Clinical Chemistry

BackgroundSepsis is a major public health challenge, and reliable biomarkers are essential for distinguishing sepsis from other conditions. Neutrophil Gelatinase-Associated Lipocalin (Neutrophil gelatinase-associated lipocalin (NGAL)) has shown promise as a diagnostic marker due to its role in the immune response. This study evaluates plasma NGAL as a diagnostic tool at the time of ICU admission. MethodsWe analysed plasma NGAL and C-reactive protein (CRP) levels in 4732 adult patients admitted to four ICUs between 2015 and 2018. All patients were retrospectively screened for Sepsis-3 criteria at ICU admission. The discriminative performance of NGAL and CRP for sepsis was assessed using receiver operating characteristic (ROC) analysis, with NGAL levels adjusted for Chronic kidney disease (CKD) and age. Patients were stratified by renal function. ResultsPlasma NGAL levels were significantly higher in septic patients (p<0.001). For the whole cohort, NGAL alone yielded an Area under the curve (AUC) of 0.67 (Confidence interval (CI) 0.66-0.69), CRP yielded an AUC of 0.72 (CI 0.71-0.73, p<0.001), and combining NGAL with CRP nominally improved discriminative performance (AUC 0.74 vs 0.72, p<0.001). Stratified analyses indicated that NGAL, together with CRP, significantly outperformed CRP alone in patients with no kidney injury and those with Acute Kidney Injury (AKI) only. In contrast, differences were not significant in patients with CKD only or CKD and AKI. ConclusionIn this large cohort, NGAL showed modest discrimination for sepsis, with a nominal improvement when combined with CRP. These findings do not indicate that NGAL meaningfully improves sepsis diagnosis in the ICU.

Blood-based biomarkers have emerged as a promising tool for the detection and monitoring of neurodegenerative diseases such as Alzheimers disease (AD), yet broad implementation of ultrasensitive protein quantification remains constrained by reliance on specialized instrumentation and centralized laboratory infrastructure. Here we present SPLASH (Solid Phase Ligation Assay with Single wasH), an ultrasensitive proximity ligation assay platform that achieves sub-pg/mL sensitivity using only standard benchtop qPCR equipment. We developed five assays targeting Alzheimers disease biomarkers - pTau-217, A{beta}1-40, A{beta}1-42, neurofilament light chain (NfL), and glial fibrillary acidic protein (GFAP) - with limits of detection ranging from 0.0005 to 0.119 pg/mL. Direct comparison with Simoa demonstrated high concordance (R2 = 0.95) for plasma pTau-217 quantification across AD-positive and AD-negative samples. We further established compatibility with dried plasma spot samples, enabling decentralized collection and quantitation without cold-chain storage. A multiplexed five-analyte panel was applied to 69 plasma samples, revealing heterogeneous biomarker profiles consistent with AD-associated patterns. By eliminating dependencies on proprietary instrumentation, SPLASH facilitates broad implementation of ultrasensitive protein quantification for neurodegenerative disease research and diagnostics.

BackgroundCandidemia is a rare but life-threatening bloodstream infection that remains difficult to predict using conventional risk stratification approaches, highlighting the need for improved predictive strategies. As a result, empiric antifungal therapy is often delayed even in high-risk patients. MethodsWe developed a deep learning model (PyTorch_EHR) to predict 7-day candidemia risk by using electronic health record data from two large cohorts (Houston Methodist Hospital System [HMHS] and MIMIC-IV), including adult inpatients who underwent at least one blood culture. Model performance was compared with logistic regression (LR), LightGBM, and established intensive care unit candidemia scores. We further implemented a two-step prediction framework integrating candidemia and 30-day mortality risk models to inform empiric antifungal decision-making. ResultsAmong 213,404 and 107,507 patients in the HMHS and MIMIC-IV cohorts, candidemia occurred in fewer than 1% (851 [0.4%] and 634 [0.6%], respectively). PyTorch_EHR outperformed LR, LightGBM, and existing candidemia scores, particularly in terms of area under the precision-recall curve (AUPRC) in HMHS and MIMIC-IV. By integrating 30-day mortality risk, the two-step framework identified an additional 20 and 28 candidemia cases beyond the one-step model, increasing coverage to 61% (121/199) and 46% (68/147) in HMHS and MIMIC-IV, respectively. Many patients identified by the two-step framework had high mortality yet did not receive empiric antifungal therapy (61.1% HMHS; 82.6% MIMIC-IV). ConclusionA two-step deep-learning framework integrating candidemia and mortality risk may support early identification of high-risk patients and facilitate timely empiric antifungal therapy. Prospective studies are warranted to confirm the findings.

