Art or Artifact: Evaluating the Accuracy, Appeal, and Educational Value of AI-Generated Imagery in DALL·E 3 for Illustrating Congenital Heart Diseases.

Artificial Intelligence (AI), particularly AI-Generated Imagery, has the potential to impact medical and patient education. This research explores the use of AI-generated imagery, from text-to-images, in medical education, focusing on congenital heart diseases (CHD). Utilizing ChatGPT's DALL·E 3, the research aims to assess the accuracy and educational value of AI-created images for 20 common CHDs. In this study, we utilized DALL·E 3 to generate a comprehensive set of 110 images, comprising ten images depicting the normal human heart and five images for each of the 20 common CHDs. The generated images were evaluated by a diverse group of 33 healthcare professionals. This cohort included cardiology experts, pediatricians, non-pediatric faculty members, trainees (medical students, interns, pediatric residents), and pediatric nurses. Utilizing a structured framework, these professionals assessed each image for anatomical accuracy, the usefulness of in-picture text, its appeal to medical professionals, and the image's potential applicability in medical presentations. Each item was assessed on a Likert scale of three. The assessments produced a total of 3630 images' assessments. Most AI-generated cardiac images were rated poorly as follows: 80.8% of images were rated as anatomically incorrect or fabricated, 85.2% rated to have incorrect text labels, 78.1% rated as not usable for medical education. The nurses and medical interns were found to have a more positive perception about the AI-generated cardiac images compared to the faculty members, pediatricians, and cardiology experts. Complex congenital anomalies were found to be significantly more predicted to anatomical fabrication compared to simple cardiac anomalies. There were significant challenges identified in image generation. Based on our findings, we recommend a vigilant approach towards the use of AI-generated imagery in medical education at present, underscoring the imperative for thorough validation and the importance of collaboration across disciplines. While we advise against its immediate integration until further validations are conducted, the study advocates for future AI-models to be fine-tuned with accurate medical data, enhancing their reliability and educational utility.

Effects of Computerised Clinical Decision Support on Adherence to VTE Prophylaxis Clinical Practice Guidelines among Hospitalised Patients.

OBJECTIVE: Venous thromboembolism (VTE) is an important patient safety concern. VTE leads to significant mortality and morbidity and a burden on healthcare resources. Despite the widespread availability of evidence-based clinical practice guidelines on VTE prophylaxis, we found that only 50.9% of our patients were receiving appropriate prophylaxis. The purpose of this study was to evaluate the impact of automation of an adapted VTE prophylaxis CPG using a clinical decision support system (the VTE-CDSS) on VTE prevention among hospitalised adult patients. DESIGN AND SETTING: A quasi-experimental study (pre- and post-implementation) was conducted at a large 900-bed tertiary teaching multi-specialty hospital in Riyadh, Saudi Arabia. PARTICIPANTS: The 1809 adult patients in the study included 871 enrolled during the pre-implementation stage and 938 enrolled during the post-implementation stage. INTERVENTION: Multi-faceted implementation interventions were utilised, including leadership engagement and support, quality and clinical champions, staff training and education and regular audit and feedback. MAIN OUTCOME MEASURE: Two rate-based process measures were calculated for each admission cohort (i.e. pre- and post-implementation): the percentage of inpatients who were assessed for VTE risk on admission and the percentage of inpatients who received appropriate VTE prophylaxis. Two outcome measures were calculated: the prevalence of hospital-acquired VTE (HA-VTE) events and the in-hospital all-cause mortality. RESULTS: The percentage of inpatients risk assessed for VTE on admission increased from 77.4% to 93.3% (P < 0.01). The percentage of patients who received appropriate VTE prophylaxis increased from 50.9% to 81.4% (P < 0.01). The HA-VTE events decreased by 50% from 0.33% to 0.15% (P < 0.01).All-cause in-hospital mortality did not significantly change after implementation of the VTE-CDSS compared with pre-implementation mortality (P > 0.05). CONCLUSION: The VTE-CDSS improved patient safety by enhancing adherence to the VTE prophylaxis best practice and adapted CPG. The multi-faceted implementation strategies approach improved the compliance rate of risk assessment and the adherence to prophylaxis recommendations and substantially reduced the HA-VTE prevalence. A successful CDSS requires a set of critical components to ensure better user compliance and positive patient outcomes.

