Tag: pubmed

Front Med (Lausanne). 2026 May 14;13:1834636. doi: 10.3389/fmed.2026.1834636. eCollection 2026.

ABSTRACT

OBJECTIVES: Artificial intelligence (AI)-enhanced virtual patient simulations are increasingly used in health professions education to improve clinical communication and diagnostic reasoning. However, the effectiveness of these technologies for psychiatric interview training has not been systematically quantified. This study aimed to systematically review and meta-analyze the existing literature evaluating the impact of AI-enhanced virtual patients on psychiatric interview performance, knowledge acquisition, and learner confidence in health professions education.

MATERIALS AND METHODS: A systematic review and meta-analysis was conducted following the PRISMA 2020 guidelines. Electronic database searches were performed in PubMed/MEDLINE, Scopus, Web of Science, and Google Scholar to identify relevant studies published between January 2000 and March 2026. Studies were included if they evaluated AI-enhanced virtual patient simulations for psychiatric interview training among medical students, psychiatry residents, clinicians, or other health professions trainees. Data extraction included study characteristics, participant populations, intervention types, and educational outcomes. Risk of bias was assessed using the Cochrane Risk of Bias Tool. Quantitative synthesis was performed using random-effects meta-analysis models, and effect sizes were calculated as standardized mean differences (SMD) with 95% confidence intervals (CI) using R statistical software.

RESULTS: A total of 560 records were identified through database searches and additional sources. After removal of duplicates and screening procedures, 10 studies met the inclusion criteria and were included in the final analysis. The studies involved approximately 450 participants, including medical students, psychiatry residents, clinicians, nursing students, and psychology trainees. AI-enhanced virtual patient interventions included conversational AI systems, virtual human simulations, large language model-based simulated patients, and AI-virtual reality training environments. The pooled analyses indicated improvements in psychiatric interview performance, knowledge acquisition, and learner confidence following AI-supported virtual patient training. Subgroup analysis demonstrated positive educational outcomes across both student and clinician populations. Risk-of-bias assessment revealed variable methodological quality across studies, with several pilot and non-randomized designs.

CONCLUSION: AI-enhanced virtual patient simulations appear to be effective educational tools for improving psychiatric interview training in health professions education. These technologies provide scalable and standardized simulation environments that support communication skill development, diagnostic reasoning, and learner confidence. Although the findings suggest promising educational benefits, further large-scale randomized controlled trials and standardized outcome assessments are needed to confirm the long-term educational impact of AI-supported virtual patient training in psychiatry.

PMID:42221128 | PMC:PMC13215799 | DOI:10.3389/fmed.2026.1834636

Front Med (Lausanne). 2026 May 14;13:1830071. doi: 10.3389/fmed.2026.1830071. eCollection 2026.

ABSTRACT

BACKGROUND: Type 2 diabetes mellitus (T2DM) is increasing rapidly in low- and middle-income countries, including Lao People’s Democratic Republic (Lao PDR). Although behavioral self-management is widely considered essential in diabetes care, evidence linking psychosocial determinants to glycemic outcomes among older adults remains inconsistent.

OBJECTIVE: This study examined the associations between diabetes-related knowledge, attitudes, self-care behaviors, and glycemic control among older adults with T2DM receiving tertiary hospital care in Lao PDR.

METHODS: A cross-sectional study was conducted among 88 adults aged ≥60 years with diagnosed T2DM attending the outpatient diabetes clinic at Setthathirath Hospital in Vientiane Capital. Structured interviews were used to assess diabetes knowledge, attitudes, and self-care practices. Glycemic control was defined as HbA1c < 7%. Pearson correlation and multivariable regression analyses were performed to examine associations between psychosocial factors and glycemic outcomes.