BackgroundThere is a need for synthetic peptide-based serologic assays that exploit avidity to replace whole antigens while enabling low-cost diagnostics in resource-limited settings. ObjectiveTo evaluate the diagnostic accuracy of a polymeric peptide-based ELISA leveraging avidity to enhance signal. MethodA 15-member SARS-CoV-2 peptide library corresponding to multiple epitope clusters and proteins was screened by indirect ELISA using pooled sera from RT-PCR-confirmed COVID-19 patients to identify peptides with possible diagnostic utility. The identified lead candidate, S559, possessed terminal cysteine-substitution to allow disulfide polymerization, and the resulting avidity gain was evaluated by comparing the apparent dissociation constant (KDapp) before and after depolymerization with N-acetylcysteine. The performance of an optimized ELISA using S559 was evaluated on 1,222 prospectively collected COVID-19 serum samples and 218 biobanked pre-COVID control serum samples. ResultsPolymeric S559 with a KDapp of 29.26 nM-1was demonstrated to have a 218% avidity gain relative to the completely depolymerized form. At pre-defined thresholds, the optimized S559 ELISA has a sensitivity and specificity of 83.39% (95%CI: 81.18% and 85.43%) and 96.79% (95%CI: 93.50% and 98.70%), respectively. At post hoc thresholds determined by Youden index, sensitivity and specificity reached 95.01 (95% CI: 93.63% - 96.16%) and 100.00% (95% CI: 98.32% - 100.00%), respectively. ConclusionHomomultivalent epitope presentation using polymeric S559 allows a highly specific immunoassay using human sera that may have important value in detecting antibodies, whether for diagnosing infection, confirming vaccination status or conducting surveillance.

BackgroundRecent global outbreaks of Mpox have posed significant diagnostic challenges, particularly in resource-limited settings. Conventional diagnostic methods are often inaccessible due to cost, logistical constraints, or lack of trained personnel. These limitations highlight the urgent need for alternative, scalable diagnostic strategies. This study explored the application of machine learning (ML) classifiers trained on clinical symptom data as a rapid, cost-effective tool for Mpox detection. MethodsAn open-access dataset of clinical symptoms from suspected Mpox cases was used to train and evaluate five supervised ML algorithms: Extra Trees, Quadratic Discriminant Analysis (QDA), Decision Trees, Perceptron, and Support Vector Classifier (SVC). Prior to training, data preprocessing steps, including normalization and handling of missing values, were performed after which model training was carried out using a stratified 80:20 train-test split. Performance was assessed using accuracy, recall, area under the receiver operating characteristic curve (ROC-AUC), and F1-score metrics. Subsequently, feature importance was analyzed using permutation-based techniques to determine the contribution of each clinical symptom to model predictions. ResultsAmong the five evaluated models, SVC, QDA, and Perceptron achieved superior and identical performance metrics, with accuracy, ROC-AUC, and F1-score values of 97.7%, and a recall of 95.5%. Each of these models correctly identified 44 true positive cases with zero false positives. In addition, QDA and SVC produced the lowest number of false negatives (2) and the highest number of true negatives (42), indicating robust discriminatory power. Feature importance analysis identified skin rash as the most predictive clinical feature, with a permutation importance score of 0.12. ConclusionsThese findings demonstrate the strong potential of machine learning classifiers for detecting Mpox based on clinical features. Incorporating these models into healthcare systems could significantly enhance early case detection, improve clinical decision-making, and bolster disease surveillance. Future research should focus on prospective validation of these ML classifiers in real-world clinical environments.

