Healthcare Informatics Research最新文献

Objectives: Online consultation scheduling is increasingly common in health services across various countries. This paper reviews articles published in the past five years and reflects on the risks and benefits of this practice, linking it to a recent Portuguese pilot project.

Methods: A search for articles from Web of Science and Scopus published since 2018 was conducted using the terms "online scheduling," "online booking," and "consultations." This search was completed in the last week of 2023.

Results: Out of 64 articles retrieved, 26 were relevant to the topic. These articles were reviewed, and their main findings, along with those from other relevant sources, were discussed.

Conclusions: Several limitations of online consultations were identified, encompassing ethical, clinical, and economic aspects. While these consultations tend to be less expensive, their accessibility varies based on factors such as the users' age, whether they reside in rural or urban areas, and the technological capabilities of different countries, indicating that access disparities may continue to widen. Confidentiality concerns also arise, varying by medical specialty, along with issues related to payment. Overall, however, both users and health professionals view the advent of online consultation booking positively. In conclusion, despite the risks identified, online consultation booking has the potential to enhance user access to health services, provided that usage limitations and technological disparities are addressed. Research production has not kept pace with rapid technological advancements.

Objectives: Sepsis is a leading global cause of mortality, and predicting its outcomes is vital for improving patient care. This study explored the capabilities of ChatGPT, a state-of-the-art natural language processing model, in predicting in-hospital mortality for sepsis patients.

Methods: This study utilized data from the Korean Sepsis Alliance (KSA) database, collected between 2019 and 2021, focusing on adult intensive care unit (ICU) patients and aiming to determine whether ChatGPT could predict all-cause mortality after ICU admission at 7 and 30 days. Structured prompts enabled ChatGPT to engage in in-context learning, with the number of patient examples varying from zero to six. The predictive capabilities of ChatGPT-3.5-turbo and ChatGPT-4 were then compared against a gradient boosting model (GBM) using various performance metrics.

Results: From the KSA database, 4,786 patients formed the 7-day mortality prediction dataset, of whom 718 died, and 4,025 patients formed the 30-day dataset, with 1,368 deaths. Age and clinical markers (e.g., Sequential Organ Failure Assessment score and lactic acid levels) showed significant differences between survivors and non-survivors in both datasets. For 7-day mortality predictions, the area under the receiver operating characteristic curve (AUROC) was 0.70-0.83 for GPT-4, 0.51-0.70 for GPT-3.5, and 0.79 for GBM. The AUROC for 30-day mortality was 0.51-0.59 for GPT-4, 0.47-0.57 for GPT-3.5, and 0.76 for GBM. Zero-shot predictions using GPT-4 for mortality from ICU admission to day 30 showed AUROCs from the mid-0.60s to 0.75 for GPT-4 and mainly from 0.47 to 0.63 for GPT-3.5.

Conclusions: GPT-4 demonstrated potential in predicting short-term in-hospital mortality, although its performance varied across different evaluation metrics.

Objectives: The use of technology has rapidly increased in the past century. Artificial intelligence (AI) and information technology (IT) are now applied in healthcare and medical education. The purpose of this study was to assess the readiness of Indonesian teaching staff and pediatric residents for AI integration into the curriculum.

Methods: An anonymous online survey was distributed among teaching staff and pediatric residents from 15 national universities. The questionnaire consisted of two sections: demographic information and questions regarding the use of IT and AI in child health education. Responses were collected using a 5-point Likert scale: strongly disagree, disagree, neutral, agree, and highly agree.

Results: A total of 728 pediatric residents and 196 teaching staff from 15 national universities participated in the survey. Over half of the respondents were familiar with the terms IT and AI. The majority agreed that IT and AI have simplified the process of learning theories and skills. All participants were in favor of sharing data to facilitate the development of AI and expressed readiness to incorporate IT and AI into their teaching tools.

Conclusions: The findings of our study indicate that pediatric residents and teaching staff are ready to implement AI in medical education.

Objectives: This study aimed to optimize early coronary heart disease (CHD) prediction using a genetic algorithm (GA)-based convolutional neural network (CNN) feature engineering approach. We sought to overcome the limitations of traditional hyperparameter optimization techniques by leveraging a GA for superior predictive performance in CHD detection.

Methods: Utilizing a GA for hyperparameter optimization, we navigated a complex combinatorial space to identify optimal configurations for a CNN model. We also employed information gain for feature selection optimization, transforming the CHD datasets into an image-like input for the CNN architecture. The efficacy of this method was benchmarked against traditional optimization strategies.

Results: The advanced GA-based CNN model outperformed traditional methods, achieving a substantial increase in accuracy. The optimized model delivered a promising accuracy range, with a peak of 85% in hyperparameter optimization and 100% accuracy when integrated with machine learning algorithms, namely naïve Bayes, support vector machine, decision tree, logistic regression, and random forest, for both binary and multiclass CHD prediction tasks.

Conclusions: The integration of a GA into CNN feature engineering is a powerful technique for improving the accuracy of CHD predictions. This approach results in a high degree of predictive reliability and can significantly contribute to the field of AI-driven healthcare, with the possibility of clinical deployment for early CHD detection. Future work will focus on expanding the approach to encompass a wider set of CHD data and potential integration with wearable technology for continuous health monitoring.

Objectives: This review presents a comprehensive overview of the rapidly evolving digital healthcare industry, aiming to provide a broad understanding of the recent landscape and directions for the future of digital healthcare.

