BACKGROUND: Artificial intelligence, particularly chatbot systems, is becoming an instrumental tool in health care, aiding clinical decision-making and patient engagement.
OBJECTIVE: This study aims to analyze the performance of ChatGPT-3.5 and ChatGPT-4 in addressing complex clinical and ethical dilemmas, and to illustrate their potential role in health care decision-making while comparing seniors\' and residents\' ratings, and specific question types.
METHODS: A total of 4 specialized physicians formulated 176 real-world clinical questions. A total of 8 senior physicians and residents assessed responses from GPT-3.5 and GPT-4 on a 1-5 scale across 5 categories: accuracy, relevance, clarity, utility, and comprehensiveness. Evaluations were conducted within internal medicine, emergency medicine, and ethics. Comparisons were made globally, between seniors and residents, and across classifications.
RESULTS: Both GPT models received high mean scores (4.4, SD 0.8 for GPT-4 and 4.1, SD 1.0 for GPT-3.5). GPT-4 outperformed GPT-3.5 across all rating dimensions, with seniors consistently rating responses higher than residents for both models. Specifically, seniors rated GPT-4 as more beneficial and complete (mean 4.6 vs 4.0 and 4.6 vs 4.1, respectively; P<.001), and GPT-3.5 similarly (mean 4.1 vs 3.7 and 3.9 vs 3.5, respectively; P<.001). Ethical queries received the highest ratings for both models, with mean scores reflecting consistency across accuracy and completeness criteria. Distinctions among question types were significant, particularly for the GPT-4 mean scores in completeness across emergency, internal, and ethical questions (4.2, SD 1.0; 4.3, SD 0.8; and 4.5, SD 0.7, respectively; P<.001), and for GPT-3.5\'s accuracy, beneficial, and completeness dimensions.
CONCLUSIONS: ChatGPT\'s potential to assist physicians with medical issues is promising, with prospects to enhance diagnostics, treatments, and ethics. While integration into clinical workflows may be valuable, it must complement, not replace, human expertise. Continued research is essential to ensure safe and effective implementation in clinical environments.

BACKGROUND: Increased workload, including workload related to electronic health record (EHR) documentation, is reported as a main contributor to nurse burnout and adversely affects patient safety and nurse satisfaction. Traditional methods for workload analysis are either administrative measures (such as the nurse-patient ratio) that do not represent actual nursing care or are subjective and limited to snapshots of care (eg, time-motion studies). Observing care and testing workflow changes in real time can be obstructive to clinical care. An examination of EHR interactions using EHR audit logs could provide a scalable, unobtrusive way to quantify the nursing workload, at least to the extent that nursing work is represented in EHR documentation. EHR audit logs are extremely complex; however, simple analytical methods cannot discover complex temporal patterns, requiring use of state-of-the-art temporal data-mining approaches. To effectively use these approaches, it is necessary to structure the raw audit logs into a consistent and scalable logical data model that can be consumed by machine learning (ML) algorithms.
OBJECTIVE: We aimed to conceptualize a logical data model for nurse-EHR interactions that would support the future development of temporal ML models based on EHR audit log data.
METHODS: We conducted a preliminary review of EHR audit logs to understand the types of nursing-specific data captured. Using concepts derived from the literature and our previous experience studying temporal patterns in biomedical data, we formulated a logical data model that can describe nurse-EHR interactions, the nurse-intrinsic and situational characteristics that may influence those interactions, and outcomes of relevance to the nursing workload in a scalable and extensible manner.
RESULTS: We describe the data structure and concepts from EHR audit log data associated with nursing workload as a logical data model named RNteract. We conceptually demonstrate how using this logical data model could support temporal unsupervised ML and state-of-the-art artificial intelligence (AI) methods for predictive modeling.
CONCLUSIONS: The RNteract logical data model appears capable of supporting a variety of AI-based systems and should be generalizable to any type of EHR system or health care setting. Quantitatively identifying and analyzing temporal patterns of nurse-EHR interactions is foundational for developing interventions that support the nursing documentation workload and address nurse burnout.