RationaleAutonomic dysfunction is a hallmark of sepsis pathophysiology, yet its quantification remains challenging. Multiscale entropy (MSE) derived from heart rate variability (HRV) offers a dynamic measure of physiological complexity and may serve as a biomarker of early deterioration associated with subsequent organ failure, vasopressor escalation, or mortality. ObjectiveTo determine whether MSE computed across multiple temporal scales during the first 24 hours of Intensive Care Unit (ICU) admission is associated with short-term mortality and longer-term organ dysfunction in patients with sepsis, and whether these relationships vary across vasopressor exposure. Unlike prior studies that focused on short-term HRV metrics, we applied MSE across multiple temporal scales and incorporated these features into machine learning models to evaluate their prognostic utility in septic shock. MethodsThis retrospective cohort study included adult ICU sepsis patients at Emory University Hospital from January 2016 to December 2019. Of 2,076 eligible patients, 958 were propensity matched into two cohorts: fluids-only and fluids-plus-vasopressor, with norepinephrine as the primary vasopressor. High-resolution electrocardiogram (ECG) waveforms were analyzed to compute MSE across 20 temporal scales. Machine learning models using (1) MSE features alone and (2) MSE combined with demographic and vital sign data (MSE-DV) were compared against traditional HRV measures based model and severity of illness scores for predicting outcomes. Model performance was assessed using the area under the receiver operating characteristic curve (AUROC), with a primary outcome of mortality at day 7 and secondary outcome of persistent organ dysfunction at day 28. ResultsIn the fluids-plus-vasopressor cohort, MSE-based models demonstrated superior predictive performance for 7-day mortality (AUROC 0.84) compared to severity of illness scores (AUROC 0.64). MSE-DV models also predicted organ dysfunction including 28-day renal (AUROC 0.75), neurological (AUROC 0.79), and respiratory (AUROC 0.71) dysfunction. Patients receiving second-line and third-line vasopressors and corticosteroids exhibited progressively lower MSE values, particularly at mid-range and long-range scales. ConclusionMSE features in the first 24 hours of ICU stay predict mortality and organ dysfunction with higher discrimination than traditional severity of illness scores. Future work should validate these findings, assess longitudinal MSE trends, and race-specific autonomic patterns to refine predictive models.

Serological surveillance is fundamental to infectious disease research and informed public-health decision making. Immunoassays used in the study of pathogen-specific immunity have historically relied on the collection of venous blood. While critical for many public-health applications, this sample collection method is invasive and resource intensive. The costs and logistical barriers associated with venous blood collection are exacerbated in resource-limited regions, and the shift to less invasive sampling methods would increase sample availability for pathogen surveillance and study of pathogen-specific immunity. To this end, we have developed and optimized a self-collected, saliva-based immunoassay capable of quantifying pathogen-specific antibody binding in saliva samples. Using samples collected from geographically and epidemiologically diverse regions of the world, we compared antigen-specific IgG levels in paired plasma and saliva samples. We observed that levels of IgG against multiple pathogens of public health concern - including SARS-CoV-2 and dengue virus (DENV) - were highly correlated in plasma and swab-collected saliva. In addition, the decay of maternally derived antibodies in saliva samples collected from infants was readily observed using this immunoassay, demonstrating the assay's sensitivity and potential for use in measuring antibody kinetics. We posit that this assay represents a climate stable, non-invasive tool that can aid in the surveillance and study of pathogen-specific immunity across a broad range of public-health indications.

Genomic surveillance of influenza viruses informs vaccine strain selection and evolutionary forecasting. Sequencing efforts vary widely across U.S. states, which raises concerns about spatial sampling bias. We evaluated how well 10,958 influenza virus genomes sampled by our group in Michigan captured the genetic diversity in 34,743 genomes circulating nationally from the 2021/22 through 2024/25 seasons. We defined seasonal hemagglutinin haplotypes and tracked their detection across states. A small number of haplotypes dominated each season, and Michigan detected all major haplotypes, even under substantial downsampling. Detection delays were primarily driven by haplotype frequency rather than geographic factors. Comparisons across states showed that higher sequencing effort improved coverage and detection timeliness, with diminishing returns at higher volumes. Rarefaction analysis confirmed that relatively few sequences were needed to capture 95% of national haplotype diversity. These findings suggest that intensive sequencing in a single well-sampled location can be broadly representative of national influenza diversity. One sentence summaryDense influenza genomic sequencing from a single U.S. state captured nearly all nationally circulating haplotype diversity, with detection timeliness primarily driven by sequencing effort and haplotype frequency.