Remote interviews for medical residency selection during the initial COVID-19 crisis: a national survey.

BACKGROUND: Medical training programs candidate's interview is an integral part of the residency matching process. During the coronavirus disease 2019 (COVID-19) pandemic, conducting these interviews was challenging due to infection prevention restrains (social distancing, namely) and travel restrictions. E-interviews were implemented by the Saudi Commission for Healthcare Specialties (SCFHS) since the matching cycle of March 2020 to hold the interviews in a safer virtual environment while maintaining the same matching quality and standards. AIM: This study was conducted to assess the medical training residency program applicants' satisfaction, stress, and other perspectives for the (SCFHS) March 2020 Matching-cycle conducted through an urgently implemented E-interviews process. METHOD: A cross-sectional, nationwide survey (Additional file 1) was sent to 4153 residency-nominated applicants to the (SCFHS) March 2020 cycle. RESULTS: Among the 510 candidates who responded, 62.2% applied for medical specialties, 20.2% applied for surgical specialties, and 17.6% applied for critical care and emergency specialties. Most respondents (61.2%) never had previous experience with web-based video conferences. Most respondents (80.2%) used the Zoom application to conduct the current E-interviews, whereas only 15.9% used the FaceTime application. 63.3% of the respondents preferred E-interviews over in-person interviews, and 60.6% rated their experience as very good or excellent. 75.7% of the respondents agreed that all their residency program queries were adequately addressed during the E-interviews. At the same time, 52.2% of them agreed that E-interviews allowed them to represent themselves accurately. 28.2% felt no stress at all with their E-interviews experience, while 41.2% felt little stressed and only 8.2% felt highly stressed. The factors that were independently and inversely associated with applicants' level of stress with E-interviews experience were their ability to represent themselves during the interviews (p = 0.001), cost-savings (p < 0.001), their overall rating of the E-interviews quality (p = 0.007) and the speed of the internet connection (p < 0.006). CONCLUSION: Videoconferencing was implemented on an urgent basis during the COVID-19 pandemic in the medical residency application process in Saudi Arabia. It was perceived as an adequate and promising tool to replace in-person interviews in the future. Applicants' satisfaction was mainly driven by good organization, cost-saving, and their ability to present themselves. Future studies to enhance this experience are warranted.

Knowledge, Attitude, and Perception of Health Care Personnel Working in Intensive Care Units of Mass Gatherings Toward the Application of Telemedicine Robotic Remote-Presence Technology: A Cross-Sectional Multicenter Study.

Background: The innovative telemedicine robotic remote-presence technology offers a promising solution to confront the challenges faced by health care personnel during events of mass gatherings by consulting expertise from offsite settings. Objective: To assess knowledge, attitude, and perceptions of health care personnel (physicians and nurses) toward telemedicine robotic remote-presence technology, at the intensive care units (ICUs) of hospitals serving mass gathering. Methods: The primary sampling unit included physicians and nurses using the sophisticated technology of telemedicine with robotic presence at the ICUs of Mina hospitals. An electronic invitation containing the survey tool was sent to all the participants from the four selected hospitals. Mean scores for knowledge and attitude questions were based on Likert scale responses. Result: The study received a final sample of 140 valid and complete responses. The findings showed overall positive attitude, but the knowledge was limited. On a maximum score of 5, the mean knowledge and attitude scores obtained were 2.55 and 3.51. The participants expressed strong agreement in using technology to seek expert opinion, increase communication among providers, and improve clinical decisions, which is an essential factor during mass gatherings. However, concerns about patient privacy and confidentiality were raised. Lack of training and insufficient knowledge regarding telemedicine and robotic systems' applications were identified as significant barriers, followed by issues related to equipment malfunction. Conclusions: Reinforcing continuous training programs to the health care staff to maximize the potential benefits of the innovative technology is suggested.

