Fragility fractures occur with low-energy trauma in patients with weakened bone strength and are associated with reductions in quality of life, function, and independence. The risk of re-fracture after a fragility fracture is significantly elevated. Prevention of secondary fractures is imperative as the population ages, given the morbidity, mortality, and costs directly attributable to these fractures. Unfortunately, treatment rates remain low despite the availability of effective therapies to reduce the risk of secondary fracture. We assessed the rate of pharmacologic treatment after vertebral, femur, or pelvic fracture within our academic medical center. In addition, we investigated potential factors, such as race or endocrinology consultation, which might influence post-fracture treatment. Our sample included 814 females over age 65, who sustained a hip, pelvis, or vertebral fracture between January 2018 and December 2021. Within the sample, 29.7% received an anti-fracture prescription within 6months of fracture. Those with osteoporosis included on their electronic health record (EHR) problem list had a significantly higher percent probability (28.6, 95% confidence interval (CI), 22.9, 34.3) of anti-fracture therapy prescription compared to those without osteoporosis on their problem list. Furthermore, Black patients were less likely to have a diagnosis of osteoporosis listed (p < 0.0001) and had a 10.9 percentage point lower probability of receiving a prescription compared to White patients (CI, 1.9, 19.9). These results suggest that the majority of patients in our health care system did not receive effective pharmacotherapy for secondary fracture prevention. The study highlights the need for more effective interventions to reduce secondary fractures.

In large-scale clinical data analysis, CSV and traditional relational database management system-based approaches are widely used but impose substantial storage and processing constraints that delay research preparation and hinder multicenter collaboration. Although column-oriented storage formats such as Apache Parquet have gained attention in data science, systematic end-to-end evaluations in clinical environments remain limited, particularly regarding efficiency and scalability. This study aimed to empirically evaluate whether a Parquet-based end-to-end pipeline could improve computational efficiency and scalability in large-scale clinical data analysis while preserving predictive performance and protecting privacy. Electronic health record data comprising 13.76 million rows from a large academic medical center in Korea were analyzed using Parquet, CSV, PostgreSQL, and DuckDB environments. Standardized SQL workloads and multilabel classification models-implemented using graphics processing unit-accelerated Extreme Gradient Boosting and classifier chain (CC) ensembles to address class imbalance-were applied to evaluate storage efficiency, time to analysis, and predictive performance. Statistical equivalence testing with prespecified clinical margins and bootstrap resampling ensured rigorous comparison, while privacy risks were assessed through advanced membership inference attacks (MIA), including shadow MIA and likelihood ratio attacks. Compared with CSV, Parquet demonstrated enhanced computational efficiency by lowering disk access from 940.2 to 44.2 seconds (95.3% reduction). End-to-end processing latency was substantially reduced across feature transformation (15.0 vs 9.3 s) and model training (8.1 vs 6.7 s). To address complex clinical correlations, we implemented CC and one-vs-rest architectures, which effectively captured interdependencies between disease labels. Classification performance remained statistically equivalent across area under the receiver operating characteristic curve, area under the precision-recall curve, accuracy, and F1-score, with all differences falling within prespecified clinical equivalence margins (P<.001). Notably, the CC ensemble demonstrated high technical rigor, minimizing Hamming loss (2.2×10-4) and ensuring robustness even in imbalanced cohorts. MIA performed at chance level (area under the curve=0.500), suggesting no measurable increase in privacy risk. By significantly mitigating data processing bottlenecks, a Parquet-based pipeline enabled high-throughput, large-scale clinical evidence generation without compromising model integrity or patient privacy. This framework provides a scalable and robust infrastructure for precision medicine, facilitating agile multicenter collaborations and real-world data analysis in resource-constrained clinical environments.

Chronic spontaneous urticaria (CSU), angioedema (AE), and atopic dermatitis (AD) are some of the common recurrent dermatological conditions that seriously affect the quality of life in patients through wheals, itching, and swelling unpredictably. Patient-reported outcome measures (PROMs) are standardized tools used to assess disease activity and the quality of life from the patient’s perspective. However, these instruments are usually underutilized in clinical settings due to various reasons such as time constraints and logistical challenges. The objective of the study is to assess and describe mobile applications that incorporate PROMs for skin allergy monitoring and management, focusing on features, usability, and clinical utility. A qualitative descriptive analysis of five selected mobile apps that integrated validated PROMs for CSU, AE, or AD, which were publicly available at major app stores, with supporting documentation or published literature, was performed. The mobile app rating scale (MARS) framework was used to appraise the apps, considering ratings found in app stores, user feedback, and literature reviews. Among the five applications evaluated, namely Target My Hives, UrCare, SymTrac, HAE TrackR, and CRUSE, there was considerable variability in how PROMs were integrated into features that support symptom tracking and facilitate interaction between patients and clinicians. Key aspects are summarized in a comparative table, followed by a narrative analysis highlighting strengths in engagement and limitations in automation. The review is based on secondary data, with no direct user testing; gaps in app design, regulatory compliance, and electronic health record integration were noted. Mobile applications with PROMs hold promise for improving care in skin allergy but need further clinical validation, advances in accessibility, and interoperability.