The red cell distribution width (RDW) is a recognized prognostic marker in sepsis, yet its dynamic changes over time and their relationship with outcomes remain unexplored. This study aimed to identify distinct RDW trajectories during the early phase of sepsis and evaluate their association with mortality. We conducted a retrospective cohort study using data from the MIMIC-IV database (n=3,813) as the derivation cohort and from the First Affiliated Hospital of Kunming Medical University (n=467) for external validation. Sepsis patients with at least seven RDW measurements within the first ten days of hospitalization were included. Group-based trajectory modeling (GBTM) was employed to identify RDW trajectories. A three-trajectory model was selected based on model fit indices and clinical interpretability: Trajectory 1 (Slow-Decrease, 32.97%), Trajectory 2 (Slow-Increase, 43.30%), and Trajectory 3 (Fluctuating-Rapid Decrease, 23.73%). In the our study, Cox models adjusted for confounders revealed that, compared to Trajectory 1, Trajectory 3 was independently associated with significantly increased 30-day (HR 1.47, 95% CI 1.17-1.84) and 90-day mortality (HR 1.54, 95% CI 1.25-1.88). Conversely, Trajectory 2 was associated with the most favorable survival rates. Kaplan-Meier analysis consistently showed the highest mortality in the Trajectory 3 group. External validation confirmed the models robustness and the consistent prognostic value of the identified trajectories. We conclude that dynamic RDW trajectories, readily identifiable from routine clinical data, provide significant prognostic information beyond single-time-point measurements and can aid in the risk stratification of sepsis patients.

Syphilis remains a major public health concern. However, current serologic assays are limited in their ability to distinguish active from previously treated disease. We applied tandem mass tag-based quantitative proteomics to plasma from 10 adults with active syphilis and 10 age- and gender-matched non-diseased controls. We identified 54 differentially regulated proteins (36 upregulated, 18 downregulated). Those proteins map to immune and inflammatory responses, acute-phase signaling, coagulation and vascular pathways, and cellular stress processes. Three sets of between 2-5 proteins achieved >99% discrimination between cases and controls. Our exploratory findings support proteomics as a potential tool to develop novel syphilis diagnostics.

BackgroundThe 2024 blood culture bottle shortage brought diagnostic resource allocation to the forefront, reflecting persistent, foundational challenges with low-value testing and empiric treatment approaches under clinical uncertainty. ObjectiveTo determine whether a machine learning approach using electronic medical record data can predict bacteremia more effectively than existing systems and practices to guide diagnostic testing and empiric treatment strategies. MethodsIn a retrospective cohort of 101,812 adult emergency department encounters (2015-2025), we first established an idealized cognitive baseline by evaluating physician and generative AI (GPT-5) application of the professional society-endorsed Fabre framework on a validation subset. We then trained an XGBoost model (Cultryx) on the full cohort to predict bacteremia, benchmarking its performance against real-world clinical heuristics (SIRS, Shapiro Rule). ResultsFor the idealized baseline, physicians applying the Fabre framework achieved 95.7% sensitivity, but GPT-5 automation failed to replicate this standard (71.6% sensitivity). In real-world benchmarking, Cultryx outperformed all clinical heuristics (AUROC 0.810). SIRS lacked specificity (41.2%), driving diagnostic overuse, while the Shapiro Rule lacked sensitivity (70.2%), missing ~30% of bacteremia cases. In contrast, when calibrated to a strict 95% sensitivity target, Cultryx achieved the highest culture volume deferral rate (26.2%, deferring ~ 15,872 bottles with predicted negative results) while maintaining a 98.9% negative predictive value. Cultryxscore, a simplified bedside tool, retained a 20.8% deferral rate. ConclusionsMachine learning provides a superior, data-driven alternative to mainstream clinical heuristics for predicting bacteremia. By maximizing culture deferment without compromising pathogen detection, Cultryx can conserve diagnostic resources, reduce unnecessary empiric antibiotic exposure, and systematically elevate patient safety. SummaryCultryx, a machine learning model for blood culture stewardship, outperforms standard clinical heuristics in predicting bacteremia. This approach could reduce culture utilization by over 26% while preserving pathogen detection, conserving diagnostic resources, reducing unnecessary antibiotic exposure, and elevating patient safety.

