To evaluate the performance of machine learning models in predicting liver metastasis in colorectal cancer (CRC) patients using the SEER database and external validation from Ningbo No.2 Hospital. The data on patients with colorectal cancer were obtained from Surveillance, Epidemiology, and End Results (SEER) database from 2010 to 2023. Patients were classified into training (n = 29017) and testing sets (n = 12437). The data were used to build eight machine learning models to predict liver metastasis in colorectal cancer patients. A total of 11 clinical variables were entered into these models. Model performance was measured with the area under the receiver operating characteristic curve (ROC) and area under precision-recall curve (AUPR). The models were visualized and interpreted using the SHAP method. In the SEER database cohort, the incidence of liver metastasis was 7.2% (2977/41,454). Of the eight machine learning models, Gradient Boosting (GB) had the best AUC (0.837) and AUPR (0.294). Upon external validation, the GB model achieved an AUC of 0.730 and an AUPR of 0.278. We explored the significance of features in the model through SHAP analysis. CEA, N stage and T stage were the heavily weighted factors used by the GB. An online calculator was developed for clinical use. The GB model demonstrates robust predictive performance for liver metastasis in CRC, validated internally and externally, and presents a potentially valuable tool for clinical decision-making.
Per- and polyfluoroalkyl substances (PFAS) represent a critical class of persistent environmental contaminants with significant ecological and human health implications. However, the rapid emergence of novel PFAS has far outpaced the development of reference mass spectral databases. Here, Neural Per- and Polyfluoroalkyl Substances Mass Spectrometry (NPFAS-MS), a transfer learning-based neural network model, was developed to predict PFAS-specific high-resolution mass spectra. NPFAS-MS was fine-tuned from a pretrained model using PFAS tandem mass (MS/MS) spectra. NPFAS-MS outperformed other in silico spectral prediction models for PFAS spectra prediction across multiple spectral similarity metrics. In library searching tasks, libraries generated by other spectral prediction models showed top-1 recall between 42.1% and 55.4%, while NPFAS-MS demonstrated 71.1%. Applying the virtual PFAS mass spectral library generated with NPFAS-MS using 10,553 PFAS structures from the U.S. EPA and NORMAN databases to groundwater and aqueous film-forming foam (AFFF) samples revealed more potential PFAS than other mass spectral databases. Specifically, 38 potential PFAS were annotated in AFFF products and 40 in groundwater samples. NPFAS-MS enabled characterization of emerging PFAS, including ultrashort-chain, unsaturated, and substituted derivatives in environmental matrices. This advancement enables comprehensive environmental monitoring of rapidly evolving PFAS contamination. NPFAS-MS and associated resources were deployed as a web-based tool at https://cosbi10.ee.ncku.edu.tw/NPFAS_MS/, enabling both structure-to-spectrum prediction and library searching against 31,659 predicted PFAS spectra.
A BigSMILES string encodes the structural connectivity of any polymer chemistry and topology as a linear string. However, multiple BigSMILES strings can encode the same ensemble, making string-based searches for polymers in digital databases challenging. This work presents a canonicalization algorithm that breaks the degeneracy of the BigSMILES language for both linear and branched polymers and can reverse-translate canonicalized structures back into BigSMILES. The algorithm was validated on broadly representative polymer chemistries and topologies from the literature. First, the BigSMILES string is mapped onto a tree automaton, a type of state machine that accommodates branch points and recognizes the same ensemble of molecules that BigSMILES encodes. The automaton can then be minimized into a unique graph with the fewest states through existing algorithms. Finally, a human-readable canonicalized BigSMILES is obtained upon translation of the state machine transition rules back into a string. This robust canonicalization algorithm allows polymers to be searched rapidly in large database systems, making data findable, accessible, interoperable, and reusable (FAIR) and enabling the development of novel data-driven approaches with BigSMILES.
