Extracting lung cancer staging descriptors from pathology reports: A generative language model approach.

BACKGROUND: In oncology, electronic health records contain textual key information for the diagnosis, staging, and treatment planning of patients with cancer. However, text data processing requires a lot of time and effort, which limits the utilization of these data. Recent advances in natural language processing (NLP) technology, including large language models, can be applied to cancer research. Particularly, extracting the information required for the pathological stage from surgical pathology reports can be utilized to update cancer staging according to the latest cancer staging guidelines. OBJECTIVES: This study has two main objectives. The first objective is to evaluate the performance of extracting information from text-based surgical pathology reports and determining pathological stages based on the extracted information using fine-tuned generative language models (GLMs) for patients with lung cancer. The second objective is to determine the feasibility of utilizing relatively small GLMs for information extraction in a resource-constrained computing environment. METHODS: Lung cancer surgical pathology reports were collected from the Common Data Model database of Seoul National University Bundang Hospital (SNUBH), a tertiary hospital in Korea. We selected 42 descriptors necessary for tumor-node (TN) classification based on these reports and created a gold standard with validation by two clinical experts. The pathology reports and gold standard were used to generate prompt-response pairs for training and evaluating GLMs which then were used to extract information required for staging from pathology reports. RESULTS: We evaluated the information extraction performance of six trained models as well as their performance in TN classification using the extracted information. The Deductive Mistral-7B model, which was pre-trained with the deductive dataset, showed the best performance overall, with an exact match ratio of 92.24% in the information extraction problem and an accuracy of 0.9876 (predicting T and N classification concurrently) in classification. CONCLUSION: This study demonstrated that training GLMs with deductive datasets can improve information extraction performance, and GLMs with a relatively small number of parameters at approximately seven billion can achieve high performance in this problem. The proposed GLM-based information extraction method is expected to be useful in clinical decision-making support, lung cancer staging and research.

Deep learning-based prediction of Clostridioides difficile infection caused by antibiotics using longitudinal electronic health records.

Clostridioides difficile infection (CDI) is a major cause of antibiotic-associated diarrhea and colitis. It is recognized as one of the most significant hospital-acquired infections. Although CDI can develop severe complications and spores of Clostridioides difficile can be transmitted by the fecal-oral route, CDI is occasionally overlooked in clinical settings. Thus, it is necessary to monitor high CDI risk groups, particularly those undergoing antibiotic treatment, to prevent complications and spread. We developed and validated a deep learning-based model to predict the occurrence of CDI within 28 days after starting antibiotic treatment using longitudinal electronic health records. For each patient, timelines of vital signs and laboratory tests with a 35-day monitoring period and a patient information vector consisting of age, sex, comorbidities, and medications were constructed. Our model achieved the prediction performance with an area under the receiver operating characteristic curve of 0.952 (95% CI: 0.932-0.973) in internal validation and 0.972 (95% CI: 0.968-0.975) in external validation. Platelet count and body temperature emerged as the most important features. The risk score, the output value of the model, exhibited a consistent increase in the CDI group, while the risk score in the non-CDI group either maintained its initial value or decreased. Using our CDI prediction model, high-risk patients requiring symptom monitoring can be identified. This could help reduce the underdiagnosis of CDI, thereby decreasing transmission and preventing complications.

Opportunities and challenges of a dynamic consent-based application: personalized options for personal health data sharing and utilization.

BACKGROUND: The principles of dynamic consent are based on the idea of safeguarding the autonomy of individuals by providing them with personalized options to choose from regarding the sharing and utilization of personal health data. To facilitate the widespread introduction of dynamic consent concepts in practice, individuals must perceive these procedures as useful and easy to use. This study examines the user experience of a dynamic consent-based application, in particular focusing on personalized options, and explores whether this approach may be useful in terms of ensuring the autonomy of data subjects in personal health data usage. METHODS: This study investigated the user experience of MyHealthHub, a dynamic consent-based application, among adults aged 18 years or older living in South Korea. Eight tasks exploring the primary aspects of dynamic consent principles-including providing consent, monitoring consent history, and managing personalized options were provided to participants. Feedback on the experiences of testing MyHealthHub was gathered via multiple-choice and open-ended questionnaire items. RESULTS: A total of 30 participants provided dynamic consent through the MyHealthHub application. Most participants successfully completed all the provided tasks without assistance and regarded the personalized options favourably. Concerns about the security and reliability of the digital-based consent system were raised, in contrast to positive responses elicited in other aspects, such as perceived usefulness and ease of use. CONCLUSIONS: Dynamic consent is an ethically advantageous approach for the sharing and utilization of personal health data. Personalized options have the potential to serve as pragmatic safeguards for the autonomy of individuals in the sharing and utilization of personal health data. Incorporating the principles of dynamic consent into real-world scenarios requires remaining issues, such as the need for powerful authentication mechanisms that bolster privacy and security, to be addressed. This would enhance the trustworthiness of dynamic consent-based applications while preserving their ethical advantages.