PurposeSepsis-associated immunothrombosis significantly contributes to high mortality, yet the role of N-glycosylation in this process remains poorly understood. This study aimed to comprehensively profile the plasma N-glycosylation landscape in sepsis and elucidate how its specific reprogramming in the complement and coagulation cascades influences immunothrombotic balance and patient outcomes. MethodsWe performed in-depth 4D-DIA proteomic and N-glycomic analyses on plasma from 43 sepsis patients and 9 healthy controls. Differential expression, weighted gene co-expression network analysis (WGCNA), and protein-glycosylation correlation analyses were used to characterize molecular features. Clinical relevance was assessed via correlation and survival analyses. ResultsExtensive N-glycosylation reprogramming was observed in sepsis plasma,with marked enrichment in complement and coagulation pathways(KEGG p=7.76x10- {superscript 2}{superscript 1}).Pro-coagulant proteins(eg,vWF,fibrinogen)showed increased abundance together with enhanced site-specific glycosylation,potentially amplifying their activity.In contrast,key anticoagulant proteins(eg,SERPINC1)displayed unchanged glycosylation at critical sites despite abundance changes,which may impair function.Survival analysis revealed distinct prognostic values of glycoproteins and specific glycosylation sites.For instance,high vWF protein levels predicted mortality(HR=2.83),whereas elevated glycosylation at vWF N211 was associated with improved survival(HR=0.135),suggesting a negative regulatory role.These glycosylation markers correlated closely with disease severity and prognosis,representing potential early-warning biomarkers independent of current clinical coagulation indicators. ConclusionOur study demonstrates widespread reprogramming of the plasma proteome and N-glycome in sepsis.We propose that decoupling of protein function from abundance through N-glycosylation in the complement-coagulation network contributes to immunothrombotic imbalance.Specific N-glycosylation sites may serve as novel prognostic biomarkers,offering new perspectives for early risk stratification and glycosylation-targeted therapies in sepsis. Key PointsO_LISepsis plasma exhibits specific N-glycosylation reprogramming overwhelmingly focused on the complement and coagulation cascade. C_LIO_LIA dominant "glycosylation-dominated co-upregulation" mode in procoagulant factors, coupled with a "silent" glycosylation state in key anticoagulants, drives prothrombotic imbalance. C_LIO_LISite-specific N-glycosylation levels provide prognostic information distinct from, and often superior to, their carrier protein abundance, offering novel early-risk biomarkers. C_LI

BackgroundPulse oximeters are typically validated on cohorts of 200-500 subjects under controlled conditions. Whether these cohorts capture the demographic heterogeneity of national clinical practice -- and whether measurement error is associated with patient outcomes -- has not been established at scale. MethodsWe analyzed paired SpO2/SaO2 readings from three independent sources spanning 209 U.S. hospitals: MIMIC-IV (1 hospital; 12,934 ICU stays), eICU-CRD (208 hospitals; 55,178 stays), and the Open Oximetry Repository (PhysioNet; 52.4 million readings with continuous melanin and perfusion indices). Bias was defined as SpO2 - SaO2. Hidden hypoxemia (SpO2 [&ge;] 94% with SaO2 < 88%) was assessed per ICU stay. Mortality was compared between hidden-hypoxemia-positive and -negative stays with multivariable logistic regression adjusting for age, sex, race, and four laboratory severity markers (cluster-robust SEs by hospital). Sensitivity analyses included landmark restriction (first 48 hours), lactate stratification, alternate thresholds, and patient-level aggregation. PPG signal quality was assessed in 125 ICU patients with demographic-linked waveform data. ResultsBias was minimal at normal perfusion but amplified under low perfusion in high-melanin patients, consistent with known optics: at very low perfusion x high melanin x severe hypoxia, mean bias reached +12.8% (n = 458,571), with 47% of readings constituting hidden severe hypoxemia. National bias in African American patients was +2.76% (n = 529,541; 208 hospitals), 62% higher than academic estimates. Across 55,178 eICU stays, hidden hypoxemia was associated with approximately doubled mortality after adjustment for age, sex, race, and illness severity (adjusted OR 1.86, 95% CI 1.69-2.04, p < 0.001), consistent across all racial groups. Hidden hypoxemia was not a pre-terminal phenomenon: 63% of events occurred >48 hours before death (median first event: 15.3 hours; mean time to death: 151 hours), and the association persisted in landmark analysis (first 48 hours only), in patients with normal lactate (adjusted OR 1.87, 95% CI 1.61-2.16), and when both restrictions were applied simultaneously (16.5% vs. 11.1%). Waveform analysis (n = 125) showed no fixed racial difference in baseline PPG AC/DC ratio (Black: 0.299, White: 0.273), suggesting the signal deficit is conditional on perfusion state. Full extraction (n = 1,545) is in progress. ConclusionsIn this multicenter retrospective analysis, national pulse oximetry variance exceeded published benchmarks and was associated with approximately doubled ICU mortality, replicated across 209 U.S. hospitals. Hidden hypoxemia was not a pre-terminal artifact: events occurred throughout the ICU stay at a constant rate, and mortality associations persisted in landmark and lactate-stratified analyses. These findings suggest that current regulatory validation standards may underestimate the real-world prevalence of demographic bias in pulse oximetry, and that perfusion-dependent mechanisms may offer a target for algorithmic correction.