Ferroptosis plays a significant role in pulmonary arterial hypertension (PAH), although its underlying mechanisms and key pathogenic genes remain unclear. Transcriptomic data from human PAH and control lung tissue were obtained from the Gene Expression Omnibus (GEO) database, whereas ferroptosis-related genes (FRGs) were sourced from the MsigDb and FerrDb databases. Differentially expressed FRGs (DE-FRGs) were identified through the intersection of FRGs with differentially expressed genes (DEGs). Functional enrichment analysis was performed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Key hub genes were identified through Least Absolute Shrinkage and Selection Operator (LASSO), support vector machine-recursive feature elimination (SVM-RFE), and weighted correlation network analysis (WGCNA). Gene set enrichment analysis (GSEA) was conducted to explore the functional roles and associated pathways of hub genes. The relationship between hub genes and immune infiltration was investigated. Expression levels of potential biomarkers were validated via Quantitative real-time polymerase chain reaction (qRT-PCR) and immunohistochemistry (IHC) in two PAH animal models (monocrotaline-induced and Sugen5416 plus hypoxia-induced PAH). Finally, molecular docking was employed to screen potential therapeutic compounds. A total of 133 DE-FRGs were identified, with KEGG and GO analyses highlighting their involvement in intracellular iron homeostasis and ferroptosis. Hub genes, notably FZD7 and NFE2, were identified using LASSO, SVM-RFE, and WGCNA. Immune infiltration analysis suggested that monocytes and neutrophils play key roles in PAH pathogenesis. Validation in PAH animal models showed significant upregulation of Fzd7 and downregulation of Nfe2 in lung tissues of both MCT- and SuHx-induced PAH models. Molecular docking identified tetrachlorodibenzodioxin (TCDD) has good binding affinity. In summary, we investigated two ferroptosis-related biomarkers, FZD7 and NFE2, in PAH using transcriptomics, offering new insights into molecular mechanisms and potential targeted therapies for the disease.
This study is to systematically evaluate the efficacy of acupuncture for PHN and provide a visual overview of treatment landscape. A systematic search was conducted in PubMed, Embase, Web of Science, China National Knowledge Infrastructure (CNKI), Chinese Scientific Journals Database (VIP), and Wanfang Database for systematic reviews (SRs) on acupuncture for PHN up to Apr 18, 2025. Studies were included if they were SRs of randomized controlled trials (RCTs) assessing traditional Chinese acupuncture interventions for PHN, and excluded if they involved non-traditional acupuncture, herpes zoster, or PHN prevention research. Two independent reviewers utilized Excel, EndNote 20, and R software for data analysis and assessed the quality of included studies using the AMSTAR2 tool. Of 351 identified records, 40 SRs met inclusion criteria, encompassing 926 RCTs, 63,493 patients, 13 types of acupuncture interventions and 29 outcomes. Acupuncture interventions, particularly fire needling, CPBLC, Fu's subcutaneous needling, plum-blossom needle, multi-acupuncture and multi-acupuncture + pharmacotherapy, showed the most robust benefits in improving effective rate, reducing visual analog scale (VAS) scores, and decreasing adverse reactions in PHN treatment. Despite most SRs reporting positive outcomes, the quality was generally low by AMSTAR2. Acupuncture could be a valuable adjunct to standard PHN treatment, offering benefits in overall efficacy, pain management and treatment safety. However, high-quality clinical trials and systematic reviews are needed to confirm these preliminary results and guide clinical practice.
Venous thromboembolism (VTE), which encompasses deep venous thrombosis (DVT) and pulmonary embolism (PE), is a common preventable complication in hospitalized patients. Risk assessment tools allow for easy stratification of patient VTE risk and have been demonstrated to reduce incidence of VTE. However, risk assessment tools remain underutilized in clinical practice. This scoping review aims to explore barriers and facilitators to VTE risk assessment usage to improve rates of hospital-acquired VTE and provide recommendations for future implementation strategies. Four databases (PubMed/MEDLINE, Embase, Cumulative Index to Nursing and Allied Health Literature, and Cochrane Database of Systematic Reviews) were searched from January 1990 through December 2025, and 59 studies were included after selection by three independent reviewers. Themes related to 'decreased provider compliance' and 'difficulty of use' were the most commonly cited barriers. For facilitators, the majority of themes surrounded 'electronic medical record integration,' 'forcing functions,' and 'education.' A prevalence of barriers and a paucity of facilitators contribute to decreased VTE risk assessment usage. Hospital administrators and clinicians should address current barriers and promote facilitators during VTE risk assessment initiatives to maximize patient quality improvement outcomes. (Prospero ID: CRD42022360033).