Understanding the stakeholders' preferences on a mobile application to reduce door to balloon time in the management of ST-elevated myocardial infarction patients - a qualitative study.

BACKGROUND: ST-elevated myocardial infarction (STEMI) is a critical and time-sensitive emergency. The survival depends on prompt initiation of treatment requiring high precision and multi-level coordination between healthcare staff. The use of a mobile application may facilitate prompt management and shorten the door-to-balloon time by capturing information at the point of care and provide immediate feedback to all healthcare staff involved in STEMI management. The objective of the present study has two primary components: (i) to explore the suggestions and opinions of stakeholders in the development of a novel mobile app for code activation in management of STEMI patients (ii) to find out the healthcare workers' expectations including facilitating steps and challenges in the activation process of the proposed mobile app. METHODS: Unstructured interviews were conducted with key informants (n = 2) to identify all stakeholders, who also helped in developing the interview protocol and prototype designs. In-depth, semi-structured, open-ended, face to face interviews were conducted on 22 stakeholders involved in managing STEMI patients. All interviews were recorded and transcribed verbatim. Data were analyzed using ATLAS.ti 8 software, allowing themes and subthemes to emerge. RESULTS: The 22 participants included in the study were cardiology physicians (n = 3), emergency consultants (n = 4), emergency room (ER) senior nurses (n = 10), and cardiac catheterization lab staff (n = 5). The main themes identified during analysis were workflow and the App. The themes identified from the interviews surrounding the App were: 1) facilitating ideas 2) management steps needed 3) features 4) preferred code activation method 5) steps of integration 6) possible benefits of the App 7) barriers and 8) possible solutions to the suggested barriers. Most of the interviewed stakeholders expressed their acceptance after viewing the proposed mobile app prototype. CONCLUSION: The study identified the mandatory features and the management steps needed from the stakeholder's perspectives. The steps for integrating the current paper-based workflow with the suggested mobile app were identified. The expected benefits of the App may include improved and faster management, accuracy, better communication, and improvement in data quality. Moreover, the possible barriers might comprise of doubtful acceptability, device-related issues, and time and data-related challenges.

Medical Doctors' Offline Computer-Assisted Digital Education: Systematic Review by the Digital Health Education Collaboration.

BACKGROUND: The widening gap between innovations in the medical field and the dissemination of such information to doctors may affect the quality of care. Offline computer-based digital education (OCDE) may be a potential solution to overcoming the geographical, financial, and temporal obstacles faced by doctors. OBJECTIVE: The objectives of this systematic review were to evaluate the effectiveness of OCDE compared with face-to-face learning, no intervention, or other types of digital learning for improving medical doctors' knowledge, cognitive skills, and patient-related outcomes. Secondary objectives were to assess the cost-effectiveness (CE) of OCDE and any adverse effects. METHODS: We searched major bibliographic databases from 1990 to August 2017 to identify relevant articles and followed the Cochrane methodology for systematic reviews of intervention. RESULTS: Overall, 27 randomized controlled trials (RCTs), 1 cluster RCT (cRCT), and 1 quasi-RCT were included in this review. The total number of participants was 1690 in addition to the cRCT, which included 24 practices. Due to the heterogeneity of the participants, interventions, and outcomes, meta-analysis was not feasible, and the results were presented as narrative summary. Compared with face-to-face learning, the effect of OCDE on knowledge gain is uncertain (ratio of the means [RM] range 0.95-1.17; 8 studies, 495 participants; very low grade of evidence). From the same comparison, the effect of OCDE on cognitive skill gain is uncertain (RM range 0.1-0.9; 8 studies, 375 participants; very low grade of evidence). OCDE may have little or no effect on patients' outcome compared with face-to-face education (2 studies, 62 participants; low grade of evidence). Compared with no intervention, OCDE may improve knowledge gain (RM range 1.36-0.98; 4 studies, 401 participants; low grade of evidence). From the same comparison, the effect of OCDE on cognitive skill gain is uncertain (RM range 1.1-1.15; 4 trials, 495 participants; very low grade of evidence). One cRCT, involving 24 practices, investigated patients' outcome in this comparison and showed no difference between the 2 groups with low-grade evidence. Compared with text-based learning, the effect of OCDE on cognitive skills gain is uncertain (RM range 0.91-1.46; 3 trials with 4 interventions; 68 participants; very low-grade evidence). No study in this comparison investigated knowledge gain or patients' outcomes. One study assessed the CE and showed that OCDE was cost-effective when compared with face-to-face learning if the cost is less than or equal to Can $200. No trial evaluated the adverse effect of OCDE. CONCLUSIONS: The effect of OCDE compared with other methods of education on medical doctors' knowledge and cognitive skill gain is uncertain. OCDE may improve doctors' knowledge compared with no intervention but its effect on doctors' cognitive skills is uncertain. OCDE may have little or no effect in improving patients' outcome.

