Unraveling spatial distribution and associated factors of bisphenol analogues in the Yangtze River Delta: A machine learning approach.

Acceptance of water reuse adoption in small communities: analysis of driving factors and policy preferences.

How and Under What Conditions Do Patient Portals Improve Cancer Screening Completion: A Scoping Review.

Acute kidney injury (AKI) is a common and serious condition associated with prolonged hospitalization, chronic kidney disease, and increased mortality. Early prediction of AKI offers an opportunity to mitigate these adverse outcomes, yet existing models often fail to generalize to the emergency department (ED) setting, particularly for patients discharged directly to the community. We sought to develop and validate machine learning models that predict new or progressive AKI within 72 hours of ED departure, addressing challenges related to missing outcome data for discharged patients. This retrospective, multicenter study of adult patients from 5 EDs within a large health-care system included adult patients with at least 1 serum creatinine measurement during their ED visit. AKI was defined using Kidney Disease Improving Global Outcomes serum creatinine-based criteria, and prediction models relied on demographic, clinical, and laboratory data routinely collected during ED care. Extreme gradient boosting algorithms were trained using 4 approaches to handle missing outcome data: incomplete case exclusion, negative outcome assumption, multiple imputation, and inverse probability weighting. Model performance was evaluated via 10-fold cross-validation and external temporal validation using area under the receiver operating characteristic curve, precision, recall, calibration curve analyses, and measurement of diagnostic performance across a range of risk thresholds. A total of 1,124,017 ED visits between 2017 and 2024 were included in the study; 5.7% (22,093) met AKI progression outcome criteria. The models demonstrated robust predictive performance for any new or progressive AKI (area under the receiver operating characteristics curve, 0.81-0.82) and severe AKI (area under the receiver operating characteristics curve, 0.87-0.88) across validation cohorts. Inverse probability weighting provided a reliable and consistent method for handling missing outcome data, ensuring accurate risk estimates for both hospitalized and discharged patients. Models performed similarly across diverse subgroups and ED sites. Machine learning models trained on routinely collected ED data can provide reliable early predictions of AKI progression, supporting actionable clinical decision making for a broad spectrum of patients. This study advances the real-world usability of such models by expanding their applicability to discharged patients and by enabling estimation of ongoing kidney risk, irrespective of AKI status on arrival.

Vagus nerve stimulation as an adjunctive therapy for super-refractory status epilepticus including NORSE: a retrospective cohort study.

This study employs first-principles calculations based on Density Functional Theory (DFT) to investigate the electronic and structural properties of boron-carbon-nitrogen (BCN) monolayers exhibiting quasi-periodic order within the unit cell, specifically following a Fibonacci-like sequence. To quantify the degree of disorder in these quasi-periodic BCN systems, we introduce a comparative metric against fully random BCN configurations. The analysis reveals a linear correlation between the degree of disorder and the electronic band width, indicating that increasing disorder enhances electron localization, with boron and nitrogen atoms acting as dopants within the graphene-like lattice. Furthermore, using a simplified tight-binding bond model, we extend the study to large-scale systems. Our results show that both the band gap and the degree of disorder become size-independent for Fibonacci BCN monolayers containing more than 288 atoms, a behavior attributed to the intrinsic self-similarity of quasi-periodic structures. Owing to electron localization, prominent peaks associated with electronic transitions between flat valence and conduction bands emerge in the absorbance spectra of large Fibonacci BCN systems. • Quasi-periodic BCN monolayers following a Fibonacci sequence are investigated. • A Pearson correlation metric quantifies disorder relative to random BCN systems. • Disorder degree linearly correlates with electronic band width and localization. • Band gap becomes size-independent for superlattices exceeding 288 atoms. • Optical spectra reveal distinct peaks from transitions between flat bands.

Can large-scale climate patterns predict nitrate export mechanisms from agricultural land?

The need for long-term support: Five-year outcomes after severe injury in older adults.