Black individuals bear a disproportionate burden of kidney diseases, including genetically mediated risk related to Apolipoprotein L1 (APOL1) gene variants. Awareness of APOL1-mediated kidney disease (AMKD) and participation in therapeutic trials remain low. Whether different engagement strategies can raise awareness and identify trial-eligible individuals is uncertain. The Community APOL1 Research Engagement (CARE) study aimed to increase AMKD awareness through culturally tailored education and screening while building a clinical trial-eligible registry, CARE registry, to support a phase 2 baricitinib trial for AMKD in the Janus kinase-STAT Inhibition to Reduce APOL1-Associated Kidney Disease (JUSTICE). The CARE study was conducted from May 2022 through July 2025 across community and clinical settings in multiple U.S. regions, in partnership with churches with predominantly Black attendees. Black adults aged 18-70 years without diabetes or dialysis dependence underwent APOL1 genotyping and kidney disease screening. Recruitment occurred via community events, electronic health record (EHR) queries, physician referrals, and self-referrals. The primary outcome was eligibility for the CARE Registry, defined by APOL1 high-risk genotype, urine-albumin-to-creatinine ratio (UACR) ≥300 mg/g and eGFR ≥25 mL/min/1.73 m2. Of 1,052 individuals approached, 789 (75%) consented to screening. Overall, 128 (17%) carried APOL1 high-risk genotypes. Community events accounted for most enrollments (83%) but yielded low rates of registry-eligible albuminuria (1%). In contrast, EHR queries and physician referrals identified higher proportions of participants with APOL1 high-risk genotypes and UACR ≥300 mg/g. Twenty-four participants met CARE Registry criteria, and seven enrolled in the JUSTICE trial. Refusal was 4% and attrition was 2%. Community engagement achieved high participation and awareness but was less efficient for identifying trial-eligible individuals than EHR- and provider-based approaches.

Cardiovascular disease (CVD) remains the leading cause of premature death in England, with ethnic minority populations disproportionately affected, largely due to differences in socioeconomic factors, exposure and/or susceptibility to CVD risk factors. Midlife risk assessment does not fully account for observed variation in CVD incidence and mortality. Early and precise quantification of risk factor burden across diverse populations is therefore essential to inform targeted prevention strategies. This study assessed the prevalence of CVD risk factors in apparently healthy individuals residing in London. This cross-sectional study included CVD-free individuals aged 30-90years residing in London and registered with general practices using the Egton Medical Information Systems (EMIS) electronic health record system. Unadjusted, crude estimates of traditional CVD risk factors were assessed across participants of different ethnicities who underwent a CVD risk assessment between 2009-2020. Among 607,327 registered individuals, 83,414 were included (52.0% women, median age 45 [IQR:36-48] years). Ethnic distribution was as follows: White (43.6%), Asian (30.1%), Black (9.7%), Chinese/Other (4.0%), Mixed (2.1%). Overall, 7.8% were current smokers, 31.5% had obesity (universally defined as body mass index (BMI) ≥ 30.0kg/m2), 48.5% had elevated blood pressure (BP ≥ 140/90mmHg), 44.9% had hypercholesterolemia (≥ 5.0mmol/l), 28.2% had elevated triglycerides (TG) ≥ 1.7mmol/l, and 25.9% had low high-density lipoprotein cholesterol (HDL-C < 1.0/1.3mmol/l for males/females, respectively). Smoking prevalence was highest among White individuals (9.7%). Obesity prevalence varied across groups, with higher proportions in Black participants (42.3%) and lower in Asian individuals (26.1%). Elevated BP was recorded more frequently in Mixed (54.9%) and Black (53.0%) participants and less frequently in those classified as Chinese/Other (42.7%). Total cholesterol ≥ 5.0mmol/L was more commonly documented in Mixed (56.8%) and White (49.8%) participants. Higher proportions of Asian individuals had elevated TG (30.9%) and low HDL-C (31.6%), while corresponding proportions were lower among Black participants (14.4% and 19.5%, respectively). This large-scale analysis of a diverse population suggests variation in CVD risk factor burden among relatively young individuals without CVD. While not implying causality, these findings reflect inequalities between ethnic groups and support an appraisal of early, tailored, and equitable public health policies to improve CVD risk management across diverse populations.