The pattern of social media use and its association with academic performance among medical students.

BACKGROUND: There are concerns that the use of social media (SM) among medical students could affect academic performance. The objectives of the study were to investigate the pattern and reasons for SM use and their association with academic performance. METHODS: A stratified random sample, frequency distribution and comparison of categorical variables with Chi-square and Fisher exact tests were used. RESULTS: Of the 97% who responded, 98% used SM. The most popular were Whatsapp (87.8%), You tube (60.8%) and Twitter (51.8%) for general use; while You tube (83.5%), Whatsapp (35.5%) and Twitter (35.3%) for learning. For general use, there was a significant higher number of visits to You tube and Facebook among male students, while the reverse was true for Instagram and Path. Around 71% visited SM >4 times/day and 55% spent 1-4 hours/day. The main reasons for SM use were entertainment (95.8%), staying up-to-date with news (88.3%), and socializing (85.5%); for academic studies (40%). There was no significant association between Grade Point Average and the frequency of daily SM use or use during lectures. CONCLUSIONS: While almost all the students used SM, only a minority used them for academic purposes. SM use was not associated with academic performance.

Health professions' students have an alarming prevalence of depressive symptoms: exploration of the associated factors.

BACKGROUND: There is a need to better understand the depression phenomenon and to clarify why some students become depressed and others don't. The purpose of this study was to compare the prevalence of depressive symptoms among health professions' (HP) students, and to explore the association between socio-demographic factors (e.g. year of study, discipline, gender) and depressive symptoms. METHODS: In this descriptive-analytic, cross-sectional study, stratified proportionate sampling strategy was used to select the study sample during the academic year 2012-2013. The students from four health professions' schools situated within a large, public university located in Riyadh, Saudi Arabia were screened for depressive symptoms using the 21-item Beck Depression Inventory (BDI II). Chi-square test, student t-test and ANOVA were used to compare different categorical variables. RESULTS: The overall response rate was 79.0 %, the highest among dental students 86.1 %, and lowest among nursing (49.7 %). The overall prevalence rate of depressive symptoms was 47.0 %; it was highest among dentistry students (51.6 %), followed by medicine (46.2 %), applied medical sciences (AMS) (45.7 %) and lowest among nursing students (44.2 %). A statistically significant association was found between the presence and severity of depressive symptoms on one hand and the female gender (p = 0.000) and year of study on the other hand. CONCLUSION: This study seems to indicate an alarming rate of depressive symptoms. Female gender, dentistry, the third year for all schools and fifth year for medicine and dentistry have the highest association with depressive symptoms. Future studies may be needed to explore further the reasons and explanations for the variation in the prevalence of depressive symptoms among these groups. The factors that deserve exploration include curricular variables and personal factors such as the students' study skills.