ObjectiveTo develop and validate a predictive model incorporating behavioral telemetry signals--documentation pattern anomalies derived from routine EHR charting--alongside clinical variables for ICU mortality prediction in patients with low acute physiologic derangement. Materials and MethodsRetrospective cohort study of 46,002 adult ICU stays from MIMIC-IV v3.1 (2008-2022) with SOFA scores 0-2, excluding neurological units. We extracted 66 variables spanning demographics, acuity, behavioral telemetry, clinical enrichment, and temporal factors. Progressive logistic regression models (M1-M7) were compared using cross-validation, DeLong tests, net reclassification improvement, and calibration analysis. ResultsOverall mortality was 9.34% (4,295 deaths). The clinical model (M5) achieved cross-validated AUROC 0.691 versus 0.639 for demographics alone (M2; {Delta}AUROC = 0.052, DeLong p = 4.41x10-47). NRI was 24.3%. Discordant care patients received 30.5% more chart events than concordant patients, with the sole deficit in neurological assessments (-15.4%), refuting the neglect hypothesis. Kaplan-Meier analysis confirmed survival separation (log-rank {chi}2 = 138.6, p = 5.32x10-32). In the most conservative subgroup (SOFA 0, no sedation, no ventilation, N = 11,158), orientation omission remained associated with mortality (adjusted OR 1.52, p = 0.027). DiscussionDeep sedation and mechanical ventilation function as mediators on the causal pathway rather than traditional confounders; the discordant care signal retains significance after full sedation adjustment. ConclusionDocumentation pattern analysis adds measurable predictive value for ICU mortality risk stratification and represents a novel signal for real-time EHR-based clinical decision support.

ObjectivesTo develop and validate pediatric adaptations of the Venous Excess Ultrasound Score (P-VExUS) for noninvasive estimation of central venous pressure (CVP) in critically ill children. DesignProspective observational study. SettingPICU of a tertiary-care teaching hospital. PatientsFifty-six mechanically ventilated children (median age 7.4 months, median weight 6.0 kg) with central venous catheters. InterventionsNone. Measurements and Main ResultsVenous Doppler ultrasonography of the inferior vena cava, hepatic, portal, and intrarenal veins was performed at the bedside. Two P-VExUS models were tested: (1) a categorical grading system (0-III) and (2) a semiquantitative point-based score (0-7). Both models showed significant associations with CVP. For predicting elevated CVP (>12 mmHg), model 1 achieved an AUROC of 0.74 (95% CI 0.61-0.85) with 45% sensitivity and 98% specificity, while model 2 demonstrated superior accuracy with an AUROC of 0.94 (95% CI 0.84-0.98), sensitivity 82%, and specificity 91% (p < 0.001). For detecting low CVP (<7 mmHg), model 2 also outperformed model 1 (AUROC 0.80 vs. 0.69, p = 0.02). Among individual venous Doppler components, intrarenal veins had the highest discriminative ability (AUROC 0.92), followed by hepatic (0.89) and portal (0.80) veins. ConclusionsTwo pediatric-specific P-VExUS models were feasible and accurate for estimating CVP in critically ill children. The point-based model (model 2) demonstrated superior diagnostic performance, supporting its potential as a noninvasive tool to assess venous congestion at the bedside. Research in ContextO_LIVenous congestion, reflected by elevated central venous pressure (CVP), is associated with adverse outcomes in critically ill children, including mortality and renal dysfunction. C_LIO_LIThe Venous Excess Ultrasound Score (VExUS) is validated in adults, but pediatric-specific adaptations and cutoff values remain poorly defined. C_LIO_LIThere is a need for noninvasive, bedside tools to estimate CVP in children and guide fluid management in the PICU. C_LI What This Study MeansO_LIThis study validates pediatric-specific adaptations of the Venous Excess Ultrasound Score (P-VExUS) for estimating CVP in critically ill children. C_LIO_LIThe semiquantitative point-based model provided more consistent and accurate discrimination of venous congestion compared with categorical grading. C_LIO_LIThese findings highlight the feasibility and potential clinical utility of venous Doppler ultrasonography as a noninvasive bedside tool in the PICU. C_LI