BACKGROUND: HALE is now a regular strategic planning indicator for all levels of the Chinese government. However, HALE measurements necessitate comprehensive data collection and intricate technology. Therefore, effectively converting numerous diseases into the years lived with disability (YLD) rate is a significant challenge for HALE measurements. Our study aimed to construct a simple YLD rate measurement model with high applicability based on the current situation of actual data resources within China to address challenges in measuring HALE target values during planning.
METHODS: First, based on the Chinese YLD rate in the Global Burden of Disease (GBD) 2019, Pearson correlation analysis, the global optimum method, etc., was utilized to screen the best predictor variables from the current Chinese data resources. Missing data for predictor variables were filled in via spline interpolation. Then, multiple linear regression models were fitted to construct the YLD rate measurement model. The Sullivan method was used to measure HALE. The Monte Carlo method was employed to generate 95% uncertainty intervals. Finally, model performances were assessed using the mean absolute error (MAE) and mean absolute percentage error (MAPE).
RESULTS: A three-input-parameter model was constructed to measure the age-specific YLD rates by sex in China, directly using the incidence of infectious diseases, the incidence of chronic diseases among persons aged 15 and older, and the addition of an under-five mortality rate covariate. The total MAE and MAPE for the combined YLD rate were 0.0007 and 0.5949%, respectively. The MAE and MAPE of the combined HALE in the 0-year-old group were 0.0341 and 0.0526%, respectively. There were slightly fewer males (0.0197, 0.0311%) than females (0.0501, 0.0755%).
CONCLUSIONS: We constructed a high-accuracy model to measure the YLD rate in China by using three monitoring indicators from the Chinese national routine as predictor variables. The model provides a realistic and feasible solution for measuring HALE at the national and especially regional levels, considering limited data.

Artificial intelligence (AI) is reshaping health care, including nursing, across Asia, presenting opportunities to improve patient care and outcomes. This viewpoint presents our perspective and interpretation of the current AI landscape, acknowledging its evolution driven by enhanced processing capabilities, extensive data sets, and refined algorithms. Notable applications in countries such as Singapore, South Korea, Japan, and China showcase the integration of AI-powered technologies such as chatbots, virtual assistants, data mining, and automated risk assessment systems. This paper further explores the transformative impact of AI on nursing education, emphasizing personalized learning, adaptive approaches, and AI-enriched simulation tools, and discusses the opportunities and challenges of these developments. We argue for the harmonious coexistence of traditional nursing values with AI innovations, marking a significant stride toward a promising health care future in Asia.

Recent breakthroughs in artificial intelligence (AI) language models have elevated the vision of using conversational AI support for mental health, with a growing body of literature indicating varying degrees of efficacy. In this paper, we ask when, in therapy, it will be easier to replace humans and, conversely, in what instances, human connection will still be more valued. We suggest that empathy lies at the heart of the answer to this question. First, we define different aspects of empathy and outline the potential empathic capabilities of humans versus AI. Next, we consider what determines when these aspects are needed most in therapy, both from the perspective of therapeutic methodology and from the perspective of patient objectives. Ultimately, our goal is to prompt further investigation and dialogue, urging both practitioners and scholars engaged in AI-mediated therapy to keep these questions and considerations in mind when investigating AI implementation in mental health.

BACKGROUND: The growth in the capabilities of telehealth have made it possible to identify individuals with a higher risk of uncontrolled diabetes and provide them with targeted support and resources to help them manage their condition. Thus, predictive modeling has emerged as a valuable tool for the advancement of diabetes management.
OBJECTIVE: This study aimed to conceptualize and develop a novel machine learning (ML) approach to proactively identify participants enrolled in a remote diabetes monitoring program (RDMP) who were at risk of uncontrolled diabetes at 12 months in the program.
METHODS: Registry data from the Livongo for Diabetes RDMP were used to design separate dynamic predictive ML models to predict participant outcomes at each monthly checkpoint of the participants\' program journey (month-n models) from the first day of onboarding (month-0 model) up to the 11th month (month-11 model). A participant\'s program journey began upon onboarding into the RDMP and monitoring their own blood glucose (BG) levels through the RDMP-provided BG meter. Each participant passed through 12 predicative models through their first year enrolled in the RDMP. Four categories of participant attributes (ie, survey data, BG data, medication fills, and health signals) were used for feature construction. The models were trained using the light gradient boosting machine and underwent hyperparameter tuning. The performance of the models was evaluated using standard metrics, including precision, recall, specificity, the area under the curve, the F1-score, and accuracy.
RESULTS: The ML models exhibited strong performance, accurately identifying observable at-risk participants, with recall ranging from 70% to 94% and precision from 40% to 88% across the 12-month program journey. Unobservable at-risk participants also showed promising performance, with recall ranging from 61% to 82% and precision from 42% to 61%. Overall, model performance improved as participants progressed through their program journey, demonstrating the importance of engagement data in predicting long-term clinical outcomes.
CONCLUSIONS: This study explored the Livongo for Diabetes RDMP participants\' temporal and static attributes, identification of diabetes management patterns and characteristics, and their relationship to predict diabetes management outcomes. Proactive targeting ML models accurately identified participants at risk of uncontrolled diabetes with a high level of precision that was generalizable through future years within the RDMP. The ability to identify participants who are at risk at various time points throughout the program journey allows for personalized interventions to improve outcomes. This approach offers significant advancements in the feasibility of large-scale implementation in remote monitoring programs and can help prevent uncontrolled glycemic levels and diabetes-related complications. Future research should include the impact of significant changes that can affect a participant\'s diabetes management.

