Journal of Medical Internet Research最新文献

Background: Public attitudes toward health issues are becoming increasingly polarized, as seen in social media comments, which vary from supportive to oppositional and frequently include uncivil language. The combined effects of comment slant and comment tone on health behavior among a polarized public need further examination.

Objective: This study aims to examine how social media users' prior attitudes toward mask wearing and their exposure to a mask-wearing-promoting post, synchronized with polarized and hostile discussions, affect their compliance with mask wearing.

Methods: The study was a web-based survey experiment with participants recruited from Amazon Mechanical Turk. A total of 522 participants provided consent and completed the study. Participants were assigned to read a fictitious mask-wearing-promoting social media post with either civil anti-mask-wearing comments (130/522, 24.9%), civil pro-mask-wearing comments (129/522, 24.7%), uncivil anti-mask-wearing comments (131/522, 25.1%), or uncivil pro-mask-wearing comments (132/522, 25.3%). Following this, the participants were asked to complete self-assessed questionnaires. The PROCESS macro in SPSS (model 12; IBM Corp) was used to test the 3-way interaction effects between comment slant, comment tone, and prior attitudes on participants' presumed influence from the post and their behavioral intention to comply with mask-wearing.

Results: Anti-mask-wearing comments led social media users to presume less influence about others' acceptance of masks (B=1.49; P<.001; 95% CI 0.98-2.00) and resulted in decreased mask-wearing intention (B=0.07; P=.03; 95% CI 0.01-0.13). Comment tone with incivility also reduced compliance with mask-wearing (B=-0.44; P=.02; 95% CI -0.81 to -0.07). Furthermore, polarized attitudes had a direct impact (B=0.86; P<.001; 95% CI 0.45-1.26) and also interacted with both the slant and tone of comments, influencing mask-wearing intention.

Conclusions: Pro-mask-wearing comments enhanced presumed influence and compliance of mask-wearing, but incivility in the comments hindered this positive impact. Antimaskers showed increased compliance when they were unable to find civil support for their opinion in the social media environment. The findings suggest the need to correct and moderate uncivil language and misleading information in online comment sections while encouraging the posting of supportive and civil comments. In addition, information literacy programs are needed to prevent the public from being misled by polarized comments.

Background: Recent studies have identified significant gaps in equity, diversity, and inclusion (EDI) considerations within the lifecycle of artificial intelligence (AI), spanning from data collection and problem definition to implementation stages. Despite the recognized need for integrating EDI principles, there is currently no existing guideline or framework to support this integration in the AI lifecycle.

Objective: This study aimed to address this gap by identifying EDI principles and indicators to be integrated into the AI lifecycle. The goal was to develop a comprehensive guiding framework to guide the development and implementation of future AI systems.

Methods: This study was conducted in 3 phases. In phase 1, a comprehensive systematic scoping review explored how EDI principles have been integrated into AI in health and oral health care settings. In phase 2, a multidisciplinary team was established, and two 2-day, in-person international workshops with over 60 representatives from diverse backgrounds, expertise, and communities were conducted. The workshops included plenary presentations, round table discussions, and focused group discussions. In phase 3, based on the workshops' insights, the EDAI framework was developed and refined through iterative feedback from participants. The results of the initial systematic scoping review have been published separately, and this paper focuses on subsequent phases of the project, which is related to framework development.

Results: In this study, we developed the EDAI framework, a comprehensive guideline that integrates EDI principles and indicators throughout the entire AI lifecycle. This framework addresses existing gaps at various stages, from data collection to implementation, and focuses on individual, organizational, and systemic levels. Additionally, we identified both the facilitators and barriers to integrating EDI within the AI lifecycle in health and oral health care.

Conclusions: The developed EDAI framework provides a comprehensive, actionable guideline for integrating EDI principles into AI development and deployment. By facilitating the systematic incorporation of these principles, the framework supports the creation and implementation of AI systems that are not only technologically advanced but also sensitive to EDI principles.

Background: Digital technologies offer the potential for low-cost, scalable delivery of interventions to promote smoking cessation.

Objective: We aimed to evaluate the effectiveness of the offer of Smoke Free-an evidence-informed, widely used app-for smoking cessation versus no support.