Prospective validation of a mobile health application for blood pressure management in patients with hypertensive disorders of pregnancy: study protocol for a randomized controlled trial.

BACKGROUND: Hypertensive disorders of pregnancy (HDP) pose significant risks to both maternal and fetal health, contributing to global morbidity and mortality. Management of HDP is complex, particularly because of concerns regarding potential negative effects on utero-placental circulation and limited therapeutic options due to fetal safety. Our study investigates whether blood pressure monitoring through a mobile health (mHealth) application can aid in addressing the challenges of blood pressure management in pregnant individuals with HDP. Additionally, we aim to assess whether this intervention can improve short-term maternal and fetal outcomes and potentially mitigate long-term cardiovascular consequences. METHODS: This prospective, randomized, single-center trial will include 580 pregnant participants who meet the HDP criteria or who have a heightened risk of pregnancy-related hypertension due to factors such as multiple pregnancies, obesity, diabetes, or a history of HDP in prior pregnancies leading to preterm birth. Participants will be randomized to either the mHealth intervention group or the standard care group. The primary endpoint is the difference in systolic blood pressure from enrollment to 1 month after childbirth. The secondary endpoints include various blood pressure parameters, obstetric outcomes, body mass index trajectory, step counts, mood assessment, and drug adherence. CONCLUSIONS: This study emphasizes the potential of mHealth interventions, such as the Heart4U application, to improve blood pressure management in pregnant individuals with HDP. By leveraging technology to enhance engagement, communication, and monitoring, this study aims to positively impact maternal, fetal, and postpartum outcomes associated with HDP. This innovative approach demonstrates the potential of personalized technology-driven solutions for managing complex health conditions. TRIAL REGISTRATION: ClinicalTrials.gov NCT05995106. Registered on 16 August 2023.

Integrating Clinical Data and Medical Imaging in Lung Cancer: Feasibility Study Using the Observational Medical Outcomes Partnership Common Data Model Extension.

BACKGROUND: Digital transformation, particularly the integration of medical imaging with clinical data, is vital in personalized medicine. The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) standardizes health data. However, integrating medical imaging remains a challenge. OBJECTIVE: This study proposes a method for combining medical imaging data with the OMOP CDM to improve multimodal research. METHODS: Our approach included the analysis and selection of digital imaging and communications in medicine header tags, validation of data formats, and alignment according to the OMOP CDM framework. The Fast Healthcare Interoperability Resources ImagingStudy profile guided our consistency in column naming and definitions. Imaging Common Data Model (I-CDM), constructed using the entity-attribute-value model, facilitates scalable and efficient medical imaging data management. For patients with lung cancer diagnosed between 2010 and 2017, we introduced 4 new tables-IMAGING_STUDY, IMAGING_SERIES, IMAGING_ANNOTATION, and FILEPATH-to standardize various imaging-related data and link to clinical data. RESULTS: This framework underscores the effectiveness of I-CDM in enhancing our understanding of lung cancer diagnostics and treatment strategies. The implementation of the I-CDM tables enabled the structured organization of a comprehensive data set, including 282,098 IMAGING_STUDY, 5,674,425 IMAGING_SERIES, and 48,536 IMAGING_ANNOTATION records, illustrating the extensive scope and depth of the approach. A scenario-based analysis using actual data from patients with lung cancer underscored the feasibility of our approach. A data quality check applying 44 specific rules confirmed the high integrity of the constructed data set, with all checks successfully passed, underscoring the reliability of our findings. CONCLUSIONS: These findings indicate that I-CDM can improve the integration and analysis of medical imaging and clinical data. By addressing the challenges in data standardization and management, our approach contributes toward enhancing diagnostics and treatment strategies. Future research should expand the application of I-CDM to diverse disease populations and explore its wide-ranging utility for medical conditions.