BACKGROUND: Large language models (LLMs) hold potential for mental health applications. However, their opaque alignment processes may embed biases that shape problematic perspectives. Evaluating the values embedded within LLMs that guide their decision-making have ethical importance. Schwartz\'s theory of basic values (STBV) provides a framework for quantifying cultural value orientations and has shown utility for examining values in mental health contexts, including cultural, diagnostic, and therapist-client dynamics.
OBJECTIVE: This study aimed to (1) evaluate whether the STBV can measure value-like constructs within leading LLMs and (2) determine whether LLMs exhibit distinct value-like patterns from humans and each other.
METHODS: In total, 4 LLMs (Bard, Claude 2, Generative Pretrained Transformer [GPT]-3.5, GPT-4) were anthropomorphized and instructed to complete the Portrait Values Questionnaire-Revised (PVQ-RR) to assess value-like constructs. Their responses over 10 trials were analyzed for reliability and validity. To benchmark the LLMs\' value profiles, their results were compared to published data from a diverse sample of 53,472 individuals across 49 nations who had completed the PVQ-RR. This allowed us to assess whether the LLMs diverged from established human value patterns across cultural groups. Value profiles were also compared between models via statistical tests.
RESULTS: The PVQ-RR showed good reliability and validity for quantifying value-like infrastructure within the LLMs. However, substantial divergence emerged between the LLMs\' value profiles and population data. The models lacked consensus and exhibited distinct motivational biases, reflecting opaque alignment processes. For example, all models prioritized universalism and self-direction, while de-emphasizing achievement, power, and security relative to humans. Successful discriminant analysis differentiated the 4 LLMs\' distinct value profiles. Further examination found the biased value profiles strongly predicted the LLMs\' responses when presented with mental health dilemmas requiring choosing between opposing values. This provided further validation for the models embedding distinct motivational value-like constructs that shape their decision-making.
CONCLUSIONS: This study leveraged the STBV to map the motivational value-like infrastructure underpinning leading LLMs. Although the study demonstrated the STBV can effectively characterize value-like infrastructure within LLMs, substantial divergence from human values raises ethical concerns about aligning these models with mental health applications. The biases toward certain cultural value sets pose risks if integrated without proper safeguards. For example, prioritizing universalism could promote unconditional acceptance even when clinically unwise. Furthermore, the differences between the LLMs underscore the need to standardize alignment processes to capture true cultural diversity. Thus, any responsible integration of LLMs into mental health care must account for their embedded biases and motivation mismatches to ensure equitable delivery across diverse populations. Achieving this will require transparency and refinement of alignment techniques to instill comprehensive human values.

BACKGROUND: Mentalization, which is integral to human cognitive processes, pertains to the interpretation of one\'s own and others\' mental states, including emotions, beliefs, and intentions. With the advent of artificial intelligence (AI) and the prominence of large language models in mental health applications, questions persist about their aptitude in emotional comprehension. The prior iteration of the large language model from OpenAI, ChatGPT-3.5, demonstrated an advanced capacity to interpret emotions from textual data, surpassing human benchmarks. Given the introduction of ChatGPT-4, with its enhanced visual processing capabilities, and considering Google Bard\'s existing visual functionalities, a rigorous assessment of their proficiency in visual mentalizing is warranted.
OBJECTIVE: The aim of the research was to critically evaluate the capabilities of ChatGPT-4 and Google Bard with regard to their competence in discerning visual mentalizing indicators as contrasted with their textual-based mentalizing abilities.
METHODS: The Reading the Mind in the Eyes Test developed by Baron-Cohen and colleagues was used to assess the models\' proficiency in interpreting visual emotional indicators. Simultaneously, the Levels of Emotional Awareness Scale was used to evaluate the large language models\' aptitude in textual mentalizing. Collating data from both tests provided a holistic view of the mentalizing capabilities of ChatGPT-4 and Bard.
RESULTS: ChatGPT-4, displaying a pronounced ability in emotion recognition, secured scores of 26 and 27 in 2 distinct evaluations, significantly deviating from a random response paradigm (P<.001). These scores align with established benchmarks from the broader human demographic. Notably, ChatGPT-4 exhibited consistent responses, with no discernible biases pertaining to the sex of the model or the nature of the emotion. In contrast, Google Bard\'s performance aligned with random response patterns, securing scores of 10 and 12 and rendering further detailed analysis redundant. In the domain of textual analysis, both ChatGPT and Bard surpassed established benchmarks from the general population, with their performances being remarkably congruent.
CONCLUSIONS: ChatGPT-4 proved its efficacy in the domain of visual mentalizing, aligning closely with human performance standards. Although both models displayed commendable acumen in textual emotion interpretation, Bard\'s capabilities in visual emotion interpretation necessitate further scrutiny and potential refinement. This study stresses the criticality of ethical AI development for emotional recognition, highlighting the need for inclusive data, collaboration with patients and mental health experts, and stringent governmental oversight to ensure transparency and protect patient privacy.