Association of Online Health Information-Seeking Behavior and Self-Care Activities Among Type 2 Diabetic Patients in Saudi Arabia.

BACKGROUND: Health information obtained from the Internet has an impact on patient health care outcomes. There is a growing concern over the quality of online health information sources used by diabetic patients because little is known about their health information-seeking behavior and the impact this behavior has on their diabetes-related self-care, in particular in the Middle East setting. OBJECTIVE: The aim of this study was to determine the online health-related information-seeking behavior among adult type 2 diabetic patients in the Middle East and the impact of their online health-related information-seeking behavior on their self-care activities. METHODS: A cross-sectional survey was conducted on 344 patients with type 2 diabetes attending inpatient and outpatient primary health care clinics at 2 teaching hospitals in Riyadh, Saudi Arabia. The main outcome measures included the ability of patients to access the Internet, their ability to use the Internet to search for health-related information, and their responses to Internet searches in relation to their self-care activities. Further analysis of differences based on age, gender, sociodemographic, and diabetes-related self-care activities among online health-related information seekers and nononline health-related information seekers was conducted. RESULTS: Among the 344 patients, 74.1% (255/344) were male with a mean age of 53.5 (SD 13.8) years. Only 39.0% (134/344) were Internet users; 71.6% (96/134) of them used the Internet for seeking health-related information. Most participants reported that their primary source of health-related information was their physician (216/344, 62.8%) followed by television (155/344, 45.1%), family (113/344, 32.8%), newspapers (100/344, 29.1%), and the Internet (96/344, 27.9%). Primary topics participants searched for were therapeutic diet for diabetes (55/96, 57%) and symptoms of diabetes (52/96, 54%) followed by diabetes treatment (50/96, 52%). Long history of diabetes, familial history of the disease, unemployment, and not seeking diabetes education were the most common barriers for online health-related information-seeking behavior. Younger age, female, marital status, higher education, higher income, and longer duration of Internet usage were associated with more online health-related information-seeking behaviors. Most (89/96, 93%) online health-related information seekers reported positive change in their behaviors after seeking online health information. Overall odds ratio (OR 1.56, 95% CI 0.63-3.28) for all self-care responses demonstrated that there was no statistically significant difference between those seeking health-related information online and non-health-related information seekers. However, health-related information seekers were better in testing their blood glucose regularly, taking proper action for hyperglycemia, and adopting nonpharmacological management. CONCLUSIONS: Physicians and television are still the primary sources of health-related information for adult diabetic patients in Saudi Arabia whether they seek health-related information online or not. This study demonstrates that participants seeking online health-related information are more conscious about their diabetes self-care compared to non-health-related information seekers in some aspects more than the others.

Correlates of cardiovascular disease risk scores in women in Riyadh, Kingdom of Saudi Arabia.

Saudi Arabia has a high prevalence of obesity and physical inactivity. We measured cardiovascular (CVD) risk scores and determined the factors associated with them in women in Riyadh, Saudi Arabia. A cross-sectional study using a self-administered questionnaire was conducted on 291 women aged ≥ 30 years. Information was collected on socio-demographics and physical health status. Anthropometric and blood pressure measurements were taken. Physical activity was measured using Kaiser's Physical Activity Survey and Godin's Leisure Time Exercise questionnaire. CVD risk scores were calculated using the non-laboratory-based Framingham Risk (FRS) prediction model for primary care. FRS scores ranged from 0.50 to 21.9. A total of 2.7% (n = 8) of women had a high FRS score (>20), 5.5% (n = 16) had intermediate scores (11-20), and 91.8% (n = 267) of women had low scores (<10) CVD risk scores. Multiple linear regression results indicated that a one-unit change in physical activity (household/caregiver index), strenuous exercise, waist circumference, number of children, television watching, and knee pain were significantly associated with -0.20 (p < .01), -0.12 (p = .03), 0.19 (p = .001), 0.29 (p < .01), 0.13 (p = .04), and 0.11 (p = .05) unit change in CVD risk scores, respectively. Household activities and strenuous exercise had a protective role in females in relation to CVD risk. Programs recommending physical activity at all levels should be encouraged.