The Brahmaputra River Basin (BRB), a large climate-sensitive transboundary river system in South Asia. This study develops a climate-driven deep learning framework to simulate and project streamflow in the basin by linking observed daily discharge with key hydro-meteorological variables. A hybrid Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) model is implemented for spatiotemporal modeling, capturing both spatial variability in precipitation and long-term temporal dependencies in streamflow dynamics. The hybrid model is compared with standalone deep learning architectures to identify the best-performing approach. Additionally, Shapley Additive Explanations (SHAP) are applied to interpret the model and quantify the contribution of each climatic variable to the predicted streamflow. The CNN–LSTM model achieved Nash–Sutcliffe Efficiency (NSE) values of 0.92 during training and 0.86 during testing, outperforming the standalone deep learning models. SHAP analysis indicates that precipitation is the dominant driver of streamflow variability, while temperature provides a secondary influence, and evaporation and snow contribute minimally. Future projections using bias-corrected CMIP6 climate scenarios (SSP2–4.5 and SSP5–8.5) suggest substantial increases in mean and extreme flows, with the 100-year return period discharge projected to rise by about 25%. These findings demonstrate that climate-driven hybrid deep learning models can provide interpretable and reliable long-term streamflow projections, supporting improved flood-risk assessment and climate adaptation planning in the basin. • CNN–LSTM predicts Brahmaputra River Basin streamflow using only climate data. • CNN–LSTM model outperformed standalone deep learning models. • Precipitation is the key driver of streamflow variability. • Future CMIP6 projections indicate intensified high flows.

Heat and hearts: An exposure-anchored computational phenotyping framework for assessing cardiovascular vulnerability during extreme heat.

Computational microbiology: Where is artificial intelligence addressing the barriers to large-scale simulations of bacterial cell envelopes?

Adult adherence to multi-dose vaccine schedules: A population-based assessment of guideline concordance.

This study evaluated donor-reported opioid use and pain management outcomes after living donor nephrectomy at a single center. Prospective data were collected through an automated text-messaging platform to assess postdischarge pain intensity, perceived pain control, and use of prescribed opioids for acute postoperative pain. This prospective study was conducted at a large academic health system in Pennsylvania. Eligible participants were adult living kidney donors (aged 18 y and older) who underwent donor nephrectomy between July 2021 and March 2023. Data were collected using an automated text-messaging system from hospital discharge through postoperative day 28. Of the eligible participants, 75 (71%) consented to participate. Donors who consented to follow-up did not differ significantly from those who did not consent in terms of demographic or perioperative characteristics. Inpatient opioid use and discharge prescribing practices were similar across groups. Of participants who responded at each follow-up, the median (interquartile range) self-reported pain intensity on day 4 was 4.0 (3.0-5.0) and decreased to 1.0 (0.0-2.0) by day 21. The median (interquartile range) perceived pain control increased to 10.0 (9.0-10.0) by day 14, with 3 (8.1%) of participants reporting opioid use by this time. From a systems perspective, these findings underscore the need to recalibrate default discharge prescribing practices. Implementing standardized order sets and feedback mechanisms may reduce excess opioid supply while maintaining living kidney donor comfort.

Progressive state transfer for BFT with larger-than-memory state

Patient portals support health management, yet enrollment remains low. This study evaluated whether timely email nudges increase patient portal enrollment compared with usual system portal invitations, whether message framing influences enrollment, and whether patients nudged to enroll go on to engage with the portal at similar rates as those who enroll organically. In this pre-registered randomized controlled trial at a large health system, 5009 patients aged ≥ 18 with recently released laboratory results were randomized over 15 days to: (1) an "Ease" email suggesting immediate access to laboratory results, upon enrollment, (2) a "Transparent" email outlining registration steps needed to access their results, or (3) no email. The primary outcome was portal enrollment within one week. Secondary outcomes included enrollment rates over four years and post-enrollment portal engagement. Post hoc analysis examined the impact of time-sensitive test results. Compared with control, receiving either nudge increased one-week portal enrollment (10.8% vs 3.9%; OR 2.99, P < .001). This difference remained significant through one year post-nudge (36.5% vs 33.1%; OR 1.23, P = .005). Message framing had no significant effect on enrollment. Among emailed patients, nudges were more effective with time-sensitive laboratory results (OR 1.46; P < .001). Among enrollees, long-term portal usage was similar across groups. Beyond standard marketing efforts, timely nudges highlighting the immediate benefit of accessing laboratory test results can meaningfully increase patient portal enrollment-especially when results are time-sensitive. These low-cost, one-time interventions can drive adoption without sacrificing long-term engagement.