Peripheral T-cell lymphoma-not otherwise specified (PTCL-NOS) is a highly aggressive and heterogeneous lymphoma subtype with a poor prognosis. This study aims to develop a machine learning-based model to predict early death (within 3 months of diagnosis) in PTCL-NOS patients using data from the SEER database (2016-2021). A total of 1,156 patients were included and randomly divided into training (n = 809) and validation (n = 347) sets. Key predictive factors were identified through Boruta and LASSO algorithms, including chemotherapy, radiotherapy, age, B symptoms, primary tumor site, Summary Stage, and Ann Arbor Stage. Seven machine learning models were constructed and evaluated using AUROC, AUPRC, calibration curves, Brier scores, and decision curve analysis. XGBoost demonstrated the best predictive performance (AUROC = 0.842 in training and 0.774 in validation). This study provides a novel and interpretable predictive tool that can aid in early risk stratification and personalized treatment planning for PTCL-NOS patients, ultimately improving clinical outcomes.
Safety-net hospital (SNH) status is associated with high perioperative morbidity in pulmonary resection. Multiple etiologies have been proposed, including delays in diagnosis and inefficient care pathways. In integrated health systems, surgical volume has been shown to improve outcomes in pulmonary resection. However, whether surgical volume can overcome the inherent challenges of SNHs is unclear. We hypothesize that surgical volume is associated with improved outcomes at SNHs. The 2016 to 2021 Nationwide Readmissions Database was queried for all adult (≥18 years) patients undergoing elective lobectomy for lung cancer. Centers in the top quartile of Medicaid or self-pay/uninsured admissions were defined as SNHs. SNHs were further stratified by lobectomy caseload as a low-volume hospital (<10 cases/y), medium-volume hospital (10-33 cases/y), or high-volume hospital (>33 cases/y). Multivariable regressions were built to consider the independent association of hospital volume on acute clinical and financial outcomes among patients treated at SNHs. Care at high-volume centers remained associated with significantly reduced likelihood of overall major morbidity (adjusted odds ratio [AOR], 0.81; 95% CI, 0.68-0.97), respiratory complications (AOR, 0.79; 95% CI, 0.65-0.96), need for blood transfusion (AOR, 0.67; 95% CI, 0.48-0.93), and nonhome discharge (AOR, 0.66; 95% CI, 0.48-0.88). Care at high-volume centers was also associated with a decrease in duration of hospitalization (β = -1.02 days; 95% CI, -1.48 to -0.54 days) and overall expenditures (β = -$4360; 95% CI, -$7020 to -$1700). Surgical volume is associated with improved outcomes in pulmonary resection at SNHs. Patients who are eligible for care only at SNHs can still benefit from undergoing pulmonary resection at a high-volume center.
Promoting remyelination is a key therapeutic goal in demyelinating diseases such as multiple sclerosis (MS), yet effective strategies remain limited. Sphingosine-1-phosphate (S1P), a ubiquitous bioactive lipid, has emerged as a key therapeutic target in MS due to its dual roles in immune regulation and neuroprotection; however, the therapeutic efficacy of current S1P-based therapies in remyelination remains unclear. This systematic review evaluated in vivo studies up to July 2025, in accordance with PRISMA guidelines, to assess the efficacy of S1P modulators on remyelination in mammalian models of demyelination. A comprehensive search across three databases identified 24 eligible studies that investigated S1P receptor (S1PR) modulation in both acute and chronic models of demyelination, with or without immune-mediated components. Fingolimod was the most extensively studied compound (16 studies). Of the 18 studies assessing demyelination outcomes, S1P modulation consistently attenuated myelin loss and oligodendrocyte depletion. In contrast, remyelination outcomes were inconsistent: among 15 studies assessing repair, most reported no significant enhancement. While fingolimod showed limited evidence on remyelination, more promising effects were observed with selective S1PR1/5 modulators such as siponimod and ponesimod. Overall, current evidence supports a model in which S1P modulators act primarily through S1PR1-mediated immunomodulation and S1PR5-associated oligodendroglial protection, preserving oligodendrocyte lineage cells rather than driving terminal differentiation or de novo remyelination. Several compounds displayed bell-shaped dose-response patterns, highlighting the importance of dosing and treatment paradigms. Collectively, these findings indicate S1PR-based therapies primarily limit demyelination, with limited evidence of remyelination, emphasising the need for more efficacious S1P modulators to improve MS outcomes.
