JAMIA Open最新文献

{"title":"Comparing audiovisual visit and in-person visit note quality.","authors":"Harry B Burke, Albert Hoang, Heidi King, Joseph Benich, Brandon Brown, Christopher Bunt, Sean Haley, Paul Hemmer, William Kelly, Renee Mallory, Louis Pangaro, Wendy Shen, Viktoria Gagarin","doi":"10.1093/jamiaopen/ooag014","DOIUrl":"https://doi.org/10.1093/jamiaopen/ooag014","url":null,"abstract":"<p><strong>Background: </strong>The pandemic dramatically increased in the frequency of audiovisual medical visits and the rate of audiovisual visits remains higher than before the pandemic. These visits have the potential to be an important clinical modality. Researchers have assessed several aspects of audiovisual visits but they have not investigated their clinical note quality. The goal of this study is to determine the quality of audiovisual and in-person clinical notes and to compare their quality.</p><p><strong>Methods: </strong>From a population of 1660 established outpatient primary care type 2 diabetic patient visits occurring in 2021 we randomly selected 100 audiovisual and 100 in-person visits. QNOTE, a validated instrument that measures clinical note quality, was used by 7 experienced primary care physicians to assess 4 key elements of the clinical note: the chief complaint, history of present illness, assessment, and plan.</p><p><strong>Results: </strong>The mean quality scores (out of 100) and their standard deviations (SD) were: overall, audiovisual 75.8 (SD 9.6), in-person 86.5 (SD 4.4), <i>P</i> < .0001; chief complaint, audiovisual 72.4 (SD 23.0), in-person 83.0 (SD 14.7), <i>P</i> < .0001; history of present illness, audiovisual 63.7 (SD 22.9), in-person 82.5 (SD, 15.8), <i>P</i> < .0001; assessment, audiovisual 82.9 (SD 14.8), in-person 89.8 (SD 10.0), <i>P</i> = .0002; and plan, audiovisual 84.1 (SD 13.7), in-person 90.9 (SD 8.7), <i>P</i> < .0001. The in-person scores are consistent with a previous QNOTE study. The rater intraclass correlation coefficient was excellent (0.81, 95% CI, 0.76-0.86).</p><p><strong>Conclusion: </strong>Audiovisual visits demonstrated a lower note quality than in-person visits. To our knowledge, this is the first study to assess the quality of audiovisual visit notes and compare them to the quality of in-person visit notes.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag014"},"PeriodicalIF":3.4,"publicationDate":"2026-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13127419/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147820849","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Area-specific autoencoder spatiotemporal graph neural networks for opioid overdose death prediction.","authors":"Xianhui Chen, Changchang Yin, John V Myers, Brandon Slover, Neena Thomas, Charles Marks, Joanne Kim, Soledad Fernández, Penn Whitley, Naleef Fareed, Ping Zhang","doi":"10.1093/jamiaopen/ooag063","DOIUrl":"10.1093/jamiaopen/ooag063","url":null,"abstract":"<p><strong>Background: </strong>Ohio has been severely impacted by the opioid crisis, with opioid overdose (OD) death rates exceeding national averages. Accurate OD death prediction supports proactive prevention and treatment allocation. Existing methods often focus on ZIP Code Tabulation Area (ZCTA)-level prediction for small-area resource allocation; however, performance at this resolution is poor due to substantial fluctuations in OD death counts, which introduce noise. This raises a critical methodological question: what is the optimal population threshold for OD death prediction that balances predictive accuracy with geographic resolution?</p><p><strong>Methods: </strong>We perform a theoretical analysis of variance and error bounds to establish the minimal population required for robust prediction. Building on this analysis, we propose an Area-specific Autoencoder Spatiotemporal Graph Neural Network (AAE-STGNN) framework for OD death count prediction using urine drug test (UDT) data as dynamic features and Social Determinants of Health (SDoH) as static features. The framework consists of two key components: (1) an <i>Area-specific Autoencoder (AAE)</i>, which learns latent spatial representations while incorporating the minimal population threshold, and (2) a <i>Spatiotemporal Graph Neural Network (STGNN)</i>, which models geographic adjacency between areas and dynamic features across time.</p><p><strong>Results: </strong>Empirical evaluations demonstrate that AAE-STGNN outperforms state-of-the-art (SOTA) approaches, achieving improved accuracy and robustness. We also provide the OD death count trend estimation to support public health decision-making.</p><p><strong>Conclusions: </strong>These findings underscore the importance of selecting an optimal spatial granularity and leveraging spatiotemporal modeling techniques for data-driven public health surveillance and targeted intervention in the opioid crisis.