OpenAI o1-Preview vs. ChatGPT in Healthcare: A New Frontier in Medical AI Reasoning.

This editorial explores the recent advancements in generative artificial intelligence with the newly-released OpenAI o1-Preview, comparing its capabilities to the traditional ChatGPT (GPT-4) model, particularly in the context of healthcare. While ChatGPT has shown many applications for general medical advice and patient interactions, OpenAI o1-Preview introduces new features with advanced reasoning skills using a chain of thought processes that could enable users to tackle more complex medical queries such as genetic disease discovery, multi-system or complex disease care, and medical research support. The article explores some of the new model's potential and other aspects that may affect its usage, like slower response times due to its extensive reasoning approach yet highlights its potential for reducing hallucinations and offering more accurate outputs for complex medical problems. Ethical challenges, data diversity, access equity, and transparency are also discussed, identifying key areas for future research, including optimizing the use of both models in tandem for healthcare applications. The editorial concludes by advocating for collaborative exploration of all large language models (LLMs), including the novel OpenAI o1-Preview, to fully utilize their transformative potential in medicine and healthcare delivery. This model, with its advanced reasoning capabilities, presents an opportunity to empower healthcare professionals, policymakers, and computer scientists to work together in transforming patient care, accelerating medical research, and enhancing healthcare outcomes. By optimizing the use of several LLM models in tandem, healthcare systems may enhance efficiency and precision, as well as mitigate previous LLM challenges, such as ethical concerns, access disparities, and technical limitations, steering to a new era of artificial intelligence (AI)-driven healthcare.

Transforming Virtual Healthcare: The Potentials of ChatGPT-4omni in Telemedicine.

The introduction of OpenAI's ChatGPT-4omni (GPT-4o) represents a potential advancement in virtual healthcare and telemedicine. GPT-4o excels in processing audio, visual, and textual data in real time, offering possible enhancements in understanding natural language in both English and non-English contexts. Furthermore, the new "Temporary Chat" feature may improve privacy and data confidentiality during interactions, potentially increasing integration with healthcare systems. These innovations promise to enhance communication clarity, facilitate the integration of medical images, and increase data privacy in online consultations. This editorial explores some future implications of these advancements for telemedicine, highlighting the necessity for further research on reliability and the integration of advanced language models with human expertise.

Exploring behavioral intention to use telemedicine services post COVID-19: a cross sectional study in Saudi Arabia.

Introduction: While telemedicine offers significant benefits, there remain substantial knowledge gaps in the literature, particularly regarding its use in Saudi Arabia. This study aims to explore health consumers' behavioral intention to use telemedicine examining the associated factors such as eHealth literacy and attitudes toward telemedicine services. Methods: A cross-sectional observational study was conducted to collect data on demographics, health status, internet skills, attitudes toward telemedicine, and eHealth literacy. An online survey was administered at two large public gatherings in Riyadh. The eHEALS-Pl scale was used to measure perceived eHealth literacy levels, and data analysis was performed using SPSS (IBM Corp. United States). Results: There were 385 participants, with an equal distribution of genders. The largest age group was 18-20 years old (57%). Nearly half of the participants were neither employed nor students, while 43% had access to governmental hospitals through employment. 71% reported proficiency in using the internet. Health-wise, 47% rated their health as excellent, and 56% did not have medical insurance. 87% expressed a high likelihood of using telemedicine if offered by a provider. Participants were categorized based on their eHealth Literacy scores, with 54% scoring low and 46% scoring high. Overall, participants showed positive attitudes toward telemedicine, with 82% agreeing that it saves time, money, and provides access to specialized care. About half of the participants perceived the process of seeing a doctor through telemedicine video as complex. Both eHealth Literacy and attitudes toward telemedicine showed a statistically significant association with the intention to use telemedicine (p < 0.001). There was a positive and significant correlation between eHealth Literacy and attitudes (ρ =0.460; p < 0.001). Multivariate ordinal regression analysis revealed that the odds for a high likelihood of intention to use telemedicine significantly increased with positive attitudes (p < 0.001). Mediation analysis confirmed the significant mediating role of attitudes toward telemedicine in the relationship between eHealth Literacy and the intention to use telemedicine. Conclusion: The findings underline the importance of enhancing health literacy and consumer attitudes toward telemedicine, particularly during the healthcare digital transformation we are experiencing globally. This is crucial for promoting increased acceptance and utilization of telemedicine services beyond the COVID-19 pandemic.