Information technology-supported integrated health service for older adults in long-term care settings.

BACKGROUND: To examine the effectiveness and safety of a data sharing and comprehensive management platform for institutionalized older patients. METHODS: We applied information technology-supported integrated health service platform to patients who live at long-term care hospitals (LTCHs) and nursing homes (NHs) with cluster randomized controlled study. We enrolled 555 patients aged 65 or older (461 from 7 LTCHs, 94 from 5 NHs). For the intervention group, a tablet-based platform comprising comprehensive geriatric assessment, disease management, potentially inappropriate medication (PIM) management, rehabilitation program, and screening for adverse events and warning alarms were provided for physicians or nurses. The control group was managed with usual care. Co-primary outcomes were (1) control rate of hypertension and diabetes, (2) medication adjustment (PIM prescription rate, proportion of polypharmacy), and (3) combination of potential quality-of-care problems (composite quality indicator) from the interRAI assessment system which assessed after 3-month of intervention. RESULTS: We screened 1119 patients and included 555 patients (control; 289, intervention; 266) for analysis. Patients allocated to the intervention group had better cognitive function and took less medications and PIMs at baseline. The diabetes control rate (OR = 2.61, 95% CI 1.37-4.99, p = 0.0035), discontinuation of PIM (OR = 4.65, 95% CI 2.41-8.97, p < 0.0001), reduction of medication in patients with polypharmacy (OR = 1.98, 95% CI 1.24-3.16, p = 0.0042), and number of PIMs use (ꞵ = - 0.27, p < 0.0001) improved significantly in the intervention group. There was no significant difference in hypertension control rate (OR = 0.54, 95% CI 0.20-1.43, p = 0.2129), proportion of polypharmacy (OR = 1.40, 95% CI 0.75-2.60, p = 0.2863), and improvement of composite quality indicators (ꞵ = 0.03, p = 0.2094). For secondary outcomes, cognitive and motor function, quality of life, and unplanned hospitalization were not different significantly between groups. CONCLUSIONS: The information technology-supported integrated health service effectively reduced PIM use and controlled diabetes among older patients in LTCH or NH without functional decline or increase of healthcare utilization. TRIAL REGISTRATION: Clinical Research Information Service, KCT0004360. Registered on 21 October 2019.

Symptoms and Conditions in Children and Adults up to 90 Days after SARS-CoV-2 Infection: A Retrospective Observational Study Utilizing the Common Data Model.

Background/Objectives: There have been widespread reports of persistent symptoms in both children and adults after SARS-CoV-2 infection, giving rise to debates on whether it should be regarded as a separate clinical entity from other postviral syndromes. This study aimed to characterize the clinical presentation of post-acute symptoms and conditions in the Korean pediatric and adult populations. Methods: A retrospective analysis was performed using a national, population-based database, which was encoded using the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). We compared individuals diagnosed with SARS-CoV-2 to those diagnosed with influenza, focusing on the risk of developing prespecified symptoms and conditions commonly associated with the post-acute sequelae of COVID-19. Results: Propensity score matching yielded 1,656 adult and 343 pediatric SARS-CoV-2 and influenza pairs. Ninety days after diagnosis, no symptoms were found to have elevated risk in either adults or children when compared with influenza controls. Conversely, at 1 day after diagnosis, adults with SARS-CoV-2 exhibited a significantly higher risk of developing abnormal liver function tests, cardiorespiratory symptoms, constipation, cough, thrombophlebitis/thromboembolism, and pneumonia. In contrast, children diagnosed with SARS-CoV-2 did not show an increased risk for any symptoms during either acute or post-acute phases. Conclusions: In the acute phase after infection, SARS-CoV-2 is associated with an elevated risk of certain symptoms in adults. The risk of developing post-acute COVID-19 sequelae is not significantly different from that of having postviral symptoms in children in both the acute and post-acute phases, and in adults in the post-acute phase. These observations warrant further validation through studies, including the severity of initial illness, vaccination status, and variant types.

Prediction of treatment response after stereotactic radiosurgery of brain metastasis using deep learning and radiomics on longitudinal MRI data.