Methods: In this 2-arm randomized controlled trial, 3143 motivated adult smokers were recruited online between August 2020 and April 2021 and randomized to receive an offer of the Smoke Free app plus follow-up (intervention arm) versus follow-up only (comparator arm). Both groups were shown a brief message at the end of the baseline questionnaire encouraging them to make a quit attempt. The primary outcome was self-reported 6-month continuous abstinence assessed 7 months after randomization. Secondary outcomes included quit attempts in the first month post randomization, 3-month continuous abstinence assessed at 4 months, and 6-month continuous abstinence at 7 months among those who made a quit attempt. The primary analysis was performed on an intention-to-treat (ITT) analysis basis. Sensitivity analyses included (1) restricting the intervention group to those who took up the offer of the app, (2) using complete cases, and (3) using multiple imputation.

Results: The effective follow-up rate for 7 months was 41.9%. The primary analysis showed no evidence of a benefit of the intervention on rates of 6-month continuous abstinence (intervention 6.8% vs comparator 7.0%; relative risk 0.97, 95% CI 0.75-1.26). Analyses of all secondary outcomes also showed no evidence of a benefit. Similar results were observed on complete cases and using multiple imputation. When the intervention group was restricted to those who took up the offer of the app (n=395, 25.3%), participants in the intervention group were 80% more likely to report 6-month continuous abstinence (12.7% vs 7.0%; relative risk 1.80, 95% CI 1.30-2.45). Equivalent subgroup analyses produced similar results on the secondary outcomes. These differences persisted after adjustment for key baseline characteristics.

Conclusions: Among motivated smokers provided with very brief advice to quit, the offer of the Smoke Free app did not have a detectable benefit for cessation compared with follow-up only. However, the app increased quit rates when smokers randomized to receive the app downloaded it.

Trial registration: ISRCTN ISRCTN85785540; https://www.isrctn.com/ISRCTN85785540.

International registered report identifier (irrid): RR2-https://onlinelibrary.wiley.com/doi/full/10.1111/add.14652.

Background: As crowdfunding sites proliferate, visual content often serves as the initial bridge connecting a project to its potential backers, underscoring the importance of image selection in effectively engaging an audience.

Objective: This paper aims to explore the relationship between images and crowdfunding success in cancer-related crowdfunding projects.

Methods: We used the Alibaba Cloud platform to detect individual features in images. In addition, we used the Recognize Anything Model to label images and obtain content tags. Furthermore, the discourse atomic topic model was used to generate image topics. After obtaining the image features and image content topics, we built regression models to investigate the factors that influence the results of crowdfunding success.

Results: Images with a higher proportion of young people (β=0.0753; P<.001), a larger number of people (β=0.00822; P<.001), and a larger proportion of smiling faces (β=0.0446; P<.001) had a higher success rate. Image content related to good things and patient health also contributed to crowdfunding success (β=0.082, P<.001; and β=0.036, P<.001, respectively). In addition, the interaction between image topics and image characteristics had a significant effect on the final fundraising outcome. For example, when smiling faces are considered in conjunction with the image topics, using more smiling faces in the rest and play theme increased the amount of money raised (β=0.0152; P<.001). We also examined causality through a counterfactual analysis, which confirmed the influence of the variables on crowdfunding success, consistent with the results of our regression models.

Conclusions: In the realm of web-based medical crowdfunding, the importance of uploaded images cannot be overstated. Image characteristics, including the number of people depicted and the presence of youth, significantly improve fundraising results. In addition, the thematic choice of images in cancer crowdfunding efforts has a profound impact. Images that evoke beauty and resonate with health issues are more likely to result in increased donations. However, it is critical to recognize that reinforcing character traits in images of different themes has different effects on the success of crowdfunding campaigns.

[This corrects the article DOI: 10.2196/54800.].

Background: Telemedicine is expanding rapidly, with public direct-to-consumer (DTC) telemedicine representing 70% of the market. A key priority is establishing clear quality distinctions between the public and private sectors. No studies have directly compared the quality of DTC telemedicine in the public and private sectors using objective evaluation methods.

Objective: Using a standardized patient (SP) approach, this study aimed to compare the quality of DTC telemedicine provided by China's public and private sectors.

Methods: We recruited 10 SPs presenting fixed cases (urticaria and childhood diarrhea), with 594 interactions between them and physicians. The SPs evaluated various aspects of the quality of care, effectiveness, safety, patient-centeredness (PCC), efficiency, and timeliness using the Institute of Medicine (IOM) quality framework. Ordinary least-squares (OLS) regression models with fixed effects were used for continuous variables, while logistic regression models with fixed effects were used for categorical variables.