RESULTS: A total of 19.3% of participants achieved glycemic control (HbA1c < 7%), with a mean HbA1c level of 9.03 ± 2.47%, indicating generally poor glycemic control. Diabetes knowledge levels were low, with 98.9% of participants classified as having low knowledge. Attitudes toward diabetes management were predominantly low (60.2%), while overall self-care behaviors were largely moderate (83.0%). Pearson correlation analysis showed no statistically significant associations between knowledge (r = -0.134, p = 0.213), attitudes (r = 0.108, p = 0.318), or self-care behaviors (r = 0.046, p = 0.671) and HbA1c levels. Multivariable regression analysis likewise identified no significant predictors of glycemic control.

CONCLUSION: Despite substantial psychosocial vulnerabilities, no statistically significant associations between psychosocial factors and glycemic control were observed in this sample. These findings may indicate a potential mismatch between psychosocial factors and glycemic outcomes; however, this interpretation should be approached with caution, given the study’s methodological limitations. Further research with larger samples and longitudinal designs is needed to better understand these relationships. This study contributes context-specific evidence from Lao PDR to the limited literature on psychosocial determinants of diabetes management in low- and middle-income countries.

PMID:42221121 | PMC:PMC13215864 | DOI:10.3389/fmed.2026.1830071

Front Med (Lausanne). 2026 May 14;13:1804544. doi: 10.3389/fmed.2026.1804544. eCollection 2026.

ABSTRACT

BACKGROUND: Chronic obstructive pulmonary disease (COPD) is a chronic respiratory disease characterized by persistent respiratory symptoms and progressive airflow limitation. Acute exacerbations of COPD (AECOPD) are significant causes of hospitalization and death among COPD patients. This study aims to identify risk factors for AECOPD exacerbations and develop a highly accurate and interpretable predictive model using various statistical and machine learning methods.

METHODS: We retrospectively analyzed data from 2,102 COPD patients admitted between 1 January 2019 and 31 December 2024. The primary outcome was AECOPD severity, defined as the need for treatment escalation. Initial feature selection was performed using LASSO regression to identify potential risk factors. To validate the model’s effectiveness and explore its superior predictive performance, the dataset was partitioned by time period and proportion: The first 70% of observations in chronological order were used as the training set, with the remaining 30% as the test set. Multiple machine learning algorithms were then employed for model construction and comparison. To enhance model interpretability, we utilized SHapley Additive exPlanations (SHAP) to illustrate the contribution of each variable to the prediction outcomes.

RESULTS: Among the six machine learning models, the extreme gradient boosting (XGBoost) model demonstrated the optimal predictive performance, achieving an area under the receiver operating characteristic curve (AUC) of 0.960 (95% confidence interval (CI): 0.940-0.980) in the training set and 0.824 (95% CI: 0.804-0.844) in the test set. In the test set, the evaluation metrics were as follows: accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were 0.805, 0.65, 0.872, 0.669, and 0.859, respectively. SHAP analysis revealed that creatinine (CREA), neutrophil percentage (NEU%), D-dimer, brain natriuretic peptide (BNP), white blood cell count (WBC), and hypertension (HTN) were important factors influencing the model output.

CONCLUSION: The XGBoost model developed in this study demonstrates robust performance in predicting AECOPD risk using routinely collected clinical and laboratory data. The integration of SHAP analysis enhances model transparency, supporting its potential utility in clinical risk stratification and early intervention.

PMID:42221119 | PMC:PMC13216505 | DOI:10.3389/fmed.2026.1804544

Front Med (Lausanne). 2026 May 14;13:1803092. doi: 10.3389/fmed.2026.1803092. eCollection 2026.

ABSTRACT

BACKGROUND: Since their implementation in Spain, adherence of hospitals to Patient Blood Management (PBM) programs has been variable, potentially influencing transfusion practices. This study aimed to evaluate, in a real-world surgical setting, the frequency and appropriateness of red blood cell (RBC) transfusion.

METHODS: A prospective multicenter study in 43 Spanish hospitals including surgical patients. Transfusion appropriateness was evaluated using evidence-based criteria based on hemoglobin thresholds and clinical conditions such as cardiovascular disease, acute hemorrhage, or high comorbidity burden. Statistical analyses identified factors associated with transfusion practices.