We developed artificial intelligence models to predict the brain metastasis (BM) treatment response after stereotactic radiosurgery (SRS) using longitudinal magnetic resonance imaging (MRI) data and evaluated prediction accuracy changes according to the number of sequential MRI scans. We included four sequential MRI scans for 194 patients with BM and 369 target lesions for the Developmental dataset. The data were randomly split (8:2 ratio) for training and testing. For external validation, 172 MRI scans from 43 patients with BM and 62 target lesions were additionally enrolled. The maximum axial diameter (Dmax), radiomics, and deep learning (DL) models were generated for comparison. We evaluated the simple convolutional neural network (CNN) model and a gated recurrent unit (Conv-GRU)-based CNN model in the DL arm. The Conv-GRU model performed superior to the simple CNN models. For both datasets, the area under the curve (AUC) was significantly higher for the two-dimensional (2D) Conv-GRU model than for the 3D Conv-GRU, Dmax, and radiomics models. The accuracy of the 2D Conv-GRU model increased with the number of follow-up studies. In conclusion, using longitudinal MRI data, the 2D Conv-GRU model outperformed all other models in predicting the treatment response after SRS of BM.

Impact of a Nationwide Medication History Sharing Program on the Care Process and End-User Experience in a Tertiary Teaching Hospital: Cohort Study and Cross-Sectional Study.

Background: Timely and comprehensive collection of a patient's medication history in the emergency department (ED) is crucial for optimizing health care delivery. The implementation of a medication history sharing program, titled "Patient's In-home Medications at a Glance," in a tertiary teaching hospital aimed to efficiently collect and display nationwide medication histories for patients' initial hospital visits. Objective: As an evaluation was necessary to provide a balanced picture of the program, we aimed to evaluate both care process outcomes and humanistic outcomes encompassing end-user experience of physicians and pharmacists. Methods: We conducted a cohort study and a cross-sectional study to evaluate both outcomes. To evaluate the care process, we measured the time from the first ED assessment to urgent percutaneous coronary intervention (PCI) initiation from electronic health records. To assess end-user experience, we developed a 22-item questionnaire using a 5-point Likert scale, including 5 domains: information quality, system quality, service quality, user satisfaction, and intention to reuse. This questionnaire was validated and distributed to physicians and pharmacists. The Mann-Whiteny U test was used to analyze the PCI initiation time, and structural equation modeling was used to assess factors affecting end-user experience. Results: The time from the first ED assessment to urgent PCI initiation at the ED was significantly decreased using the patient medication history program (mean rank 42.14 min vs 28.72 min; Mann-Whitney U=346; P=.03). A total of 112 physicians and pharmacists participated in the survey. Among the 5 domains, "intention to reuse" received the highest score (mean 4.77, SD 0.37), followed by "user satisfaction" (mean 4.56, SD 0.49), while "service quality" received the lowest score (mean 3.87, SD 0.79). "User satisfaction" was significantly associated with "information quality" and "intention to reuse." Conclusions: Timely and complete retrieval using a medication history-sharing program led to an improved care process by expediting critical decision-making in the ED, thereby contributing to value-based health care delivery in a real-world setting. The experiences of end users, including physicians and pharmacists, indicated satisfaction with the program regarding information quality and their intention to reuse.

Impact of metformin on cardiovascular and kidney outcome based on kidney function status in type 2 diabetic patients: a multicentric, retrospective cohort study.

Metformin is the primary treatment for type 2 diabetes mellitus (T2DM) due to its effectiveness in improving clinical outcomes in patients with preserved renal function, however, the evidence on the effectiveness of metformin in various renal functions is lacking. We performed a retrospective, multicenter, observational study used data of patients with T2DM obtained from three tertiary hospitals' databases. Patients given metformin within run-in periods and with at least one additional prescription formed the metformin cohort. A control cohort comprised those prescribed oral hypoglycemic agents other than metformin and never subsequently received a metformin prescription within observation period. For patients without diabetic nephropathy (DN), the outcomes included events of DN, major adverse cardiovascular events (MACE), and major adverse kidney events (MAKE). After 1:1 propensity matching, 1994 individuals each were selected for the metformin and control cohorts among T2DM patients without baseline DN. The incidence rate ratios (IRR) for DN, MACEs, and MAKEs between cohorts were 1.06 (95% CI 0.96-1.17), 0.76 (0.64-0.92), and 0.45 (0.33-0.62), respectively. In cohorts with renal function of CKD 3A, 3B, and 4, summarized IRRs of MACEs and MAKEs were 0.70 (0.57-0.87) and 0.39 (0.35-0.43) in CKD 3A, 0.83 (0.74-0.93) and 0.44 (0.40-0.48) in CKD 3B, and 0.71 (0.60-0.85) and 0.45 (0.39-0.51) in CKD 4. Our research indicates that metformin use in T2DM patients across various renal functions consistently correlates with a decreased risk of overt DN, MACE, and MAKE.