Addressing disparities and improving outcomes for patients in a neonatal intensive care unit (NICU) requires accurate and reliable reporting of race and ethnicity. We sought to increase the percentage of infants with caregiver-reported race and ethnicity at NICU discharge to greater than 90% using quality improvement (QI) methods. We first identified drivers including standardization, increasing reliability, and addressing gaps in education and health literacy. Tests of change included revising forms to enhance clarity and support equity, addressing barriers to the reliable processing of paper forms, real-time auditing and clinical decision support in the electronic health record, and development of a patient portal prebirth electronic form. Statistical process control charts were used to track the outcome measure, the percentage of patients in a NICU who had caregiver-reported race and ethnicity documented at discharge, and the process measure of the percentage of infants whose forms were received within 7 days of admission. The average percentage of patients with caregiver-reported race and ethnicity documented at NICU discharge increased from 38% to 94%. The percentage of forms received within 7days of admission increased from 81% to 96%. Improvements have been sustained for more than 18months. We improved the documentation of caregiver-reported race and ethnicity among patients discharged from our NICU through rigorous QI methods. Accurate and reliable capture of race and ethnicity data should help identify and address disparities and improve care for all infants.

Objective: To develop a technical product based on the analysis of electronic health record data and population registry reports, using public health system care criteria, to enable planning of the weekly work schedule for doctors and nurses. The aim is to reorganize the professional agenda to better balance spontaneous and scheduled demand. Methods: A cross-sectional study conducted in primary care in a municipality in Minas Gerais, Brazil, with three Family Health units covering the entire population. Data from the Azul Family Health Strategy were analyzed. Variables such as age group, sex, shift, type of service, clinical conduct, and diagnosis were assessed using descriptive statistics. Based on the demand profile and care parameters, a standard weekly schedule was created to reorganize professionals’ agendas, incorporating both spontaneous and scheduled demand. Results: An imbalance was found between spontaneous and scheduled demand, with a predominance of unscheduled visits. The proposed solution reorganized the professionals’ weekly agenda, allocating time for scheduled care, health education, administrative activities, and planning, balancing care parameters with local needs. Conclusions: Reorganizing the work schedule can enhance care quality, increase service resolution, and strengthen primary care principles by valuing health promotion and allowing time for planning and administrative tasks, ultimately improving service organization and user access.

Background: Glomerular disease (GD) and diabetic nephropathy are both leading causes of end-stage kidney disease (ESKD) in the United States. Much is known about each individually, but less of any interactions between the two. There is emerging evidence that factors specific to glomerular disease, such as immunosuppression, may increase the risk of diabetes, which in turn could compound glomerular filtration rate (GFR) decline through the mechanisms of diabetic nephropathy. Understanding the epidemiology of prediabetes and diabetes in glomerular disease patients may inform improved screening and prevention practices in this population and may lead to strategies that mitigate progression to ESKD. The aim of this study is to delineate risk factors for prediabetes in glomerular disease. Methods: Data was extracted from University of Michigan and Kidney Research Network electronic health record registry with patients classified by age at glomerular disease at diagnosis or first nephrology appointment (child (age&lt;18y, n=406) and adult (≥ 18y, n=339)). A Cox proportional hazards model was calculated using prediabetes after kidney disease onset as the outcome, adjusted for age, sex, race, weight, hypertension, and defined relevant drug prescriptions. A subgroup analysis was performed to track the progression from prediabetes to diabetes. Results: 148 patients (19.9% of cohort) developed prediabetes in study follow-up. Adult GD patients were more likely than pediatric GD patients to progress (HR: 1.73 [95%CI: 1.19 - 2.50]), as were patients with uncontrolled hypertension (HR: 9.61 [95%CI: 3.02 - 30.61]) and controlled hypertension (HR: 6.50 [95%CI: 1.91 - 22.18]). The use of beta blockers, statins, or diuretics was also associated with higher prediabetes risk (HR: 2.87 [95%CI: 1.98 - 4.17]). Conclusions: Adult age, worsening control of hypertension, and certain medications were associated with increased prediabetes risk in pre-existing glomerular disease. More data, in particular prospective data, is needed to refine risk relationships and incidence data.