ObjectiveTo explore and describe the components and underpinning theories of nurse-initiated transitional (hospital-to-home) care and to evaluate the effects of interventions on activities of daily living (ADLs), quality of life (QoL), depression, anxiety, self-efficacy, stroke-related knowledge, patient satisfaction and healthcare service utilisation in older stroke survivors and self-efficacy, caregiver burden, QoL and satisfaction with care in caregivers.Data sourcesMEDLINE, CINAHL, Cochrane Library, Health and Medical Collection, Nursing and Allied Health, Web of Science, Health Source: Nursing/Academic Edition and Scopus databases were searched in February 2024 and updated in March 2026.Review methodsRandomised controlled trials (RCTs) and cluster-RCTs were included. Risk of bias was assessed using the Cochrane Risk of Bias tool, and evidence quality was rated with GRADE. Meta-analysis was undertaken using random-effects models.ResultsSeventeen trials were included. Transitional care interventions were guided by various theoretical frameworks and had multiple components. Interventions improved ADLs at 1 month (standardised mean difference (SMD) 0.54; 95% CI 0.12-0.97, four studies) and 3 months (SMD 0.43; 95% CI 0.12-0.74, seven studies), increased patient satisfaction and reduced hospital readmissions up to 3 months. Interventions may improve the Mental Component Score at 1 month, Role Limitations due to Physical Problems and General Health of QoL at 6 months. The certainty of evidence was low to very low.ConclusionNurse-initiated transitional care can improve ADLs and QoL, while reducing hospital readmissions among older stroke survivors during the hospital-to-home transition. Multicomponent interventions combining home visits and follow-up phone calls may enhance improvements in ADLs. High-quality studies are needed to clarify long-term effects.PROSPERO IDCRD42024517619.
Background:The COVID-19 pandemic disrupted work environments worldwide, increasing productivity loss through absenteeism and presenteeism. Identifying key associated factors is essential for informing workplace health strategies during public health crises. Methods/Project: A systematic review and meta-analysis were conducted following PRISMA guidelines, using comprehensive searches of seven electronic databases from inception through January 2024. Studies were systematically selected based on predefined eligibility criteria, and 24 studies examining individual and work-related factors associated with work productivity loss were included. Risk of bias was assessed using the Joanna Briggs Institute critical appraisal tools. Correlation coefficients were synthesized using a random-effects meta-analysis of correlations in STATA 17.0, and heterogeneity was evaluated using the I2 statistic and Cochran's Q test. Findings: Twenty-one factors were analyzed. Job stress, fear of COVID-19, mental health problems, job insecurity, turnover intention, exhaustion, and job demands exhibited moderate positive correlations with productivity loss during the COVID-19 pandemic. Fear of COVID-19 and mental health problems showed relatively large positive correlations with presenteeism. General health status was the factor most strongly associated with absenteeism, exhibiting a moderate negative correlation. Conclusions/Application to Practice: These findings identify key individual and work-related determinants of productivity loss during pandemics. The results support the development of targeted workplace health promotion, mental health support, and preparedness strategies to mitigate productivity loss during future public health emergencies.