Identifying facilitators of and barriers to the adoption of dynamic consent in digital health ecosystems: a scoping review.

BACKGROUND: Conventional consent practices face ethical challenges in continuously evolving digital health environments due to their static, one-time nature. Dynamic consent offers a promising solution, providing adaptability and flexibility to address these ethical concerns. However, due to the immaturity of the concept and accompanying technology, dynamic consent has not yet been widely used in practice. This study aims to identify the facilitators of and barriers to adopting dynamic consent in real-world scenarios. METHODS: This scoping review, conducted in December 2022, adhered to the PRISMA Extension for Scoping Reviews guidelines, focusing on dynamic consent within the health domain. A comprehensive search across Web of Science, PubMed, and Scopus yielded 22 selected articles based on predefined inclusion and exclusion criteria. RESULTS: The facilitators for the adoption of dynamic consent in digital health ecosystems were the provision of multiple consent modalities, personalized alternatives, continuous communication, and the dissemination of up-to-date information. Nevertheless, several barriers, such as consent fatigue, the digital divide, complexities in system implementation, and privacy and security concerns, needed to be addressed. This study also investigated current technological advancements and suggested considerations for further research aimed at resolving the remaining challenges surrounding dynamic consent. CONCLUSIONS: Dynamic consent emerges as an ethically advantageous method for digital health ecosystems, driven by its adaptability and support for continuous, two-way communication between data subjects and consumers. Ethical implementation in real-world settings requires the development of a robust technical framework capable of accommodating the diverse needs of stakeholders, thereby ensuring ethical integrity and data privacy in the evolving digital health landscape.

Risk Prediction of Emergency Department Visits in Patients With Lung Cancer Using Machine Learning: Retrospective Observational Study.

BACKGROUND: Patients with lung cancer are among the most frequent visitors to emergency departments due to cancer-related problems, and the prognosis for those who seek emergency care is dismal. Given that patients with lung cancer frequently visit health care facilities for treatment or follow-up, the ability to predict emergency department visits based on clinical information gleaned from their routine visits would enhance hospital resource utilization and patient outcomes. OBJECTIVE: This study proposed a machine learning-based prediction model to identify risk factors for emergency department visits by patients with lung cancer. METHODS: This was a retrospective observational study of patients with lung cancer diagnosed at Seoul National University Bundang Hospital, a tertiary general hospital in South Korea, between January 2010 and December 2017. The primary outcome was an emergency department visit within 30 days of an outpatient visit. This study developed a machine learning-based prediction model using a common data model. In addition, the importance of features that influenced the decision-making of the model output was analyzed to identify significant clinical factors. RESULTS: The model with the best performance demonstrated an area under the receiver operating characteristic curve of 0.73 in its ability to predict the attendance of patients with lung cancer in emergency departments. The frequency of recent visits to the emergency department and several laboratory test results that are typically collected during cancer treatment follow-up visits were revealed as influencing factors for the model output. CONCLUSIONS: This study developed a machine learning-based risk prediction model using a common data model and identified influencing factors for emergency department visits by patients with lung cancer. The predictive model contributes to the efficiency of resource utilization and health care service quality by facilitating the identification and early intervention of high-risk patients. This study demonstrated the possibility of collaborative research among different institutions using the common data model for precision medicine in lung cancer.

Patient-Centered Economic Burden of Exudative Age-Related Macular Degeneration: Retrospective Cohort Study.