Objective: This study aimed to identify reliable risk factors for residual/recurrent cervical intraepithelial lesions in patients with negative margins after cold-knife conization. Methods: A total of 2352 women with HSILs (high-grade squamous intraepithelial lesions) with negative margins who underwent cold-knife conization between January 2014 and December 2020 were included; in total, 1411 women were assigned to the development cohort, and 941 women were assigned to the validation cohort. Multivariate logistic regression was used to build four predictive models based on the different combinations of follow-up data (Model A: preoperative factors; Model B: first-follow-up data; Model C: second-follow-up data; Model D: data from both follow-ups). The accuracy, sensitivity, specificity, false-positive rate (FPR), false-negative rate (FNR), and area under the receiver operating characteristic curve (AUC) were evaluated on the validation cohort. The predictive power of risk factors was further validated using six machine learning algorithms. Results: Model D demonstrated the highest AUC of 0.91 (95% CI, 0.87 to 0.96) in the validation cohort, whereas Models A, B, and C achieved AUCs of 0.69 (95% CI, 0.59 to 0.78), 0.88 (95% CI, 0.80 to 0.95), and 0.89 (95% CI, 0.81 to 0.97) respectively. The six machine learning methods achieved consistent results. Kaplan-Meier (KM) survival curves demonstrated that our models could effectively stratify patients with all models (p < 0.05 for all models). Conclusion: Our model, which is based on preoperative and follow-up factors, can serve as a complementary screening procedure for the early detection or prediction of recurrence after cold-knife conization in HSIL patients.

BACKGROUND: For patients with portal hypertension (PH), portal vein thrombosis (PVT) is a fatal complication after splenectomy. Postoperative platelet elevation is considered the foremost reason for PVT. However, the value of postoperative platelet elevation rate (PPER) in predicting PVT has never been studied.
OBJECTIVE: To investigate the predictive value of PPER for PVT and establish PPER-based prediction models to early identify individuals at high risk of PVT after splenectomy.
METHODS: We retrospectively reviewed 483 patients with PH related to hepatitis B virus who underwent splenectomy between July 2011 and September 2018, and they were randomized into either a training (n = 338) or a validation (n = 145) cohort. The generalized linear (GL) method, least absolute shrinkage and selection operator (LASSO), and random forest (RF) were used to construct models. The receiver operating characteristic curves (ROC), calibration curve, decision curve analysis (DCA), and clinical impact curve (CIC) were used to evaluate the robustness and clinical practicability of the GL model (GLM), LASSO model (LSM), and RF model (RFM).
RESULTS: Multivariate analysis exhibited that the first and third days for PPER (PPER1, PPER3) were strongly associated with PVT [odds ratio (OR): 1.78, 95% confidence interval (CI): 1.24-2.62, P = 0.002; OR: 1.43, 95%CI: 1.16-1.77, P < 0.001, respectively]. The areas under the ROC curves of the GLM, LSM, and RFM in the training cohort were 0.83 (95%CI: 0.79-0.88), 0.84 (95%CI: 0.79-0.88), and 0.84 (95%CI: 0.79-0.88), respectively; and were 0.77 (95%CI: 0.69-0.85), 0.83 (95%CI: 0.76-0.90), and 0.78 (95%CI: 0.70-0.85) in the validation cohort, respectively. The calibration curves showed satisfactory agreement between prediction by models and actual observation. DCA and CIC indicated that all models conferred high clinical net benefits.
CONCLUSIONS: PPER1 and PPER3 are effective indicators for postoperative prediction of PVT. We have successfully developed PPER-based practical models to accurately predict PVT, which would conveniently help clinicians rapidly differentiate individuals at high risk of PVT, and thus guide the adoption of timely interventions.