Alzheimer's disease (AD) is characterized by progressive cognitive decline accompanied by profound disturbances in cerebral energy metabolism. Mitochondrial dysfunction has long been implicated in AD pathophysiology; however, the specific contribution of mitochondrial enzymes in human disease remains fragmented across heterogeneous studies. Enzymes regulating carbon entry into the tricarboxylic acid cycle, oxidative phosphorylation, and redox balance represent key metabolic control points whose dysfunction may contribute to neuronal vulnerability. To systematically synthesize human evidence on mitochondrial enzyme alterations in Alzheimer's disease and to evaluate the feasibility of quantitative meta-analysis based on current reporting practices. A systematic literature search was conducted in PubMed, Scopus, and Web of Science from database inception through January 2026 in accordance with PRISMA 2020 guidelines. Studies were included if they investigated mitochondrial enzymes in human postmortem brain tissue, human-derived cellular models, or peripheral biospecimens. Risk of bias was assessed using the ROBINS-I tool. The feasibility of meta-analysis was evaluated based on the availability and comparability of group-level summary statistics. Fifteen studies met the eligibility criteria and were included in the final synthesis. Mitochondrial enzymes involved in carbon entry into the tricarboxylic acid cycle, oxidative phosphorylation, redox regulation, and neurotransmitter-linked mitochondrial metabolism were the most frequently investigated targets. Direct enzyme-activity evidence most consistently implicated selected metabolic control points, particularly PDHC and αKGDHC, whereas additional studies supported mitochondrial impairment through protein or post-translational modification changes, respiratory dysfunction, redox alterations, or RNA-regulatory mechanisms. Quantitative meta-analysis was not feasible due to heterogeneous assay methodologies, variable normalization strategies, and inconsistent reporting of group-level summary statistics. Human evidence consistently implicates mitochondrial enzyme dysfunction as a central metabolic feature of Alzheimer's disease. However, progress toward cumulative quantitative synthesis remains limited by methodological heterogeneity and incomplete reporting of enzyme activity outcomes. Standardized measurement and reporting of mitochondrial enzyme alterations will be essential to advance mechanistic understanding and enable future meta-analytic integration.
Surgeon case volume has been linked with outcomes across many orthopaedic procedures, but its influence on distal radius fracture fixation remains uncertain. (1) For distal radius fracture surgery, at what surgeon annual case volume does the risk of complications plateau? (2) For distal radius fracture surgery, at what surgeon annual case volume does the risk of revision surgery plateau? A retrospective, population-based study was performed using administrative health databases in Ontario, Canada, accessed through the Institute for Clinical Evaluative Sciences, an independent, nonprofit research institute that houses linkable, individual-level health administrative data for Ontario's publicly funded healthcare system. Between 2010 and 2020, a total of 27,945 adult patients (≥ 18 years of age) underwent surgical fixation for acute isolated distal radius fracture. After applying prespecified inclusion and exclusion criteria, including exclusion of patients with open fractures, polytrauma, compartment syndrome, neurovascular injury, emergent presentations, incomplete administrative records, or prior distal radius surgery, a final cohort of 13,389 patients (48% of the initial cohort) was included (71% [9533] females; mean ± SD age 56 ± 15 years). Surgeon annual case volume, defined as the number of distal radius fracture fixations performed in the preceding year, was the primary exposure. The primary outcome was a composite of complications, including postoperative complications or revision surgery up to 10 years after the index procedure; revision surgery was also analyzed separately. Cox proportional hazards models were adjusted for demographics, comorbidities, fracture type (intraarticular versus extraarticular), fixation method, and hospital type (teaching versus nonteaching). Restricted cubic spline models were used to assess nonlinearity and identify potential volume thresholds. Surgeons performing < 5 distal radius fracture fixations annually had the highest hazards of both composite complications and revision surgery. Complication hazards declined with increasing surgeon volume and stabilized after approximately 20 procedures per year; consistent with this threshold, surgeons performing 20 to 24 procedures annually demonstrated a 37% lower hazard of complications compared with surgeons performing < 5 procedures per year (HR 0.63 [95% confidence interval (CI) 0.49 to 0.81]; p = 0.004). Revision surgery hazards likewise declined with increasing surgeon volume but plateaued at a lower threshold of approximately 10 procedures per year; surgeons performing 10 to 14 procedures annually had a 56% lower hazard of revision surgery compared with surgeons performing < 5 procedures per year (HR 0.44 [95% CI 0.33 to 0.60]; p < 0.001). Surgeons who perform distal radius fracture fixation infrequently may benefit from focused strategies to support maintenance of procedural proficiency including continuing professional development and enhanced surgical training. At a systems level, the lower risk of complications observed among surgeons performing at least 20 procedures per year have implications for training programs, ongoing competence frameworks, and health-system planning, particularly in settings where referral options may be limited. Level III, prognostic study.