BackgroundImmunocompromised (IC) individuals are at increased risk for persistent SARS-CoV-2 infections and can develop new viral mutations and lineages not seen in the community. In this case report, a persistent SARS-CoV-2 infection (330 days) in an IC patient is examined for viral mutations and mutations associated with cryptic lineages. Case PresentationThe patient was followed in a longitudinal study examining persistent SARS-CoV-2 in IC patients. The patient provided stool and nasal swab samples biweekly until 28 days post-enrollment, then monthly, and then quarterly after 12 month post enrollment until the participant was no longer positive for SARS-CoV-2. Staff performed RT-qPCR and viral sequencing on the samples. Viral mutations from the XBK lineage were already present in the initial sample. By the end of the infection period, there were 40 fixed consensus changes from XBK of which two mutations were typical for cryptic lineages. Mutations increased steadily over time, with most mutations fixed by day 253, including the cryptic typical mutations. ConclusionThis case demonstrates the potential for persistent SARS-CoV-2 infections to develop mutations and lineages in IC patients and highlights the need for continued SARS-CoV-2 monitoring and treatment in this vulnerable population.

ObjectivePreterm birth (PTB) is a leading cause of neonatal morbidity and mortality worldwide, with India alone contributing nearly 27% of the global PTB burden. Although alterations in the vaginal microbiome have been implicated in PTB, its association in the Indian context is underexplored. This study aimed to investigate the association of vaginal microbiome and PTB in Indian women at the time of delivery. Study designThe vaginal swabs were collected at the time of delivery from 72 women (31 term, 41 preterm) admitted to a tertiary care hospital in Western India. Microbial DNA was extracted, and the V3-V4 region of the 16S rRNA gene was sequenced. Community composition, alpha and beta diversity, and differential taxonomic abundance were assessed using bioinformatics pipelines. ResultsAt the time of delivery, there were no significant differences in alpha or beta diversity between term and preterm groups. Principal coordinate and unsupervised clustering analyses showed no group-wise segregation. The relative abundance of individual Lactobacillus species, including L. iners and L. helveticus, did not differ significantly between the two groups. However, a modest difference in the relative abundance of Streptococcus was observed between the two groups after adjustment. ConclusionThis study found no major microbial shifts in the vaginal microbiome associated with preterm birth in this cross sectional cohort of Indian women, suggesting that vaginal dysbiosis at the time of delivery may not be a principal driver of PTB in this population. These findings underscore the need for larger, longitudinal, and ethnically diverse studies using standardized methodologies better to understand the microbiomes role in PTB risk.

Clear cell renal cell carcinoma (ccRCC) is the leading cause of kidney cancer-related death, but how the tumor microenvironment shapes patient survival is not completely understood. Here, we describe the characterization of ccRCC tumor ecosystems from 498 patients using imaging mass cytometry with a focus on tumor, myeloid, and T cell landscapes. Data from more than 3 million single cells is analyzed using machine-learning to identify key ecosystem features that outperform basic clinical data for predicting patient survival. We define three survival ecotypes of ccRCC: Poor ecotypes, correlate with the worst survival, have high levels of ICAM1 and CD44 expression in tumor cells and are enriched in M2-like macrophages and interactions of exhausted CD8+ T cells with macrophages. Favorable ecotypes are characterized by high levels of VHL on tumor cells and of HLADR on myeloid cells and contain Th1-like CD4+ T cells. Medium ecotypes have the highest endothelial cell density and various immune-to-tumor interactions. Multi-omic characterization of these ecotypes using targeted genomic sequencing and metabolic imaging reveals distinct genomic and metabolic features, including BAP1 mutations in Poor and VHL monodriver/wild-type status in Favorable patients. We show that deep learning allows ecotype prediction directly from standard pathology H&E images. We validate the ecotypes and their associated molecular characteristics with orthogonal omics data across five clinical cohorts and more than 2,500 patients. These analyses highlight an overall survival benefit for Medium patients treated with immunotherapy. In summary, our study distills the survival-relevant information encoded in the ccRCC tumor microenvironment into prognostic survival ecotypes, which may inform clinical decision making in the future.