BACKGROUND: Exudative age-related macular degeneration (AMD), one of the leading causes of blindness, requires expensive drugs such as anti-vascular endothelial growth factor (VEGF) agents. The long-term regular use of effective but expensive drugs causes an economic burden for patients with exudative AMD. However, there are no studies on the long-term patient-centered economic burden of exudative AMD after reimbursement of anti-VEGFs. OBJECTIVE: This study aimed to evaluate the patient-centered economic burden of exudative AMD for 2 years, including nonreimbursement and out-of-pocket costs, compared with nonexudative AMD using the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM). METHODS: This retrospective cohort study was conducted using the OMOP CDM, which included 2,006,478 patients who visited Seoul National University Bundang Hospital from June 2003 to July 2019. We defined the exudative AMD group as patients aged >50 years with a diagnosis of exudative AMD and a prescription for anti-VEGFs or verteporfin. The control group was defined as patients aged >50 years without a diagnosis of exudative AMD or a prescription for anti-VEGFs or verteporfin. To adjust for selection bias, controls were matched by propensity scores using regularized logistic regression with a Laplace prior. We measured any medical cost occurring in the hospital as the economic burden of exudative AMD during a 2-year follow-up period using 4 categories: total medical cost, reimbursement cost, nonreimbursement cost, and out-of-pocket cost. To estimate the average cost by adjusting the confounding variable and overcoming the positive skewness of costs, we used an exponential conditional model with a generalized linear model. RESULTS: We identified 931 patients with exudative AMD and matched 783 (84.1%) with 2918 patients with nonexudative AMD. In the exponential conditional model, the total medical, reimbursement, nonreimbursement, and out-of-pocket incremental costs were estimated at US $3426, US $3130, US $366, and US $561, respectively, in the first year and US $1829, US $1461, US $373, and US $507, respectively, in the second year. All incremental costs in the exudative AMD group were 1.89 to 4.25 and 3.50 to 5.09 times higher in the first and second year, respectively, than those in the control group (P<.001 in all cases). CONCLUSIONS: Exudative AMD had a significantly greater economic impact (P<.001) for 2 years on reimbursement, nonreimbursement, and out-of-pocket costs than nonexudative AMD after adjusting for baseline demographic and clinical characteristics using the OMOP CDM. Although economic policies could relieve the economic burden of patients with exudative AMD over time, the out-of-pocket cost of exudative AMD was still higher than that of nonexudative AMD for 2 years. Our findings support the need for expanding reimbursement strategies for patients with exudative AMD given the significant economic burden faced by patients with incurable and fatal diseases both in South Korea and worldwide.

Development and Validation of a Prognostic Classification Model Predicting Postoperative Adverse Outcomes in Older Surgical Patients Using a Machine Learning Algorithm: Retrospective Observational Network Study.

BACKGROUND: Older adults are at an increased risk of postoperative morbidity. Numerous risk stratification tools exist, but effort and manpower are required. OBJECTIVE: This study aimed to develop a predictive model of postoperative adverse outcomes in older patients following general surgery with an open-source, patient-level prediction from the Observational Health Data Sciences and Informatics for internal and external validation. METHODS: We used the Observational Medical Outcomes Partnership common data model and machine learning algorithms. The primary outcome was a composite of 90-day postoperative all-cause mortality and emergency department visits. Secondary outcomes were postoperative delirium, prolonged postoperative stay (≥75th percentile), and prolonged hospital stay (≥21 days). An 80% versus 20% split of the data from the Seoul National University Bundang Hospital (SNUBH) and Seoul National University Hospital (SNUH) common data model was used for model training and testing versus external validation. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) with a 95% CI. RESULTS: Data from 27,197 (SNUBH) and 32,857 (SNUH) patients were analyzed. Compared to the random forest, Adaboost, and decision tree models, the least absolute shrinkage and selection operator logistic regression model showed good internal discriminative accuracy (internal AUC 0.723, 95% CI 0.701-0.744) and transportability (external AUC 0.703, 95% CI 0.692-0.714) for the primary outcome. The model also possessed good internal and external AUCs for postoperative delirium (internal AUC 0.754, 95% CI 0.713-0.794; external AUC 0.750, 95% CI 0.727-0.772), prolonged postoperative stay (internal AUC 0.813, 95% CI 0.800-0.825; external AUC 0.747, 95% CI 0.741-0.753), and prolonged hospital stay (internal AUC 0.770, 95% CI 0.749-0.792; external AUC 0.707, 95% CI 0.696-0.718). Compared with age or the Charlson comorbidity index, the model showed better prediction performance. CONCLUSIONS: The derived model shall assist clinicians and patients in understanding the individualized risks and benefits of surgery.

Snoring-related polygenic risk and its relationship with lifestyle factors in a Korean population: KoGES study.