Cancer diagnoses impact adherence to antidiabetic medications, but limited research has focused on patients with prostate cancer and type 2 diabetes (T2DM). We investigated adherence trajectories to oral antidiabetic medications one year before and after a prostate cancer diagnosis and identified risk factors. This retrospective cohort study used the 2011-2021 MarketScan Commercial and Medicare Supplemental databases. We included newly diagnosed prostate cancer patients with T2DM with continuous insurance enrollment. We applied group-based trajectory modeling with a beta distribution to evaluate adherence patterns before and after prostate cancer diagnosis. Model covariates included age, total number of medications, number of antidiabetic medications, the Charlson Comorbidity Index (CCI), cost, insurance type, and complicated diabetes from the year before diagnosis. Metastasis and cancer treatments were included in the model after diagnosis. The study included 7864 patients (mean age = 74.5 ± 7.1). Three adherence trajectories were identified before diagnosis: consistently high adherence, steady decliners, and consistently low adherence. After diagnosis, a fourth trajectory revealing a moderate decline emerged. Over half (61.2%) changed adherence patterns after diagnosis. Among those with consistently high adherence before diagnosis, 57.8% transitioned to a lower adherence trajectory. In contrast, 58.5% of steady decliners and 57.6% of consistently low adherents transitioned to a higher adherence trajectory after diagnosis. Predictors of high adherence included older age, fewer antidiabetic medications, lower CCI, and complicated diabetes before diagnosis. After diagnosis, fewer antidiabetic medications and complicated diabetes remained predictive of high adherence. Patterns of adherence to oral antidiabetic medications undergo substantial changes after a prostate cancer diagnosis. Targeted interventions are needed to support and facilitate effective diabetes management in this population.
The application of machine learning (ML) models in healthcare management offers high potential. In particular, resource allocation and operational decision-making in intensive care units (ICUs) can benefit from ML predictions, leading to improvements in patient outcomes and operational efficiency. However, the generalizability of these models across diverse hospital settings with potentially different patient populations remains a critical challenge. This study examines the generalizability of ML-based ICU outcome prediction models built using external data. We utilize data from two sources: a European University Hospital (EUH) dataset from Universitätsklinikum Carl Gustav Carus Dresden, Germany and the Medical Information Mart for Intensive Care (MIMIC)-IV database, representing different healthcare systems and patient populations. Our approach evaluates multiple models of varying architectures and complexity across three common prediction tasks in ICU settings (mortality, length of stay, and readmission), analyzes the impact of data availability on model performance, and applies interpretability techniques to identify features and scenarios where models succeed or fail in new environments. We found that locally trained models generally outperform those using external data when sufficient local data is available. Low and medium complexity models, such as generalized additive models, demonstrate significantly superior generalizability compared to high complexity models and require substantially less local data for high-quality predictions, offering evidence-based guidance for healthcare managers dealing with limited data resources. Our results demonstrate how interpretability techniques can identify dataset differences that hinder generalizability, providing valuable insights for healthcare practitioners in implementing ML solutions across diverse hospitals. This research contributes to the development of more generalizable and interpretable ML models in healthcare.