Reference Hallucination Score for Medical Artificial Intelligence Chatbots: Development and Usability Study.

BACKGROUND: Artificial intelligence (AI) chatbots have recently gained use in medical practice by health care practitioners. Interestingly, the output of these AI chatbots was found to have varying degrees of hallucination in content and references. Such hallucinations generate doubts about their output and their implementation. OBJECTIVE: The aim of our study was to propose a reference hallucination score (RHS) to evaluate the authenticity of AI chatbots' citations. METHODS: Six AI chatbots were challenged with the same 10 medical prompts, requesting 10 references per prompt. The RHS is composed of 6 bibliographic items and the reference's relevance to prompts' keywords. RHS was calculated for each reference, prompt, and type of prompt (basic vs complex). The average RHS was calculated for each AI chatbot and compared across the different types of prompts and AI chatbots. RESULTS: Bard failed to generate any references. ChatGPT 3.5 and Bing generated the highest RHS (score=11), while Elicit and SciSpace generated the lowest RHS (score=1), and Perplexity generated a middle RHS (score=7). The highest degree of hallucination was observed for reference relevancy to the prompt keywords (308/500, 61.6%), while the lowest was for reference titles (169/500, 33.8%). ChatGPT and Bing had comparable RHS (ß coefficient=-0.069; P=.32), while Perplexity had significantly lower RHS than ChatGPT (ß coefficient=-0.345; P<.001). AI chatbots generally had significantly higher RHS when prompted with scenarios or complex format prompts (ß coefficient=0.486; P<.001). CONCLUSIONS: The variation in RHS underscores the necessity for a robust reference evaluation tool to improve the authenticity of AI chatbots. Further, the variations highlight the importance of verifying their output and citations. Elicit and SciSpace had negligible hallucination, while ChatGPT and Bing had critical hallucination levels. The proposed AI chatbots' RHS could contribute to ongoing efforts to enhance AI's general reliability in medical research.

ChatGPT-3.5 System Usability Scale early assessment among Healthcare Workers: Horizons of adoption in medical practice.

Artificial intelligence (AI) chatbots, such as ChatGPT, have widely invaded all domains of human life. They have the potential to transform healthcare future. However, their effective implementation hinges on healthcare workers' (HCWs) adoption and perceptions. This study aimed to evaluate HCWs usability of ChatGPT three months post-launch in Saudi Arabia using the System Usability Scale (SUS). A total of 194 HCWs participated in the survey. Forty-seven percent were satisfied with their usage, 57 % expressed moderate to high trust in its ability to generate medical decisions. 58 % expected ChatGPT would improve patients' outcomes, even though 84 % were optimistic of its potential to improve the future of healthcare practice. They expressed possible concerns like recommending harmful medical decisions and medicolegal implications. The overall mean SUS score was 64.52, equivalent to 50 % percentile rank, indicating high marginal acceptability of the system. The strongest positive predictors of high SUS scores were participants' belief in AI chatbot's benefits in medical research, self-rated familiarity with ChatGPT and self-rated computer skills proficiency. Participants' learnability and ease of use score correlated positively but weakly. On the other hand, medical students and interns had significantly high learnability scores compared to others, while ease of use scores correlated very strongly with participants' perception of positive impact of ChatGPT on the future of healthcare practice. Our findings highlight the HCWs' perceived marginal acceptance of ChatGPT at the current stage and their optimism of its potential in supporting them in future practice, especially in the research domain, in addition to humble ambition of its potential to improve patients' outcomes particularly in regard of medical decisions. On the other end, it underscores the need for ongoing efforts to build trust and address ethical and legal concerns of AI implications in healthcare. The study contributes to the growing body of literature on AI chatbots in healthcare, especially addressing its future improvement strategies and provides insights for policymakers and healthcare providers about the potential benefits and challenges of implementing them in their practice.