Whereas lifestyle-related factors are recognized as snoring risk factors, the role of genetics in snoring remains uncertain. One way to measure the impact of genetic risk is through the use of a polygenic risk score (PRS). In this study, we aimed to investigate whether genetics plays a role in snoring after adjusting for lifestyle factors. Since the effect of polygenic risks may differ across ethnic groups, we calculated the PRS for snoring from the UK Biobank and applied it to a Korean cohort. We sought to evaluate the reproducibility of the UK Biobank PRS for snoring in the Korean cohort and to investigate the interaction of lifestyle factors and genetic risk on snoring in the Korean population. In this study, we utilized a Korean cohort obtained from the Korean Genome Epidemiology Study (KoGES). We computed the snoring PRS for the Korean cohort based on the UK Biobank PRS. We investigated the relationship between polygenic risks and snoring while controlling for lifestyle factors, including sex, age, body mass index (BMI), alcohol consumption, smoking, physical activity, and sleep time. Additionally, we analyzed the interaction of each lifestyle factor and the genetic odds of snoring. We included 3526 snorers and 1939 nonsnorers from the KoGES cohort and found that the PRS, a polygenic risk factor, was an independent factor for snoring after adjusting for lifestyle factors. In addition, among lifestyle factors, higher BMI, male sex, and older age were the strongest lifestyle factors for snoring. In addition, the highest adjusted odds ratio for snoring was higher BMI (OR 1.98, 95% CI 1.76-2.23), followed by male sex (OR 1.54, 95% CI 1.28-1.86), older age (OR 1.23, 95% CI 1.03-1.35), polygenic risks such as higher PRS (OR 1.18, 95% CI 1.08-1.29), drinking behavior (OR 1.18, 95% CI 1.03-1.35), late sleep mid-time (OR 1.17, 95% CI 1.02-1.33), smoking behavior (OR 0.99, 95% CI 0.82-1.19), and lower physical activity (OR 0.92, 95% CI 0.85-1.00). Our study identified that the UK Biobank PRS for snoring was reproducible in the Korean cohort and that genetic risk served as an independent risk factor for snoring in the Korean population. These findings may help to develop personalized approaches to reduce snoring in individuals with high genetic risk.

Development and validation of a prediction model for evaluating extubation readiness in preterm infants.

Successful early extubation has advantages not only in terms of short-term respiratory morbidities and survival but also in terms of long-term neurodevelopmental outcomes in preterm infants. However, no consensus exists regarding the optimal protocol or guidelines for extubation readiness in preterm infants. Therefore, the decision to extubate preterm infants was almost entirely at the attending physician's discretion. We identified robust and quantitative predictors of success or failure of the first planned extubation attempt before 36 weeks of post-menstrual age in preterm infants (<32 weeks gestational age) and developed a prediction model for evaluating extubation readiness using these predictors. Extubation success was defined as the absence of reintubation within 72 h after extubation. This observational cohort study used data from preterm infants admitted to the neonatal intensive care unit of Seoul National University Bundang Hospital in South Korea between July 2003 and June 2019 to identify predictors and develop and test a predictive model for extubation readiness. Data from preterm infants included in the Medical Informative Medicine for Intensive Care (MIMIC-III) database between 2001 and 2008 were used for external validation. From a machine learning model using predictors such as demographics, periodic vital signs, ventilator settings, and respiratory indices, the area under the receiver operating characteristic curve and average precision of our model were 0.805 (95% confidence interval [CI], 0.802-0.809) and 0.917, respectively in the internal validation and 0.715 (95% CI, 0.713-0.717) and 0.838, respectively in the external validation. Our prediction model (NExt-Predictor) demonstrated high performance in assessing extubation readiness in both internal and external validations.

Current Status and Key Issues of Data Management in Tertiary Hospitals: A Case Study of Seoul National University Hospital.

OBJECTIVES: In the era of the Fourth Industrial Revolution, where an ecosystem is being developed to enhance the quality of healthcare services by applying information and communication technologies, systematic and sustainable data management is essential for medical institutions. In this study, we assessed the data management status and emerging concerns of three medical institutions, while also examining future directions for seamless data management. METHODS: To evaluate the data management status, we examined data types, capacities, infrastructure, backup methods, and related organizations. We also discussed challenges, such as resource and infrastructure issues, problems related to government regulations, and considerations for future data management. RESULTS: Hospitals are grappling with the increasing data storage space and a shortage of management personnel due to costs and project termination, which necessitates countermeasures and support. Data management regulations on the destruction or maintenance of medical records are needed, and institutional consideration for secondary utilization such as long-term treatment or research is required. Government-level guidelines for facilitating hospital data sharing and mobile patient services should be developed. Additionally, hospital executives at the organizational level need to make efforts to facilitate the clinical validation of artificial intelligence software. CONCLUSIONS: This analysis of the current status and emerging issues of data management reveals potential solutions and sets the stage for future organizational and policy directions. If medical big data is systematically managed, accumulated over time, and strategically monetized, it has the potential to create new value.