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag063"},"PeriodicalIF":3.4,"publicationDate":"2026-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13127421/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147821818","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large language models for automated and audience-tailored labeling of latent classes.","authors":"Fatemeh Gholi Zadeh Kharrat, Rob Werfelmann, Glen Ep Ropella, Wolf Mehling, C Anthony Hunt, Jeffrey Lotz, Thomas A Peterson","doi":"10.1093/jamiaopen/ooag058","DOIUrl":"https://doi.org/10.1093/jamiaopen/ooag058","url":null,"abstract":"<p><strong>Objective: </strong>This study compares multiple LLMs, including ChatGPT, DeepSeek, and Llama, to generate meaningful, audience-adapted labels for the existing latent classes among patients with chronic low back pain (cLBP).</p><p><strong>Methods: </strong>Phenotypes were derived from baseline data from two cohorts within the NIH HEAL BACPAC consortium: BACKHOME, a large nationwide e-cohort (train set: <i>N</i> = 3025), and COMEBACK, a deep phenotyping cohort (test set: <i>N</i> = 450). The analysis included pain characteristics, psychosocial factors, lifestyle habits, and social determinants of health. ChatGPT-4o (OpenAI), DeepSeek-R1, and Llama 3.3 (Meta) were applied to generate class labels for each combination of audience (clinician, patient, and caregiver), tone (formal, empathetic, and informal), and technicality (high, medium, and low).</p><p><strong>Results: </strong>Latent Class Model (LCM) identified four distinct behavioral phenotypes in patients with cLBP: <i>High Distress and Maladaptive Behaviors</i>, <i>Resilient and Adaptive Coping</i>, <i>Intermediate Maladaptive Patterns</i>, and <i>Emotionally Regulated with High Pain Burden</i>. Previously validated by domain experts, these profiles served as the basis for automated labeling using three LLMs (ChatGPT-4o, DeepSeek-R1, and Llama 3.3). Using different tones and complexity levels, each model produced class labels specific to clinicians, patients, and caregivers. The generated class names for all LLMs closely matched expert-defined traits like <i>emotional regulation</i>, <i>resilience</i>, and <i>high distress</i>, indicating strong conceptual alignment and the capacity of LLMs to generate precise, audience-specific labels for intricate behavioral and psychological profiles.</p><p><strong>Conclusions: </strong>These results highlight the possibility of integrating LLM-driven labeling into research and clinical practice, helping to achieve more transparent knowledge translation, improved decision-making, and personalized care.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag058"},"PeriodicalIF":3.4,"publicationDate":"2026-04-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13127414/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147821036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Text mining methods for automated data extraction from health technology assessment reports of medicines using classical natural language processing and generative artificial intelligence.","authors":"Jan-Willem Versteeg, Marie L De Bruin, Maarten Schermer, Shiva Nadi Najafabadi, Modhurita Mitra, Christine Leopold, Aukje K Mantel-Teeuwisse, Wim G Goettsch, Lourens T Bloem","doi":"10.1093/jamiaopen/ooag051","DOIUrl":"https://doi.org/10.1093/jamiaopen/ooag051","url":null,"abstract":"<p><strong>Objective: </strong>This proof of concept for utilizing automatic data extraction methods to extract health technology assessment (HTA) attributes from HTA reports of medicines aimed to explore which attributes could be extracted and how accurately, using different data extraction methods. This enables easy access to insights into HTA recommendations for policymaking and policy-related research.</p><p><strong>Materials and methods: </strong>In total, 14 relevant attributes (eg, assessment outcome or date) were identified for extraction using two classical natural language processing (NLP) methods (rule-based and classification models) and a generative AI method (large language model (LLM)-based, i.e., Claude 3 Opus). The performance of these techniques was compared using 50 HTA reports published by the National Institute for Health and Care Excellence (NICE, United Kingdom).</p><p><strong>Results: </strong>All three methods were able to extract certain attributes with high accuracy, with differences between the extraction methods and the type of attribute. The LLM-based extraction was the only method able to extract attributes on a medicine-indication combination level. The LLM-based extraction performed best (88-98% semantical accuracy for 12/14 attributes). Extraction of Outcome relative effectiveness analyses (REA) and Comparator was the most challenging and had the lowest accuracy (∼70% for the LLM-based extraction).</p><p><strong>Discussion & conclusion: </strong>Automatic data extraction for relevant attributes from HTA reports is possible, but there is still room for improvement. LLM-based extraction outperformed the two NLP methods, but challenges regarding the use of commercial software and reproducibility remain. Future research should focus on expanding the system to other HTA organizations and further refining the LLM-based extraction.