Integrating ChatGPT in Medical Education: Adapting Curricula to Cultivate Competent Physicians for the AI Era.

The rapid advancements in artificial intelligence (AI) language models, particularly ChatGPT (OpenAI, San Francisco, California, United States), necessitate the adaptation of medical education curricula to cultivate competent physicians in the AI era. In this editorial, we discuss short-term solutions and long-term adaptations for integrating ChatGPT into medical education. We recommend promoting digital literacy, developing critical thinking skills, and emphasizing evidence-based relevance as quick fixes. Long-term adaptations include focusing on the human factor, interprofessional collaboration, continuous professional development, and research and evaluation. By implementing these changes, medical educators can optimize medical education for the AI era, ensuring students are well prepared for a technologically advanced future in healthcare.

ChatGPT Surpasses 1000 Publications on PubMed: Envisioning the Road Ahead.

The exponential growth of ChatGPT in medical literature, amassing over 1000 PubMed citations by August 2023, underscores a pivotal juncture in the convergence of artificial intelligence (AI) and healthcare. This remarkable rise not only showcases its potential to revolutionize medical academia but also indicates its impending influence on patient care and healthcare systems. Notwithstanding this enthusiasm, one-third of these citations are editorials or commentaries, stressing a gap in empirical research. Alongside its potential, there are concerns about ChatGPT becoming a "Weapon of Mass Deception" and the need for rigorous evaluations to counter inaccuracies. The World Association of Medical Editors has released guidelines emphasizing that AI tools should not be manuscript co-authors and advocates for clear disclosures in AI-assisted academic works. Interestingly, ChatGPT achieved its citation milestone within nine months, compared to Google's 14 years. As Large Language Models (LLMs), like ChatGPT, become more integral in healthcare, issues surrounding data protection, patient privacy, and ethical implications gain prominence. As the future of LLM research unfolds, key areas of interest include its efficacy in clinical settings, its role in telemedicine, and its potential in medical education. The journey ahead necessitates a harmonious partnership between the medical community and AI developers, emphasizing both technological advancements and ethical considerations.

Healthcare's New Horizon With ChatGPT's Voice and Vision Capabilities: A Leap Beyond Text.

The integration of artificial intelligence (AI) in healthcare is responsible for a paradigm shift in medicine. OpenAI's recent augmentation of their Generative Pre-trained Transformer (ChatGPT) large language model (LLM) with voice and image recognition capabilities (OpenAI, Delaware) presents another potential transformative tool for healthcare. Envision a healthcare setting where professionals engage in dynamic interactions with ChatGPT to navigate the complexities of atypical medical scenarios. In this innovative landscape, practitioners could solicit ChatGPT's expertise for concise summarizations and insightful extrapolations from a myriad of web-based resources pertaining to similar medical conditions. Furthermore, imagine patients using ChatGPT to identify abnormalities in medical images or skin lesions. While the prospects are diverse, challenges such as suboptimal audio quality and ensuring data security necessitate cautious integration in medical practice. Drawing insights from previous ChatGPT iterations could provide a prudent roadmap for navigating possible challenges. This editorial explores some possible horizons and potential hurdles of ChatGPT's enhanced functionalities in healthcare, emphasizing the importance of continued refinements and vigilance to maximize the benefits while minimizing risks. Through collaborative efforts between AI developers and healthcare professionals, another fusion of AI and healthcare can evolve into enriched patient care and enhanced medical experience.