Background: Trust in artificial intelligence (AI) remains a critical barrier to the adoption of AI in mental health care. This study explores the formation of trust in an AI mental health model and its human-computer interface among clinicians at a web-based mental health clinic in the Region of Southern Denmark with national coverage.
Objective: This study aims to explore clinicians' perspectives on how trust is built in the context of an AI-supported mental health screening model and to identify the factors that influence this process.
Methods: This was a qualitative case study using semistructured interviews with clinicians involved in the pilot of a mental health AI model. Thematic analysis was used to identify key factors contributing to trust formation.
Results: Clinicians' initial attitudes toward AI were shaped by prior positive experiences with AI and their perception of AI's potential to reduce cognitive load in routine screening. Trust development followed a sequential pattern resembling a "trust journey": (1) sense-making, (2) risk appraisal, and (3) conditional decision to rely. Trust formation was supported by the explainability of the model, particularly through (1) visualization of confidence and uncertainty through violin plots, aligning with the clinicians' expectations of decision ambiguity; (2) feature attribution for and against predictions, which mirrored clinical reasoning; and (3) use of pseudo-sumscores in the AI model, which increased interpretability by presenting explanations in familiar clinical formats. Trust was contextually bounded to low-risk clinical scenarios, such as preinterview patient screening, and contingent on safety protocols (eg, suicide risk flagging). The use of both structured and unstructured patient data was seen as key to expanding trust into more complex clinical contexts. Participants also expressed a need for ongoing evaluation data to reinforce and maintain trust.
Conclusions: Clinicians' trust in AI tools is contextually and sequentially constructed, influenced by both model performance and alignment with clinical reasoning. Interpretability features were essential in establishing intrinsic trust, particularly when presented in ways that resonate with clinical norms. For broader acceptance and responsible deployment, trust must be supported by rigorous evaluation data and the inclusion of clinically relevant data types in model design.
Background: Technology is rapidly reshaping conventional hospital environments into smart spaces, enhancing care, improving clinical workflows, and reducing workloads. However, successful implementation depends not only on the effectiveness of the technology but also on organizational readiness for change.
Objective: This study aimed to identify the key enablers and barriers to readiness for change for a smart hospital ward initiative.
Methods: We conducted a qualitative study to gauge organizational readiness for change for a smart ward initiative. Using purposive sampling, we captured diverse views from clinicians, IT staff, operational support staff, and health care redesign staff. Data were coded deductively under 3 key domains in Weiner's theory of organizational readiness: change efficacy, change commitment, and contextual factors. Subthemes were derived inductively under each domain.
Results: We interviewed 19 participants, including clinicians and support staff. Six subthemes emerged: (1) perceived valence and feasibility; (2) transparency and trust in management; (3) shared understanding and readiness to act; (4) resources, training, and staff capability; (5) innovation culture; and (6) past experiences. Participants viewed the initiative as valuable and were motivated to change, citing that the institution's innovation culture was a key enabler. However, there were key barriers, including unclear timelines, inconsistent training, limited resources, and a lack of infrastructure to support innovation. Concerns about overreliance on technology were also prominent, with staff wary of its impact on clinical judgment and system reliability.
Conclusions: Enabling readiness for the smart ward initiative requires transparent communication of timelines and project awareness, particularly for ground staff, the development of training frameworks, and adequate prioritization of innovation. Alleviating commonly reported technology concerns, such as overreliance, loss of human touch, and system reliability, will also be key to adoption and sustainability.
Background: Health messages are integral to smoking cessation interventions. Common approaches to health message development include expert-crafted messages and audience-generated messages, which produce messages that can be monotonic, didactic, and limited in number. We introduce an alternative approach to health message development that relies on user-generated content available on open-content platforms as a source of health messages.
Objective: We examined the acceptability of user-generated content curated from Twitter (subsequently rebranded X) as a source of health support messages in a newly developed smoking cessation mobile intervention called Quit Journey and the optimal timing and frequency with which health messages can be deployed to support app users in real time.
Methods: A total of 12 semistructured focus groups were held with 38 young adults with low socioeconomic status who smoked cigarettes, wanted to quit, and were aged 18 to 29 years. Focus groups were held virtually on GoTo Meeting, audio recorded, and transcribed verbatim. Deductive thematic analysis was used, with themes based on 5 constructs from the second unified theory of acceptance and use of technology (ie, effort expectancy, facilitating conditions, hedonic motivation, performance expectancy, and social influence) and negative, neutral, and positive sentiment.
Results: Participants perceived user-generated content positively (56/108, 51.9% of the quotes) and focused on their perceived usefulness (37/108, 34.3% of the quotes). User-generated content was perceived as authentic, nonrepetitive support from people with similar real-life experiences. Negative or sarcastic user-generated content elicited negative reactions from participants. Participants preferred receiving 3 or fewer daily messages, ideally before cravings. Suggestions focused on the need to screen user-generated content before its inclusion in the app library and allow app users to customize message frequency and timing.
Conclusions: User-generated content was deemed an acceptable source of health messages. This content can improve the efficacy and effectiveness of smoking cessation interventions by increasing their pool of unique messages that may be better received and more persuasive than expert-curated content. User-generated content can be used to curate health messages for all medical conditions and behaviors with relevant publicly available online content for integration in behavioral interventions given its high volume, brevity, and narrative-like nature. Future research is needed to investigate the effects of user-generated content on health behaviors and identify the theoretical mechanisms for these effects.