Asthma is a chronic respiratory condition shaped by environmental, physiological, and behavioral factors. Accurate prediction of asthma severity is vital for personalized care and reducing exacerbations. While Machine Learning (ML) has been widely explored in asthma prediction, many existing models lack generalizability, robustness, or comprehensive integration of multiple predictors, limiting their clinical applicability. This study presents a robust ML-based model for classifying asthma severity by incorporating diverse patient and environmental features. A supervised learning approach was employed using a publicly available dataset of 1,010 records with 14 features, including demographics, clinical symptoms, and environmental indicators (temperature, wind speed, and humidity). The dataset was pre-processed and stratified to balance severity classes. Three models—Decision Tree, Support Vector Machine (SVM), and Random Forest (RF) were evaluated using standard metrics: accuracy, precision, recall, F1-score, and ROC AUC. Among them, the RF model showed superior performance, achieving 96.70% accuracy, a 0.9668 F1-score, and a 0.9831 ROC AUC. Feature importance analysis highlighted environmental factors, particularly temperature and humidity, as key predictors of asthma severity. These results underscore RF's effectiveness in providing accurate, interpretable predictions and addressing limitations of earlier models. The proposed model offers a data-driven framework for real-time severity forecasting, supporting early interventions and personalized treatment. It holds promise for integration into clinical decision support systems, thereby enhancing asthma management and optimizing healthcare resource use. It is therefore recommended that the proposed model be operationalized within clinical triage protocols, embedded into electronic health record (EHR) infrastructures, or integrated into mobile health (mHealth) applications to facilitate data-driven, proactive asthma management across diverse care settings in Nigeria.

Youth use of GLP-1RAs is increasing. This study described GLP-1RA prescription patterns and barriers to treatment at a pediatric integrated weight management clinic. This retrospective cohort included youth 12 to 17years with BMI at least 95th% for age and sex, with at least 1 visit at an integrated weight management program from January 2023 to August 2025. We identified youth with at least 1 GLP-1RA prescription in the electronic health record. We assessed demographic factors (age, sex, race, ethnicity, insurance payer, preferred language) and health factors (BMI, Type 2 diabetes, results of ALT, cholesterol, and hemoglobin A1c testing). Logistic regression assessed for an association between GLP-1RA prescription and demographic and health factors. Manual medical record review of a subsample of 102 youth with GLP-1RA prescriptions described reasons for interruptions in use. Of 1647 youth, 325 (20%) had at least 1 GLP-1RA prescription. Odds of prescription increased with increasing age, increasing BMI, abnormal laboratory testing results, and non-Hispanic white or Hispanic race and ethnicity (compared with non-Hispanic Black). Odds of a prescription decreased with a preferred language other than English. In medical record review, 65 youth (64%) experienced GLP-1RA treatment interruptions, most commonly related to cost and insurance coverage. At one institution's integrated weight management program, 20% of potentially eligible youth were prescribed GLP-1RAs. Prescriptions were more likely for older patients and those with comorbid conditions, and less likely for Black or non-English speaking patients, reflecting known pediatric health disparities. Barriers to treatment were common after the prescription.

Abstract Background Electronic health record (EHR) data, a key form of routinely collected patient data, offer great potential for medical research and the development of artificial intelligence (AI) tools. However, because these data are primarily gathered for health care rather than research, it often lacks the quality needed for AI training, raising both methodological and ethical concerns. While previous studies have reviewed the ethical implications of both routinely collected patient data and AI separately, their intersection, where AI is applied to such data, remains largely unexplored. Objective This study aimed to examine the ethical challenges that arise at the intersection of EHR data and AI development and to derive practice-oriented recommendations using the Dutch LEAPfROG (Leveraging Real-World Data to Optimize Pharmacotherapy Outcomes in Multimorbid Patients Using Machine Learning and Knowledge Representation Methods) project as a guiding case. Methods We used a mixed methods design combining a scoping literature review with a systematic search and 2 stakeholder workshops structured by the guidance ethics approach, reflecting a staged and iterative process aligned with the LEAPfROG project’s development phases. The review identified 25 relevant publications from 2014 to 2024. The workshops, conducted with 17 and 13 participants respectively, included patients, clinicians, ethicists, data officers, and AI developers. Both workshops used dialogue to identify ethical values, impacts, and action points, focusing on a case study of drug-induced acute kidney injury. Results The analysis highlighted four major themes: (1) data privacy, transparency, and consent, including challenges of meaningful consent and risks of reidentification; (2) public trust and regulatory challenges, such as fragmented oversight and inconsistent governance; (3) fair representation and model generalizability, where incomplete or biased data may perpetuate health inequities; and (4) responsible AI integration in clinical practice, including concerns about clinical tropism, administrative burden, and the risk of overreliance on AI outputs. Both literature and stakeholder perspectives underscore the risk of decontextualization when EHR data are reused and emphasize the importance of clearly defining the purpose of data reuse to ensure real-world applicability and foster trust. Conclusions Responsible AI development requires explicit attention to how EHR data are produced, interpreted, and governed in practice, recognizing that data quality and meaning are shaped by the clinical, institutional, and social contexts in which they originate. Technical solutions or top-down regulation alone are insufficient. Instead, stakeholder-led and context-sensitive approaches are needed to define the purposes, risks, and benefits of medical AI. Grounded in the realities of health care practice and in the perspectives of patients, clinicians, and data custodians, these approaches can strengthen transparency, fairness, and clinical relevance, ensuring that EHR data are used ethically and effectively to support equitable and trustworthy AI innovation.