ObjectiveLipedema, which mainly affects women, is a chronic and progressive disorder characterized by abnormal adipose tissue accumulation in the limbs. Despite its clinical importance, research on lipedema remains limited. Bibliometric analysis provides a quantitative way to evaluate the literature, identify trends, and assess research impact.Materials and methodsGlobal lipedema research was analyzed in the Web of Science database using the terms "lipedema", "lipoedema", and "lipolymphedema" for publications indexed through March 2025. Articles were classified by publication type, year, country of origin, journal quartile, and citation count. Citation analyses excluded publications from 2024 and 2025 because citation accumulation was incomplete. Only English original articles and reviews were included, while editorials, meeting abstracts, and non-indexed sources were excluded.ResultsOf 610 records identified, 382 met the inclusion criteria. The analysis identified the main contributing countries and highlighted knowledge gaps and opportunities for multidisciplinary collaboration in the evolving field of lipedema research.ConclusionsThis study provides a global overview of lipedema-related research and its scholarly development. It also highlights the need for further studies on the pathophysiology, diagnosis, and treatment of lipedema.
Antipsychotic medications are associated with increased cardiovascular morbidity and mortality in patients with serious mental illness (SMI). Clozapine-induced myocarditis and cardiomyopathy are well established; however, emerging pharmacovigilance and observational data suggest that cardiotoxic risk extends beyond clozapine to other antipsychotic agents. We conducted a narrative review of the literature (2000-2026), including pharmacovigilance databases (VigiBase, FAERS), cohort studies, case series, and consensus guidelines. Emphasis was placed on epidemiology, mechanisms, comparative risk, clinical presentation, diagnosis, and practical management strategies relevant to psychiatric practice. Clozapine demonstrates the highest risk of myocarditis (estimated incidence 0.3-3%), typically within the first 4-8 weeks of treatment. Pharmacovigilance data consistently identify secondary signals for quetiapine and olanzapine, with weaker but present associations for risperidone, aripiprazole, and other agents. Proposed mechanisms include hypersensitivity myocarditis, inflammatory cytokine activation, oxidative stress, and mitochondrial dysfunction. Early recognition through symptom monitoring and selective biomarker use significantly improves outcomes. Cardiotoxicity associated with antipsychotics is not limited to clozapine. Psychiatrists should adopt a risk-stratified approach incorporating cautious titration, early symptom recognition, and targeted monitoring. Greater awareness and multidisciplinary collaboration are essential to optimize both psychiatric and cardiovascular outcomes.
Cutoff values for the Berg Balance Scale (BBS) and the Mini-Balance Evaluation Systems Test (Mini-BESTest) are available, but the recommendation for each reference standard is unclear. To investigate the diagnostic and prognostic accuracy of the BBS and the Mini-BESTest cutoff values for each reference standard in individuals with stroke. A systematic search was conducted of 6 databases and the reference list of included articles through December 24, 2025. We included the diagnostic and predictive accuracy studies reporting cutoff values and the sensitivity and specificity of the BBS or Mini-BESTest in individuals with stroke, across all the reported reference standards. Two reviewers independently extracted data from the included studies and assessed quality assessment using the QUADAS-2 tool. Meta-analysis was performed for reference standards with ≥2 studies. Fifty-five studies involving 6865 participants were included. The studies used 23 different reference standards. A high risk of bias was identified in 48 studies. For diagnosis of independent walking, the BBS showed higher accuracy (area under the curve [AUC] 0.917, sensitivity 88.0%, specificity 85.1%) than the Mini-BESTest (AUC 0.790, sensitivity 73.6%, specificity 80.2%). For community ambulation, the Mini-BESTest showed higher accuracy (AUC 0.882, sensitivity 80.0%, specificity 83.4%) than the BBS (AUC 0.844, sensitivity 81.4%, specificity 74.6%). For the prediction of independent walking, the BBS showed higher accuracy (AUC 0.858, sensitivity 77.8%, specificity 80.5%), whereas the Mini-BESTest was not analyzed because only one study was available. For the prediction of falling, the BBS showed moderate accuracy (AUC 0.704, sensitivity 61.8%, specificity 71.1%), whereas the Mini-BESTest was not analyzed because only one study was available. The BBS showed good accuracy for diagnosing and predicting walking independence, although most studies were at high risk of bias. Clinicians may consider selecting between the BBS and Mini-BESTest according to the outcome of interest.