Impact of glycemic control on the progression of aortic stenosis: a single-center cohort study using a common data model.

BACKGROUND: Diabetes mellitus (DM) is a well-established risk factor for the progression of degenerative aortic stenosis (AS). However, no study has investigated the impact of glycemic control on the rate of AS progression. We aimed to assess the association between the degree of glycemic control and the AS progression, using an electronic health record-based common data model (CDM). METHODS: We identified patients with mild AS (aortic valve [AV] maximal velocity [Vpeak] 2.0-3.0 m/sec) or moderate AS (Vpeak 3.0-4.0 m/sec) at baseline, and follow-up echocardiography performed at an interval of ≥ 6 months, using the CDM of a tertiary hospital database. Patients were divided into 3 groups: no DM (n = 1,027), well-controlled DM (mean glycated hemoglobin [HbA1c] < 7.0% during the study period; n = 193), and poorly controlled DM (mean HbA1c ≥ 7.0% during the study period; n = 144). The primary outcome was the AS progression rate, calculated as the annualized change in the Vpeak (△Vpeak/year). RESULTS: Among the total study population (n = 1,364), the median age was 74 (IQR 65-80) years, 47% were male, the median HbA1c was 6.1% (IQR 5.6-6.9), and the median Vpeak was 2.5 m/sec (IQR 2.2-2.9). During follow-up (median 18.4 months), 16.1% of the 1,031 patients with mild AS at baseline progressed to moderate AS, and 1.8% progressed to severe AS. Among the 333 patients with moderate AS, 36.3% progressed to severe AS. The mean HbA1c level during follow-up showed a positive relationship with the AS progression rate (ß = 2.620; 95% confidence interval [CI] 0.732-4.507; p = 0.007); a 1%-unit increase in HbA1c was associated with a 27% higher risk of accelerated AS progression defined as △Vpeak/year values > 0.2 m/sec/year (adjusted OR = 1.267 per 1%-unit increase in HbA1c; 95% CI 1.106-1.453; p < 0.001), and HbA1c ≥ 7.0% was significantly associated with an accelerated AS progression (adjusted odds ratio = 1.524; 95% CI 1.010-2.285; p = 0.043). This association between the degree of glycemic control and AS progression rate was observed regardless of the baseline AS severity. CONCLUSION: In patients with mild to moderate AS, the presence of DM, as well as the degree of glycemic control, is significantly associated with accelerated AS progression.

Automated detection of intracranial aneurysms using skeleton-based 3D patches, semantic segmentation, and auxiliary classification for overcoming data imbalance in brain TOF-MRA.

Accurate and reliable detection of intracranial aneurysms is vital for subsequent treatment to prevent bleeding. However, the detection of intracranial aneurysms can be time-consuming and even challenging, and there is great variability among experts, especially in the case of small aneurysms. This study aimed to detect intracranial aneurysms accurately using a convolutional neural network (CNN) with 3D time-of-flight magnetic resonance angiography (TOF-MRA). A total of 154 3D TOF-MRA datasets with intracranial aneurysms were acquired, and the gold standards were manually drawn by neuroradiologists. We also obtained 113 subjects from a public dataset for external validation. These angiograms were pre-processed by using skull-stripping, signal intensity normalization, and N4 bias correction. The 3D patches along the vessel skeleton from MRA were extracted. Values of the ratio between the aneurysmal and the normal patches ranged from 1:1 to 1:5. The semantic segmentation on intracranial aneurysms was trained using a 3D U-Net with an auxiliary classifier to overcome the imbalance in patches. The proposed method achieved an accuracy of 0.910 in internal validation and external validation accuracy of 0.883 with a 2:1 ratio of normal to aneurysmal patches. This multi-task learning method showed that the aneurysm segmentation performance was sufficient to be helpful in an actual clinical setting.