RESULTS: The overall perioperative RBC transfusion rate was 9.7%, with the highest rates in cardiac (52.9%), vascular (17.9%), and orthopedic (12.3%) surgeries. RBC transfusion was associated with older patients with comorbidities, lower preoperative hemoglobin, higher ASA score and worse surgical meters and postoperative outcomes. Transfused patients showed significantly lower 60-day survival. Critically, 43% of transfusions were inappropriate, while transfusion omission (1.9%) may represent a clinical concern that warrants further investigation. Inappropriate transfusion was more frequent in older comorbidity patients according to Charlson Comorbidity Index in urgent surgery. In multivariable analysis, age was a factor associated with inappropriate transfusion, by cons, surgical blood loss was the main protective factor against inappropriate transfusion.

CONCLUSION: As far as we know, this is the first Spanish multicenter study evaluating transfusion appropriateness in surgical scenario. Despite a lower overall transfusion rate than international figures, nearly half of transfusions were inappropriate and transfusion omission, also represents a real clinical concern. Implementation of decision-support tools and strengthened PBM protocols are needed to address factors associated with inappropriate transfusion, such as age, and to optimize patient safety and resource use.

PMID:42221112 | PMC:PMC13216454 | DOI:10.3389/fmed.2026.1803092

Front Med (Lausanne). 2026 May 14;13:1848131. doi: 10.3389/fmed.2026.1848131. eCollection 2026.

ABSTRACT

INTRODUCTION: This study aimed to develop and validate a novel formula for calculating intraocular lens (IOL) refractive power based on Gaussian optics and thick-lens imaging.

METHODS: This study was conducted at Ningbo Traditional Chinese Medicine Hospital between October 2021 and October 2023. A total of 54 patients (84 eyes) with age-related cataracts (ARC) undergoing phacoemulsification and IOL implantation were included. The effective lens position was defined as (ACD + W × preoperative LT), where ACD is the anterior chamber depth and LT is the lens thickness. A new IOL power calculation formula was derived using stepwise multiple linear regression, incorporating key ocular parameters including axial length (adjusted for central corneal thickness), keratometry, and the effective lens position. The performance of the new formula was compared with five established formulas: Barrett Universal II, Haigis, Hoffer Q, SRK/T, and Holladay I. For each formula, we compared the mean and median predicted error (PE), the mean and median absolute predicted error (APE), and the proportions of eyes with within ±0.25 D, ±0.50 D, and ±1.00 D.

RESULTS: The newly developed formula demonstrated excellent bias control, with a median prediction error of 0.060 D that was not statistically different from zero (p = 0.480). In contrast, the Barrett Universal II (0.450 D, p = 0.006), Hoffer Q (0.280 D, p = 0.024), and SRK/T (0.515 D, p = 0.0004) showed significant hyperopic shifts. The mean error of the new formula (0.071 D) was significantly lower than that of Barrett Universal II, Hoffer Q, and Holladay I formulas (all p < 0.01) and comparable to that of the Haigis formula (p = 0.226). Its accuracy (mean absolute error, 0.461 D) was comparable to that of all other formulas. The new formula achieved the highest proportion of eyes within ±0.25 D (43.3%), outperforming all other formulas.

DISCUSSION: The proposed IOL calculation formula, which is based on a double-thick-lens imaging model, provides improved control of systematic bias and competitive predictive accuracy. This approach offers a promising framework for clinical applications of personal IOL power calculations.

PMID:42221106 | PMC:PMC13216231 | DOI:10.3389/fmed.2026.1848131

Front Med (Lausanne). 2026 May 15;13:1801925. doi: 10.3389/fmed.2026.1801925. eCollection 2026.