Type 2 diabetes mellitus (T2DM) and bladder urothelial carcinoma (BLCA) are two kinds of diseases that seriously threaten human health. Their pathogenesis is complex and involves the interaction of multiple genes and multiple pathways. Recent epidemiological studies have shown that the risk of BLCA in patients with T2DM is significantly higher than that in non-diabetic people, suggesting that there may be a potential biological correlation between the two. Genomic studies have opened up new ways to reveal the common genetic characteristics of T2DM and BLCA. However, most of the current studies only focus on a single disease, and the comorbidity mechanism of these two diseases still needs to be further explored. Firstly, the datasets of BLCA and T2DM were downloaded from the The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases respectively. Differential expression genes (DEGs) were identified using the Limma package. Weighted gene co-expression network analysis (WGCNA) was employed to determine the co-expression modules related to BLCA and T2DM, and the common potential target genes were obtained. Correlation analysis and enrichment analysis were conducted on these target genes. Then, the best diagnostic biomarker - plasminogen activator (PLAU) was selected using machine learning algorithms. Additionally, the role of PLAU in the progression of T2DM and BLCA was confirmed through immunohistochemistry, Western Blot, and Edu experiments. Finally, small molecule compounds targeting PLAU were discovered through molecular docking and virtual screening, and the inhibitory effect of these small molecules on the progression of bladder urothelial carcinoma was verified through experiments. This study conducted a combined limma and WGCNA analysis on the T2DM and BLCA datasets to identify 42 common potential target genes, which were enriched in pathways such as innate immunity. Using machine learning algorithms such as LASSO and SVM, PLAU was identified as the best diagnostic marker for T2DM combined with BLCA. It was significantly highly expressed in both T2DM and BLCA samples, and high expression of PLAU predicted a shorter overall survival period for BLCA patients. Experimental results confirmed that PLAU was highly expressed in BLCA tissues and increased with the severity of malignancy. Knockdown (sh-PLAU) of PLAU could inhibit cancer cell proliferation and migration in a high-glucose environment, while overexpression (oe-PLAU) still promoted cancer cell progression in a low-glucose environment. Finally, molecular docking virtual screening revealed that the small molecule compound epigallocatechin gallate (EGCG) could target and inhibit PLAU, and effectively inhibited the proliferation and invasion of BLCA cells in experiments. The results of this study reveal the role of PLAU, a common characteristic gene of T2DM and BLCA, whose high expression drives tumor progression and poor prognosis. Moreover, small molecule drugs targeting PLAU, such as EGCG, have therapeutic potential. This study provides a new direction for accurate diagnosis and treatment of BLCA patients with T2DM.
Narrative medicine is defined as medicine practiced with the competence to absorb, interpret, and respond to narratives. We hereby present a resource compiling narrative medicine texts, aiming to make narratives created by patients and/or their families fully accessible to citizens, by developing a documentary database and describing its characteristics. Active bibliographic search, March-June 2022 for narratives in Spanish and/or Catalan written after the year 2000 by patients and/or their companions. Subsequently, narratives up to June 2024 were included. The compilation is available in a searchable and open-source web ( https://osf.io/pk9b3/ ). Three hundred seventeen narratives, 50.14% written by women, are showing an increase from 2020 onwards. Texts are related to cancer/hematological diseases (45.11%), mental illnesses (10.41%), neurodegenerative diseases (9.4%). Personal stories (28.7%), autobiographical (11.29%), companion stories (5%), children's or young adult stories/narratives (8.87%). There are studies, websites, and digital platforms that recognize the importance of narrative as part of the therapeutic process and how it improves the experience of illness (either one's own or that of a family member). Despite this, to date, no one had compiled a collection of patient texts in Spanish and Catalan. For this reason, we believe our database is innovative and can pave the way for improving the patient-professional relationship.