional expression (emotional voice, neutral voice; emotional face, neutral face). Participants' language was analysed for emotional content using Linguistic Inquiry and Word Count software. A series of three-way ANOVA and ANCOVA were conducted to evaluate the effect of digital human face type, voice type, and participant gender on emotional content in participant language. A digital human with no face was associated with more first-person singular pronoun use than a neutral face and an emotional face digital human. A digital human with no face and a neutral voice received more general negative emotion language than a digital human with no face and an emotional voice. Findings suggest the presence of a face and emotion in the voice may improve emotional responses to digital humans. Results provide evidence for aspects of the theoretical framework of embodied agent-patient communication.

Reducing Patient Loneliness With Artificial Agents: Design Insights From Evolutionary Neuropsychiatry.

Loneliness is a growing public health issue that substantially increases the risk of morbidity and mortality. Artificial agents, such as robots, embodied conversational agents, and chatbots, present an innovation in care delivery and have been shown to reduce patient loneliness by providing social support. However, similar to doctor and patient relationships, the quality of a patient's relationship with an artificial agent can impact support effectiveness as well as care engagement. Incorporating mammalian attachment-building behavior in neural network processing as part of an agent's capabilities may improve relationship quality and engagement between patients and artificial agents. We encourage developers of artificial agents intended to relieve patient loneliness to incorporate design insights from evolutionary neuropsychiatry.

Enhancing treatment effectiveness through social modelling: A pilot study.

OBJECTIVE: Medical treatments take place in social contexts; however, little research has investigated how social modelling might influence treatment outcomes. This experimental pilot study investigated social modelling of treatment effectiveness and placebo treatment outcomes. DESIGN: Fifty-nine participants took part in the study, ostensibly examining the use of beta-blockers (actually placebos) for examination anxiety. Participants were randomly assigned to observe a female confederate report positive treatment effects (reduced heart rate, relaxed, calm) or feeling no different. MAIN OUTCOME MEASURES: Heart rate, anxiety and blood pressure were assessed, as were symptoms and attributed side effects. RESULTS: Heart rate decreased significantly more in the social modelling compared to control condition, p = .027 (d = .63), and there were trends towards effects in the same direction for both anxiety, p = .097 (d = .46), and systolic blood pressure, p = .077 (d = .51). Significant pre-post placebo differences in heart rate, anxiety and diastolic blood pressure were found in the social modelling group, ps < .007 (ds = .77-1.37), but not the control condition, ps > .28 (ds = .09-.59). CONCLUSIONS: Social observation of medication effectiveness enhanced placebo effectiveness in heart rate, and showed a trend towards enhancing treatment effectiveness in both anxiety and systolic blood pressure. Social modelling may have utility in enhancing the effectiveness of many active medical treatments.