ABSTRACT

BACKGROUND: Early risk stratification is crucial for improving outcomes in critically ill patients with acute myocardial infarction (AMI). Albumin-derived composite indices hold promise as convenient and effective predictive tools, but their relative efficacy and clinical utility remain unclear.

METHODS: This two-cohort retrospective analysis utilized a derivation cohort from the MIMIC-IV public database and an external validation cohort from the ICU of Guizhou Medical University Affiliated Hospital. Six albumin-derived composite indices were evaluated. Statistical analyses employed Cox proportional hazards regression models to assess their association with mortality. Predictive performance was compared using the area under the receiver operating characteristic curve (AUC) and Delong’s test. A multivariate risk prediction model was developed based on key prognostic variables selected by multiple machine learning algorithms.

RESULTS: The study included 4,850 critically ill AMI patients (4,210 in the derivation cohort, 640 in the validation cohort). Multivariable-adjusted analysis identified the red cell distribution width to Albumin Ratio (RAR), Urea nitrogen to Albumin Ratio (UAR), and Lactate Dehydrogenase to Albumin Ratio (LDAR) as independent predictors of 28-day ICU mortality. Among these, LDAR demonstrated the strongest predictive ability, with an AUC of 0.702 in the derivation cohort, a finding robustly validated externally (AUC = 0.703). Subgroup analysis indicated consistent predictive value across most populations but revealed a significant interaction with hyperlipidemia. Incorporating LDAR into traditional critical illness scores (e.g., APACHE II, SOFA) significantly improved their predictive discrimination (all Delong’s test p < 0.05). A comprehensive model integrating 7 key variables (including LDAR, urea nitrogen, and lactate) selected by machine learning showed good and robust discriminative performance in both internal and external validation (AUCs of 0.767 and 0.735, respectively), significantly outperforming five traditional risk scores (all Delong’s test p < 0.05).

CONCLUSION: Among the six albumin-derived composite indices, LDAR offers the best independent and incremental predictive value for 28-day ICU mortality in critically ill AMI patients. Its interaction with hyperlipidemia suggests potential for targeted risk stratification. The machine learning model incorporating LDAR and other variables demonstrates robust performance, providing a promising tool for the early clinical identification of high-risk patients.

PMID:42221102 | PMC:PMC13219332 | DOI:10.3389/fmed.2026.1801925

Front Med (Lausanne). 2026 May 15;13:1823549. doi: 10.3389/fmed.2026.1823549. eCollection 2026.

ABSTRACT

OBJECTIVE: This study aims to determine the prevalence of uncontrolled asthma and its associated factors among adult asthmatic patients attending hospitals in Burao, Somaliland, in 2025.

METHOD: A cross-sectional study was conducted at four hospitals from May 1 to June 15, 2025, among 363 adult asthma patients selected using systematic random sampling. Data were collected using a structured, interviewer-administered questionnaire. Asthma diagnosis was confirmed by spirometry performed within 3 months before enrollment, demonstrating reversible airflow obstruction. Descriptive statistics summarized the sample characteristics. Variables with p < 0.25 in bivariable analysis were entered into multivariable logistic regression. Model fit was assessed using the Hosmer-Lemeshow test (p = 0.423), and multicollinearity was checked using the variance inflation factor (VIF < 2 for all variables). A p-value < 0.05 was considered statistically significant.

RESULT: Out of 382 adult asthma patients approached, 363 were included in the study, yielding a response rate of 98.4%. The mean age of participants was 47.6 ± 9.98 years, with 56.8% being male. Uncontrolled asthma was identified in 57.9% of the participants. Multivariable analysis revealed significant associations with poor medication adherence (AOR = 27.89; 95% CI: 7.75-100.30; p < 0.001), persistent asthma severity (AOR = 3.42; 95% CI: 1.27-9.23; p = 0.015), a family history of asthma (AOR = 27.51; 95% CI: 2.51-301.60; p < 0.001), biomass fuel use (AOR = 8.56; 95% CI: 3.08-23.78; p < 0.001), and occupational dust exposure (AOR = 68.65; 95% CI: 3.80-1239.19; p = 0.004). Due to small cell counts for dust exposure (n = 40 exposed) and family history (n = 113 with family history, including only 1 well-controlled case), these estimates have wide confidence intervals and should be interpreted cautiously. Exact logistic regression was performed as a sensitivity analysis, which confirmed the direction and statistical significance of these associations. The presence of air conditioning was also associated with increased odds (AOR = 6.26; 95% CI: 1.44-27.21; p = 0.014). Good asthma knowledge approached significance as a protective factor (AOR = 0.45; 95% CI: 0.20-1.04; p = 0.060).