Background: Needleless access devices are essential for intravenous therapy but can be a source of contamination and catheter-related bloodstream infections (CRBSIs) if not disinfected properly. The BD PosiFlush SafeScrub (Becton, Dickinson and Company) is designed to support aseptic nontouch technique (ANTT) by incorporating a built-in reminder to "scrub-the-hub" before flushing. This feature can help improve compliance with disinfection practices and may reduce the risk of microbial contamination.
Objective: This study aimed to evaluate compliance with scrubbing before flushing using the BD PosiFlush SafeScrub in a simulated clinical environment compared with standard disinfection and flushing practices (alcohol swabs and prefilled saline syringes).
Methods: A cross-sectional, within-subjects, simulated-use compliance study was conducted with health care professionals familiarized with BD PosiFlush SafeScrub (a prefilled BD PosiFlush Syringe with an integrated disinfecting unit). Compliance was defined according to the disinfection procedure specified for each scenario; participants were considered compliant with standard practice if they followed their own institutional policy (ranging from 5 to 30 seconds or based on stroke counts), while compliance with BD PosiFlush SafeScrub required scrubbing for at least 10 seconds with a minimum of 8 clockwise and 8 counterclockwise rotations, in accordance with the instructions for use (IFU). Compliance with disinfection was monitored and recorded for both the BD PosiFlush SafeScrub and standard disinfection and flushing practice.
Results: Compliance with catheter hub disinfection was assessed among 60 participants for BD PosiFlush SafeScrub and 57 participants for standard practice. During preaccess procedures (Flush 1), BD PosiFlush SafeScrub achieved 46% compliance versus 21% with standard practice, representing an absolute improvement of 25% and a 119% relative improvement (P<.001). During the postmedication procedure (Flush 2), compliance was 22% with BD PosiFlush SafeScrub compared with 13% for standard practice, corresponding to a 9% absolute improvement and 69% relative improvement, although not statistically significant (P=.12). Overall, the compliance rate was 34% (81/240 interactions) in the BD PosiFlush SafeScrub group compared with 17% (39/228 interactions) in the standard practice group, representing an absolute improvement of 17% and a relative improvement of 100% (P<.001).
Conclusions: The BD PosiFlush SafeScrub, with its integrated disinfection unit, yielded approximately double the scrub-the-hub compliance (34%) before flushing compared to the standard practice of alcohol pads and prefilled saline syringes (17%), supporting its role in facilitating adherence to ANTT, which may reduce microbial growth.
Background: Internet-based cognitive behavioral therapy (CBT) provides psychological interventions to individuals with mild depressive symptoms.
Objective: This study aimed to examine the potential cost-effectiveness of internet-based guided-CBT in university students with mild depressive symptoms from the perspective of service providers in Hong Kong.
Methods: The outcomes of low-intensity guided internet-based CBT and in-person CBT in a hypothetical cohort of university students with mild depressive symptoms were examined using a 5-year decision-analytic model. Model inputs were obtained from published literature and local data. Model outcomes included direct medical cost, school dropouts, and quality-adjusted life years (QALYs). Sensitivity analyses were conducted on all model parameters.
Results: Compared to the in-person group, the internet group gained higher QALYs by 0.0211 QALYs, lowered school dropouts by 0.052%, and saved US $249 in the base-case analysis. In one-way sensitivity analysis, the internet group gained higher QALYs at a lower cost than the in-person group throughout the variation of all model inputs. Probabilistic sensitivity analysis showed that the internet group was cost-effective (at willingness-to-pay threshold was US $48,119/QALY) in 99.7% of the 10,000 Monte Carlo simulations.
Conclusions: Internet-based CBT appears to be the cost-effective option when compared to in-person CBT for university students with mild depressive symptoms from the perspective of service providers in Hong Kong.
Background: Digital mental health tools, such as text messaging and online resources, are increasingly utilized to support well-being. However, user satisfaction across these formats remains insufficiently explored.
Objective: The study assessed participants' engagement, perceived impact, and overall satisfaction with the Text4Support program and the e-mental health resources.
Methods: This randomized controlled study was conducted in Nova Scotia, Canada. Participants were assigned to either the Text4Support group, which received daily supportive text messages, or the Control group, which received a single text message with a link to the Nova Scotia Health Mental Health and Addiction Program e-mental health resources. Responses to various aspects of the interventions were evaluated using a 5-point Likert scale, while overall satisfaction was measured on a scale from 0 to 10. The chi-square test and Fisher's exact test were employed for data analysis.
Results: A total of 69 in the control group and 130 in the Text4Support group completed the satisfaction survey. The overall mean satisfaction score in the control group was 5.1 (SD 2.3), and the overall mean satisfaction score for the Text4Support group was 7.1 (SD 2.2). Compared to the control 3 group, participants in the Text4Support group reported greater engagement and positive program impact. Whereas 53% of Text4Support recipients always read the messages, only 39.1% of the control group sometimes accessed the e-health resources. Participants allocated to the Text4Support group were reported to sometimes take positive action upon reading the messages (42.3% vs. 33.3%). A significantly higher proportion of Text4Support users strongly agreed or agreed that the messages were supportive (81.4% vs 41.5%), positive (88.4% vs 49.2%), and helpful in coping with stress (44.2% vs 11.9%), loneliness (40.3% vs 13.4%), and improving mental well-being (51.2% vs 17.9%). In contrast, the majority of responses from the control group were largely neutral.
Conclusions: Results showed that Text4Support group participants were significantly more satisfied with the program than those receiving standard e-health resources. This highlights that daily supportive text messaging is an effective, low-cost adjunct to care delivery and mental health improvement. These findings suggest that aggregate, brief, and low-cost text-based interventions have great potential for increasing health access and engagement, particularly among traditionally disadvantaged populations with limited access to traditional services.
Clinicaltrial:
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：