To describe obesity treatments in real-world settings after discontinuation of semaglutide or tirzepatide and variability in weight change post-discontinuation. This retrospective cohort study used electronic health records from 1 January 2021, to 30 June 2025, in a large health system in Ohio and Florida. Adults with overweight or obesity who initiated injectable semaglutide or tirzepatide for obesity or T2D between 2021 and 2023 and discontinued the medication within 3-12 months were included. Main outcomes were reinitiation of the index medication or receipt of an alternative treatment, body weight changes from baseline to index medication discontinuation and from discontinuation to 1-year post discontinuation. A total of 7938 patients (mean [SD] age, 55.7 [13.4] years; 5061 [63.8%] female) were identified. During 1-year post-discontinuation, 19.6% restarted the index medication and 35.2% received an alternative obesity treatment, including starting another medication (27.4%), lifestyle modification visit (13.7%) and metabolic and bariatric surgery (0.6%). Mean percentage weight change from baseline to discontinuation was -8.4% [95% CI, -8.7%, -8.1%] when treating obesity and -4.4% [95% CI, -4.7%, -4.2%] when treating T2D. Mean percentage weight change from discontinuation to 1 year later was 0.5% [95% CI, 0.0%, 1.0%] when treating obesity and -1.3% [95% CI, -1.6%, -1.0%] when treating T2D; however, there was considerable individual-level variability. In this large sample of patients who discontinued semaglutide or tirzepatide, reinitiation of the original medication or receipt of alternative obesity treatment was common. The average weight change 1-year post-discontinuation was relatively small; however, there was considerable individual-level variability.

Exploring key risk factors for loss to follow-up after hospitalization for acute stroke.

From tracking to thinking: Facilitating post-exercise reflection by a large language model-mediated journaling system

Venous thromboembolism (VTE) is a leading cause of preventable perioperative mortality in otolaryngology, highlighting the urgent need for validated risk assessments such as the Caprini score to guide prophylaxis and enhance patient safety. The objective of this study was to determine VTE outcomes following otolaryngology surgeries, stratified by Caprini scores, otolaryngological subspecialties, and socioeconomic conditions represented as Area Deprivation Index (ADI). This study was conducted via a multi-year retrospective review of 9036 otolaryngology procedures performed at multiple tertiary care hospitals within a large health system from June 2016 through December 2022. Logistic regression and multivariable models were used to investigate the relationship between VTE and Caprini score, subspecialty, and ADI. A total of 9036 otolaryngology surgeries were analyzed. Of those, 174 (1.9%) patients experienced a VTE and 33 (0.4%) of those patients experienced a VTE within 2 weeks after otolaryngological surgery (VTE Subset). A Caprini score of ≥ 6 resulted in a VTE (VTE: OR Unadjusted = 1.49, OR Adjusted = 1.47, p < 0.001; VTE Subset: OR Unadjusted = 1.43, OR Adjusted = 1.43, p < 0.001). In those who experienced a VTE, a Caprini score of ≥ 6 yielded a sensitivity of 0.672, specificity of 0.809, positive predictive value (PPV) of 0.065, and negative predictive value (NPV) of 0.992. In patients who experienced a VTE following otolaryngology surgery, a Caprini score of ≥ 6 showed a sensitivity of 0.727, specificity of 0.802, PPV of 0.013, and NPV of 0.999. Caprini guidelines can function as a robust tool for ruling out VTE risk in otolaryngology surgical patients regardless of subspecialty and ADI. III.

The COVID-19 pandemic highlighted pediatric health care disparities and disrupted routine care, including social needs assessments. The American Academy of Pediatrics recommends universal screening for Social Determinants of Health (SDOH), yet implementation remains inconsistent in primary care settings. This quality improvement (QI) project aimed to implement a standardized, sustainable SDOH screening and referral process in pediatric primary care, hypothesizing that structured interventions would improve screening rates. This QI initiative was conducted from January to September 2023 across six practices within a large pediatric health system. Eligible patients (ages 0-19) included those attending their first well visit of the calendar year. The SMART aim targeted a 50% increase in SDOH screening compliance, from 28% at baseline to 42% over 9 months. Using the Consolidated Framework for Implementation Research and two Plan-Do-Study-Act cycles, the team addressed key implementation barriers and refined interventions. The primary measure was screening completion rate; the balancing measure was the number of refusals to screen. SDOH screening rates increased from 28% to 55%, with eligible patient volumes ranging from 2400 to 5500. All six practices demonstrated statistically significant improvements (P < 0.001). Positive screens ranged from 3.3% to 8% of patients screened. Screening refusals increased significantly (P < 0.001). Standardized SDOH screening, implemented through structured QI methods and stakeholder engagement, significantly improved screening rates in pediatric primary care. Future studies should assess referral effectiveness, clinical outcomes, cost-effectiveness, and strategies to mitigate patient discomfort and systemic barriers.