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag051"},"PeriodicalIF":3.4,"publicationDate":"2026-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13120885/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147783741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring nurses' documentation prioritization strategies to alleviate EHR documentation burden: a phenomenological study.","authors":"Rosemary Mugoya, Jennifer Thate, Fan Hao, Sarah C Rossetti, Po-Yin Yen","doi":"10.1093/jamiaopen/ooag056","DOIUrl":"https://doi.org/10.1093/jamiaopen/ooag056","url":null,"abstract":"<p><strong>Objective: </strong>This study aims to understand how inpatient nurses determine and prioritize necessary documentation within the context of the Excessive Documentation Burden (ExDocBurden) in Electronic Health Records (EHRs).</p><p><strong>Methodology: </strong>A phenomenological approach was used to explore inpatient nurses' lived experiences of prioritizing EHR documentation. Interpretive phenomenology guided the study design, focusing on how nurses prioritize documentation. Purposive sampling recruited 14 registered nurses (RNs) from acute and critical care settings. Data was collected through semi-structured interviews and analyzed using Colaizzi's 7-step and Smith's Interpretive Phenomenology Analysis.</p><p><strong>Results: </strong>Five themes emerged: (1) Advocating for Quality Patient Care Environment and Patient Needs, (2) What to Document in Near-Real Time Versus What Can Wait, (3) EHR-Driven Documentation and the Erosion of Nurse Autonomy, (4) Unnecessary (Frequent and Redundant) Documentation, and (5) Fear, Frustration, and Punitive Pressure in Charting. Nurses prioritized patient care over EHR documentation and frequently encountered unnecessary and redundant documentation tasks that did not contribute to patient needs. Defensive charting practices driven by fear of litigation further compounded nurses' emotional strain.</p><p><strong>Discussion: </strong>The study emphasizes the importance of empowering nurses by minimizing non-value-added documentation and enabling them to exercise their clinical judgment. Streamlining documentation processes can help alleviate the emotional and mental strain on nurses, enabling a more patient-centered approach to care.</p><p><strong>Conclusion: </strong>Understanding how experienced nurses prioritize documentation in the context of ExDocBurden provides valuable insights to ameliorate EHR Burden. Nurses drive quality of patient care; consequently, supporting nurse-driven documentation enhances both patient care quality and organizational needs.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag056"},"PeriodicalIF":3.4,"publicationDate":"2026-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13120879/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147783675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Feasibility of identifying factors related to Alzheimer's disease and related dementia in real-world data.","authors":"Yu Huang, Qian Li, Aokun Chen, Yongqiu Li, Yu-Neng Chuang, Xia Hu, Serena Jingchuan Guo, Xing He, Yijiang Pang, Jiayu Zhou, Yonghui Wu, Yi Guo, Jiang Bian","doi":"10.1093/jamiaopen/ooag060","DOIUrl":"10.1093/jamiaopen/ooag060","url":null,"abstract":"<p><strong>Objective: </strong>This study aimed to provide a comprehensive understanding of factors associated with Alzheimer's disease (AD) and AD-related dementias (AD/ADRD), which could aid in studies to develop new treatments for AD/ADRD patients and identify high-risk populations for prevention.</p><p><strong>Scope and method: </strong>In our study, we summarized the risk factors for AD/ADRD by reviewing existing meta-analyses and review articles on risk and preventive factors for AD/ADRD. From this literature review and the identified AD/ADRD factors, we examined the accessibility of these risk and preventive factors in both structured and unstructured Electronic Health Records (EHRs) data.</p><p><strong>Results: </strong>In total, we extracted 401 factors in 10 categories from the identified studies. To share our findings, we created an interactive knowledge graph of these risk factors and the relationships among them to assist in the design of future AD/ADRD studies that aim to use large collections of real-world data (RWD) to generate real-world evidence (RWE).</p><p><strong>Discussion and conclusion: </strong>Most factors, including conditions, medications, biomarkers, and procedures, are accessible from structured EHRs. For those not accessible from structured EHRs, clinical narratives serve as promising sources of information. However, evaluating genomic factors using RWD remains to be a challenge, possibly due to the fact that genetic testing for AD/ADRD is still uncommon and poorly documented in both structured and unstructured EHRs. Considering the continuously and rapidly evolving research on AD/ADRD, automated literature mining via natural language processing (NLP) methods offers a way to automatically update our knowledge graph.