BackgroundQuality measurement in intensive care emphasizes task completion--whether assessments were documented and protocols followed. Electronic health record (EHR) systems capture these signals in real time, yet current metrics cannot distinguish task completion from cognitive clinical engagement. A prior analysis demonstrated that omission of orientation assessment predicted a 4.29-fold increase in hospital mortality among low-acuity ICU patients [1]. Whether combining this marker with routine task-completion data yields a computable phenotype with independent prognostic value has not been studied. ObjectiveTo define, validate, and characterize "discordant care"--a computable EHR phenotype defined as completion of [&ge;]6 of 8 routine nursing assessments without orientation assessment documentation--as a predictor of hospital mortality, distinguishing patient-level confounding from care process signal. MethodsRetrospective cohort study using MIMIC-IV v3.1 (2008-2022), including 46,004 adult ICU stays with SOFA scores 0-2 and length of stay [&ge;]24 hours in non-neurological ICUs. Primary exposure: discordant care, computed from structured nursing flowsheet data within 24 hours of admission. Primary outcome: hospital mortality. Progressive covariate adjustment included mechanical ventilation, sedation, and diagnosis. ResultsDiscordant care was present in 8891 patients (19.3%), with 69.7% mechanically ventilated versus 25.3% of concordant patients. Two overlapping signals were identified: a patient-level signal driven by ventilation/sedation (full adjustment OR 1.19, 95% CI 1.09-1.30) and a care process signal in non-ventilated patients (OR 2.14, 1.87-2.44; N=30,314). Among non-ventilated SOFA 0 patients, OR was 2.60 (2.13-3.18; N=16,295). The signal was present across all 7 major diagnosis categories. Quantitative bias analysis indicated unmeasured delirium could attenuate but likely not fully explain the non-ventilated signal. ConclusionsDiscordant care identifies two phenomena: a patient-level signal from ventilation/sedation and a care process signal where assessable patients receive routine care without cognitive engagement (OR 2.14-2.60). This care process signal is invisible to existing quality metrics and detectable in real time. Prospective validation with systematic delirium screening is needed.

Applying deep learning models to RNA-Seq data poses substantial challenges, primarily due to the high dimensionality of the data and the limited sample sizes. To address these issues, this study introduces an advanced deep learning pipeline that integrates feature engineering with data augmentation. The engineering application focuses on biomedical engineering, specifically the classification of RNA-Seq datasets for disease diagnosis. The proposed framework was initially validated on synthetic datasets generated from Naive Bayes, where MLP-based augmentation yielded a notable improvement in predictive performance. Building on this foundation, we applied the approach to chromophobe renal cell carcinoma (KICH) RNA-Seq data from The Cancer Genome Atlas (TCGA). Following standard preprocessing steps normalization, transformation, and dimensionality reduction, the analysis concentrated on three main aspects: augmentation strategies, preprocessing methods, and explainable AI (XAI) techniques in relation to classification outcomes. Feature selection was performed through PCA, Boruta, and RF-based methods. Three augmentation strategies linear interpolation, SMOTE, and MixUp were evaluated. To maintain methodological rigor, augmentation was applied exclusively to the training set, while the test set was held out for unbiased evaluation. Within this framework, we conducted a comparative assessment of multiple deep learning architectures, including MLP, GNN, and the recently proposed Kolmogorov-Arnold networks (KAN). The GNN achieved the highest classification accuracy (99.47%) when trained with MixUp augmentation combined with RF feature selection, and achieved the best F1 score (0.9948). Consequently, the GNN-based XAI framework was applied to the RF dataset enriched with MixUp. XAI analyses identified the top 20 most influential genes, such as HNF4A, DACH2, MAPK15, and NAT2, which played the greatest role in classification, thereby confirming the biological plausibility of the model outputs. To further validate model robustness, cervical cancer and Alzheimers RNA-Seq datasets were also tested, yielding consistent and reliable results. Overall, the findings highlight the value of incorporating data augmentation into deep learning models for RNA-Seq analysis, not only to improve predictive performance but also to enhance biological interpretability through explainable AI approaches.