Results: Significant quality differences were observed between public and private DTC telemedicine. Physicians from private platforms were significantly more likely to adhere to clinical checklists (adjusted β 15.22, P<.001); provide an accurate diagnosis (adjusted odds ratio [OR] 3.85, P<.001), an appropriate prescription (adjusted OR 3.87, P<.001), and lifestyle modification advice (adjusted OR 6.82, P<.001); ensure more PCC (adjusted β 3.34, P<.001); and spend more time with SPs (adjusted β 839.70, P<.001), with more responses (adjusted β 1.33, P=.001) and more words (adjusted β 50.93, P=.009). However, SPs on private platforms waited longer for the first response (adjusted β 505.87, P=.001) and each response (adjusted β 168.33, P=.04) and paid more for the average visit (adjusted β 40.03, P<.001).

Conclusions: There is significant quality inequality in different DTC telemedicine platforms. Private physicians might provide a higher quality of service regarding effectiveness and safety, PCC, and response times and words. However, private platforms have longer wait times for their first response, as well as higher costs. Refining online reviews, establishing standardized norms and pricing, enhancing the performance evaluation mechanism for public DTC telemedicine, and imposing stricter limitations on the first response time for private physicians should be considered practical approaches to optimizing the management of DTC telemedicine.

Background: Artificial intelligence (AI) chatbots such as ChatGPT are expected to impact vision health care significantly. Their potential to optimize the consultation process and diagnostic capabilities across range of ophthalmic subspecialties have yet to be fully explored.

Objective: This study aims to investigate the performance of AI chatbots in recommending ophthalmic outpatient registration and diagnosing eye diseases within clinical case profiles.

Methods: This cross-sectional study used clinical cases from Chinese Standardized Resident Training-Ophthalmology (2nd Edition). For each case, 2 profiles were created: patient with history (Hx) and patient with history and examination (Hx+Ex). These profiles served as independent queries for GPT-3.5 and GPT-4.0 (accessed from March 5 to 18, 2024). Similarly, 3 ophthalmic residents were posed the same profiles in a questionnaire format. The accuracy of recommending ophthalmic subspecialty registration was primarily evaluated using Hx profiles. The accuracy of the top-ranked diagnosis and the accuracy of the diagnosis within the top 3 suggestions (do-not-miss diagnosis) were assessed using Hx+Ex profiles. The gold standard for judgment was the published, official diagnosis. Characteristics of incorrect diagnoses by ChatGPT were also analyzed.

Results: A total of 208 clinical profiles from 12 ophthalmic subspecialties were analyzed (104 Hx and 104 Hx+Ex profiles). For Hx profiles, GPT-3.5, GPT-4.0, and residents showed comparable accuracy in registration suggestions (66/104, 63.5%; 81/104, 77.9%; and 72/104, 69.2%, respectively; P=.07), with ocular trauma, retinal diseases, and strabismus and amblyopia achieving the top 3 accuracies. For Hx+Ex profiles, both GPT-4.0 and residents demonstrated higher diagnostic accuracy than GPT-3.5 (62/104, 59.6% and 63/104, 60.6% vs 41/104, 39.4%; P=.003 and P=.001, respectively). Accuracy for do-not-miss diagnoses also improved (79/104, 76% and 68/104, 65.4% vs 51/104, 49%; P<.001 and P=.02, respectively). The highest diagnostic accuracies were observed in glaucoma; lens diseases; and eyelid, lacrimal, and orbital diseases. GPT-4.0 recorded fewer incorrect top-3 diagnoses (25/42, 60% vs 53/63, 84%; P=.005) and more partially correct diagnoses (21/42, 50% vs 7/63 11%; P<.001) than GPT-3.5, while GPT-3.5 had more completely incorrect (27/63, 43% vs 7/42, 17%; P=.005) and less precise diagnoses (22/63, 35% vs 5/42, 12%; P=.009).

Conclusions: GPT-3.5 and GPT-4.0 showed intermediate performance in recommending ophthalmic subspecialties for registration. While GPT-3.5 underperformed, GPT-4.0 approached and numerically surpassed residents in differential diagnosis. AI chatbots show promise in facilitating ophthalmic patient registration. However, their integration into diagnostic decision-making requires more validation.