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag060"},"PeriodicalIF":3.4,"publicationDate":"2026-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13108722/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147783683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Conceptualization and development of the Maryland suicide data warehouse to improve mortality prediction.","authors":"Christopher Kitchen, Paul Nestadt, Holly C Wilcox, Tom Richards, Elyse C Lasser, Hadi Kharrazi","doi":"10.1093/jamiaopen/ooag062","DOIUrl":"https://doi.org/10.1093/jamiaopen/ooag062","url":null,"abstract":"<p><strong>Objectives: </strong>Research in suicide risk prediction often suffers from the lack of comprehensive data on patient suicide death, which differs from suicide attempt or suicidal behaviors. This study aimed to develop a population-wide multi-source harmonized data warehouse suitable for suicide death risk prediction.</p><p><strong>Materials and methods: </strong>The Maryland Suicide Data Warehouse (MSDW) was conceived as a statewide database that addresses limitations in prior suicide research. To develop MSDW, multiple patient-level statewide data sources were linked using the statewide health information exchange infrastructure. Manner of death, the standard outcome in suicide death research, was determined by the state medical examiner. Health services data were linked from multiple data sources such as electronic health records, hospital discharge data, and administrative insurance claims. Data were structured as a common format that preserves observations at their lowest level of analysis. Data features were included based on known or hypothesized psychiatric or suicide risk factors.</p><p><strong>Results: </strong>The warehouse contains a mix of records across data sources for patient diagnoses, clinical encounters, procedures, area of residence, pharmacy fills and laboratory findings. MSDW represents 104,517 decedents reported by the OCME between 2012 and 2020, 5,059 classified as suicides.</p><p><strong>Conclusions: </strong>The MSDW is a statewide data warehouse that allows users to conduct population health research, predictive modeling and observational studies for multiple outcomes. It has multiple overlapping clinical records that improve the completeness and timeliness of data. It is a high-quality statewide data warehouse for conducting suicide prediction research and assessing risk for surveillance and intervention.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag062"},"PeriodicalIF":3.4,"publicationDate":"2026-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13105295/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147783662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Aim high, stay private: differentially private synthetic data enables public release of behavioral health information with high utility.","authors":"Mohsen Ghasemizade, Juniper Lovato, Chris Danforth, Peter Sheridan Dodds, Laura S P Bloomfield, Matthew Price, Joseph Near","doi":"10.1093/jamiaopen/ooag066","DOIUrl":"https://doi.org/10.1093/jamiaopen/ooag066","url":null,"abstract":"<p><strong>Objective: </strong>Sharing behavioral health and wearable data poses privacy challenges, as traditional de-identification remains vulnerable to re-identification. Differential privacy (DP) provides mathematical guarantees through a tunable privacy budget, <math><mi>ϵ</mi></math> . This study evaluates the feasibility of generating and releasing DP synthetic behavioral health data with high analytical utility, identifying practical <math><mi>ϵ</mi></math> values for public data sharing.</p><p><strong>Materials and methods: </strong>We analyzed physiological data from wearable devices and self-reported data from Phase 1 of the Lived Experiences Measured Using Rings Study (LEMURS), which tracked sleep, stress, and well-being among first-year college students. Three DP synthetic data generators: AIM, MST, and PATECTGAN, were evaluated across privacy budgets ranging from <math><mrow><mi>ϵ</mi> <mo>=</mo> <mn>1</mn></mrow> </math> to 100. Utility was assessed using L1/L2 errors, correlation, regression, UMAP, and assessed vulnerability via privacy attacks.</p><p><strong>Results: </strong>AIM outperformed MST and PATECTGAN in preserving both statistical and analytical properties of the original data. For the Survey dataset, the lowest marginal errors occurred at <math><mrow><mi>ϵ</mi> <mo>=</mo> <mn>5</mn></mrow> </math> and 10. Correlation, regression, and UMAP analyses confirmed that AIM-generated data closely replicated original relationships at moderate <math><mi>ϵ</mi></math> values.</p><p><strong>Discussion: </strong>Choice of privacy budget is still an open question, and it is task-agnostic and dataset-specific. Moderate privacy budgets ( <math><mrow><mn>5</mn> <mo>≤</mo> <mi>ϵ</mi> <mo>≤</mo> <mn>10</mn></mrow> </math> ) maintained key associations between physiological and psychological measures while ensuring privacy. AIM's workload-aware design effectively allocated noise toward relevant features, enhancing performance.</p><p><strong>Conclusion: </strong>A privacy budget of <math><mrow><mi>ϵ</mi> <mo>=</mo> <mn>5</mn></mrow> </math> offers a practical balance between data utility and participant privacy for LEMURS behavioral health data sharing.