Methods: This review examines the key trends in sectors of the digital healthcare industry, which can be divided into four main categories: digital hardware, software solutions, platforms, and enablers. We discuss electroceuticals, wearables, standalone medical software, non-medical health management services, telehealth, decentralized clinical trials, and infrastructural systems such as health data systems. The review covers both global and domestic perspectives, addressing definitions, significance, revenue trends, major companies, regulations, and socioenvironmental factors.

Results: Diverse growth patterns are evident across digital healthcare sectors. The applications of electroceuticals are expanding. Wearables are becoming more ubiquitous, facilitating continuous health monitoring and data collection. Artificial intelligence in standalone medical software is demonstrating clinical efficacy, with regulatory frameworks adapting to support commercialization. Non-medical health management services are expanding their scope to address chronic conditions under professional guidance. Telemedicine and decentralized clinical trials are gaining traction, driven by the need for flexible healthcare solutions post-pandemic. Efforts to build robust digital infrastructure with health data are underway, supported by data banks and data aggregation platforms.

Conclusions: Advancements in digital healthcare create a dynamic, transformative landscape, integrating, complementing, and offering alternatives to traditional paradigms. This evolution is driven by continuous innovation, increased stakeholder participation, regulatory adaptations promoting commercialization, and supportive initiatives. Ongoing discussions about optimal digital technology integration and effective healthcare strategy implementation are essential for progress.

Objectives: Smoking remains the leading cause of preventable disease. However, smokers have shown poor compliance with smoking cessation clinics. Smartphone applications present a promising opportunity to improve this compliance. This study aimed to explore the relationship between nicotine dependence, smartphone usage patterns, and anticipated compliance with a smoking cessation application among smokers, with the goal of informing future development of such applications.

Methods: A total of 53 current smokers were surveyed using a questionnaire. Nicotine dependence was assessed using the Fagerstrom Test for Nicotine Dependence (FTND). Variables included the number of hours spent using a phone, willingness to quit smoking, number of previous quit attempts, desired number of text messages about smoking cessation, expected duration of application usage, and FTND scores. Kendall's partial correlation, adjusted for age, was employed for the analysis.

Results: The amount of time smokers spent on their mobile devices was negatively correlated with the number of smoking cessation text messages they wanted to receive (τ coefficient = -0.210, p = 0.026) and the duration they intended to use the cessation application (τ coefficient = -0.260, p = 0.006). Conversely, the number of desired text messages was positively correlated with the intended duration of application usage (τ coefficient = 0.366, p = 0.00012).

Conclusions: Smokers who spent more time on their mobile devices tended to prefer using the cessation application for shorter periods, whereas those who desired more text messages about smoking cessation were more inclined to use the application for longer durations.

Objectives: In Indonesia, the poor prognosis and high hospital readmission rates of patients with heart failure (HF) have yet to receive focused attention. However, machine learning (ML) approaches can help to mitigate these problems. We aimed to determine which ML models best predicted HF severity and hospital readmissions and could be used in a patient self-monitoring mobile application.

Methods: In a retrospective cohort study, we collected the data of patients admitted with HF to the Siloam Diagram Heart Center in 2020, 2021, and 2022. Data was analyzed using the Orange data mining classification method. ML support algorithms, including artificial neural network (ANN), random forest, gradient boosting, Naïve Bayes, tree-based models, and logistic regression were used to predict HF severity and hospital readmissions. The performance of these models was evaluated using the area under the curve (AUC), accuracy, and F1-scores.

Results: Of the 543 patients with HF, 3 (0.56%) were excluded due to death on admission. Hospital readmission occurred in 138 patients (25.6%). Of the six algorithms tested, ANN showed the best performance in predicting both HF severity (AUC = 1.000, accuracy = 0.998, F1-score = 0.998) and readmission for HF (AUC = 0.998, accuracy = 0.975, F1-score = 0.972). Other studies have shown variable results for the best algorithm to predict hospital readmission in patients with HF.

Conclusions: The ANN algorithm performed best in predicting HF severity and hospital readmissions and will be integrated into a mobile application for patient self-monitoring to prevent readmissions.

Objectives: This study compared the scientific publication speeds of Korean medical journals before and during the coronavirus disease 2019 (COVID-19) era.

Methods: We analyzed 2,064 papers from 43 international Korean medical journals, selecting 12 papers annually from 2019 to 2022. We assessed publication speed indicators, including the time from submission to revision and from submission to publication. Additionally, we examined variations in publication speed based on journal and paper characteristics, including whether the studies were related to COVID-19.

Results: Among the 43 journals analyzed, 39.5% disclosed the peer review duration from submission to the first decision, and 11.6% reported their acceptance rates. The average time from submission to acceptance was 127.0 days in 2019, 126.1 days in 2020, 124.6 days in 2021, and 126.4 days in 2022. For COVID-19-related studies, the average time from submission to revision was 61.4 days, compared to 105.1 days for non-COVID-19 studies; from submission to acceptance, it was 87.4 days for COVID-19-related studies and 127.1 days for non-COVID-19 studies. All indicators for COVID-19-related studies showed shorter durations than those for non-COVID-19 studies, and the proportion of studies accepted within 30 or 60 days was significantly higher for COVID-19-related studies.

Conclusions: This study investigated the publication speed of Korean international medical journals before and during the COVID-19 pandemic. The pandemic influenced journals' review and publication processes, potentially impacting the quality of academic papers. These findings provide insights into publication speeds during the COVID-19 era, suggesting that journals should focus on maintaining the integrity of their publication and review processes.