This study compared real-world overall survival and the risk of physical comorbidities and mental health conditions among patients aged &lt;65 years with versus without inherited retinal diseases (IRDs) in the United States (US). Optum® Electronic Health Record data (January 2014–January 2023) were evaluated for IRD (patients with ≥2 medical visits with an IRD diagnosis; index date: second such medical visit) and non-IRD (patients without an IRD diagnosis; index date: random medical visit) cohorts. Baseline demographics were balanced between cohorts using propensity score matching (2:1). Outcome measures were overall survival (date of death due to any cause) and presence of physical comorbidities and mental health conditions (medical visit with a corresponding diagnosis code). In total, 4594 patients with IRD were matched to 9188 patients without IRD (mean age: 38.7 vs. 38.2 years, 53.9% vs. 55.1% female, mean follow-up: 53.1 vs. 52.8 months). Over 84 months, patients with versus without IRD had a 24% higher risk of death (overall survival: 95.8% vs. 96.7%; hazard ratio: 1.24; 95% confidence interval: 1.00–1.53; p = 0.046) and were at significantly higher risk for each evaluated physical comorbidity and mental health condition (all p &lt; 0.05). The development of novel therapies is thus needed to address the clinical burden of IRD.

Gastroschisis is the most common newborn abdominal wall defect. Gastroschisis classification is based on the absence (simple gastroschisis [SG]) or presence (complex gastroschisis [CG]) of bowel morbidity. The severity of critical organ dysfunction with gastroschisis is unknown.This was a multicenter, retrospective cohort study of infants with gastroschisis (birth weight ≥ 1.8 kg and gestational age ≥ 35 weeks) admitted to the University of Florida Health NICU between January 1, 2012, and April 1, 2023, and the Johns Hopkins NICU between July 1, 2016, and December 31, 2024. All data was collected from the electronic health record. CG was defined as the presence of atresia, necrosis, perforation, volvulus, jejunostomy, resection, or short bowel syndrome. Hourly organ dysfunction was quantified by the neonatal sequential organ failure assessment (nSOFA) score (measures respiratory, cardiovascular, and hematologic dysfunction with a range from 0 to 15 [severe]).We identified 120 patients with gastroschisis (49% male; 90 with SG). Compared with patients with SG, neonates with CG had greater maximum nSOFA scores (median: 2 [IQR]: [0, 4] vs. 3 [1, 7]; p = 0.02). The coefficient of variation on cumulative nSOFA scores calculated at 24-hour intervals after birth as a measure of organ dysfunction heterogeneity for SG patients was 278 to 332% and was 216 to 266% for CG patients.This is the first high-granularity quantification of critical organ dysfunction in gastroschisis patients. We found a low overall severity of critical organ dysfunction among all patients. Substantial heterogeneity was present in both groups. The nSOFA may help to identify a subset of patients with critical organ dysfunction outside of bowel morbidity. · Gastroschisis is the most common abdominal wall defect; the severity of organ dysfunction is unclear.. · nSOFA measures critical organ dysfunction; its role in identifying high-risk gastroschisis is unknown.. · Most infants had minimal organ dysfunction; half of SG and one-third of CG had no organ failure.. · Organ dysfunction varied widely within groups; nSOFA may improve risk detection and trial design..

Positive airway pressure therapy and cardiovascular events in obstructive sleep apnoea: an observational clinical cohort study.

Effectiveness of monovalent vaccines in preventing death and severe disease among Omicron-infected and hospitalized patients in China.

The impact of wait times on treatment engagement and outcomes in military-connected mental health clinics.

Health Information Systems and Electronic Health Record Optimization: Strategies to Enhance Electronic Health Record Usability and Clinician Workflow.

A survey on electronic health record driven multimodal representation learning

Evaluating patient acceptability of clinical pharmacist engagement following clinical decision support.

Prevalence of Fecal Incontinence in Older Adults: A Systematic Review and Meta-Analysis.