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag066"},"PeriodicalIF":3.4,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13099409/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147783655","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Provider survey to assess the usability and acceptability of an automated electronic health record-based tool for atrial fibrillation to improve anticoagulation management.","authors":"Brittany A Carlson, Catherine P Benziger, Melissa L Harry, Nicole A Groth, Clayton I Allen, Laura A Freitag","doi":"10.1093/jamiaopen/ooag055","DOIUrl":"10.1093/jamiaopen/ooag055","url":null,"abstract":"<p><strong>Objective: </strong>Implemented a tool to identify high-risk patients with atrial fibrillation (AF) with a CHA₂DS₂-VASc score of ≥2 (males) or ≥3 (females) who are not treated with oral anticoagulation. Aimed to evaluate the acceptability and usability of the \"AF or Flutter not on Anticoagulant\" electronic health record-based Care Gap (AF Care Gap) alert and associated best practice advisory (BPA) for clinicians managing patients with AF.</p><p><strong>Materials and methods: </strong>An electronic survey was sent to 490 primary care and cardiology providers at Essentia Health (Duluth, MN, USA) to evaluate the usability, acceptability, and obtain feedback post-implementation. We excluded providers who did not complete the consent (<i>n = </i>9), give consent (<i>n = </i>15), complete the survey (<i>n = </i>340), or see AF patients (<i>n = </i>5).</p><p><strong>Results and discussion: </strong>Survey response rate was 25% (<i>N = </i>121); 51% reported prior use of the AF Care Gap (<i>N = </i>62) with the majority (73%) in family medicine. Most users and nonusers reported they were \"likely/extremely likely\" to start a conversation about anticoagulation and use the AF Care Gap or BPA in their future practice (84%). Of those who used it, 75% of providers were \"likely/extremely likely\" to prescribe anticoagulation. Most users would recommend it to others (67%). On the System Usability Scale, the AF Care Gap scored 72.5/100. The acceptance was 27/35 using a modified Theoretical Framework of Acceptability questionnaire.</p><p><strong>Conclusion: </strong>Survey respondents report above average usability and acceptability. Future evaluation of the AF Care Gap tool utilization and persistent gaps in anticoagulation management are still needed to improve management for high-risk AF patients.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag055"},"PeriodicalIF":3.4,"publicationDate":"2026-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13091096/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147724047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring the prognostic value of serum albumin in critically ill cancer patients: an observational cohort study utilizing machine learning from large intensive care unit databases.","authors":"Guiyue Wang, Limei Yuan, Zhenguo Song, Ying Shen, Jiaxu Li, Xiaobei Zhang, Yuan Li, Kaili Yu, Chengqi Deng, Minhui Yi, Kaiyuan Wang, Huiqin Mo","doi":"10.1093/jamiaopen/ooag059","DOIUrl":"10.1093/jamiaopen/ooag059","url":null,"abstract":"<p><strong>Objectives: </strong>Patients with malignant tumors are admitted to the ICU for diverse reasons. However, the clinical utility of serum albumin as a prognostic biomarker remains unclear.</p><p><strong>Materials and methods: </strong>Patients with malignant tumors were screened from the Medical Information Mart for Intensive Care IV (MIMIC-IV, v3.1). This study employed Kaplan-Meier curves, Cox proportional-hazards models, restricted cubic splines (RCS), receiver operating characteristic (ROC) curves, and subgroup analyses to evaluate serum albumin associated with all-cause mortality. For mortality-risk prediction, we applied machine-learning algorithms and used SHapley Additive exPlanations (SHAP) to identify the most influential predictors among critically ill cancer patients.</p><p><strong>Results: </strong>A total of 1,739 patients with malignancy were included. The Kaplan-Meier curves showed significantly higher all-cause mortality in the hypoalbuminemia group (serum albumin < 30 g/L) than in the control group at each time point. Multivariable Cox regression models confirmed that hypoalbuminemia was independently associated with 28-day mortality (HR 1.74; 95% CI 1.34-2.27). Serum albumin exhibited a superior predictive capacity for long-term mortality (90-day and 1-year), with AUCs of 0.676 and 0.664, respectively, notably higher than those of the SOFA score (0.617 and 0.579). External validation using data from Tianjin Cancer Hospital yielded consistent results. The Machine learning model identified BUN, serum albumin, respiratory rate, heart rate, and SOFA as the top predictors for 14- and 28- day mortality.</p><p><strong>Conclusion: </strong>Hypoalbuminemia was independently associated with increased all-cause mortality. Serum albumin measured at ICU admission serves as a prognostic biomarker for identifying high-risk cancer patient groups.</p>","PeriodicalId":36278,"journal":{"name":"JAMIA Open","volume":"9 2","pages":"ooag059"},"PeriodicalIF":3.4,"publicationDate":"2026-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13091094/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147724028","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}