CONCLUSION: The burden of uncontrolled asthma in Burao City is high, with modifiable factors such as medication adherence and follow-up care playing a vital role. Targeted interventions addressing these factors are crucial to improving asthma control in this setting.

PMID:42221096 | PMC:PMC13219360 | DOI:10.3389/fmed.2026.1823549

Front Med (Lausanne). 2026 May 15;13:1807805. doi: 10.3389/fmed.2026.1807805. eCollection 2026.

ABSTRACT

BACKGROUND: Vascular calcification (VC), characterized by calcium deposition in arterial walls, is a major risk factor for cardiovascular morbidity and mortality. While genome-wide association studies (GWAS) have identified susceptibility loci for specific vascular beds, such as coronary artery calcification (CAC) and abdominal aortic calcification (AAC), single-phenotype studies may overlook pleiotropic variants. This study aims to elucidate the shared genetic architecture of CAC and AAC and translate these findings into biological insights and potential therapeutic targets.

METHODS: We performed a multivariate genome-wide analysis integrating summary statistics for CAC and AAC from individuals of European ancestry. To prioritize candidate genes, we applied four complementary mapping strategies, including positional mapping, multivariate set-based association test, transcriptome-wide association study, and multi-marker analysis of genomic annotation. Findings were further characterized using tissue-specific expression profiling, Gene Ontology enrichment, and cell-type specificity analysis. Therapeutic potential and safety were subsequently evaluated using OpenTargets for druggability assessment and phenome-wide association studies (PheWAS) to assess horizontal pleiotropy. Finally, experimental validation was conducted to verify the genetic findings.

RESULTS: The multivariate analysis identified seven genome-wide significant loci. Cross-referencing the four gene-mapping strategies highlighted a consensus set of robust candidate genes, with CDKN2B supported by all methods, and strong multi-method support for ADAMTS7, PHACTR1, and MORF4L1. Pathway analysis identified lipid homeostasis and cell cycle regulation as key functional modules. Cell-type specificity analysis demonstrated that candidate genes were enriched in endothelial cells. Druggability assessments identified HDAC9 as a target for approved drugs potentially repurposed for VC, while PheWAS results suggested a predicted lack of severe genetic pleiotropy for most candidates, with the notable exception of CDKN2A, which showed associations with neoplasms. Quantitative real-time PCR confirmed significantly altered expression of most candidate genes, including ADAMTS7, CDKN2A, CDKN2B, CXCL12, FHL5, HDAC9, MORF4L1, PDGFD, and PHACTR1, in the experimental group.

CONCLUSION: This study demonstrates that CAC and AAC share a substantial genetic basis, reinforcing the concept of VC as a systemic pathological process driven by common mechanisms. By rigorously prioritizing candidate genes and mapping them to specific cell types, we provide a comprehensive genetic map of VC and highlight potentially safe targets for future therapeutic development.

PMID:42221083 | PMC:PMC13218890 | DOI:10.3389/fmed.2026.1807805

Front Med (Lausanne). 2026 May 15;13:1793058. doi: 10.3389/fmed.2026.1793058. eCollection 2026.

ABSTRACT

INTRODUCTION: The clinical-epidemiological profile of Chagas disease (CD) has changed over the past decades, leading to decreased levels of physical activity (PA), which may be associated with changes in markers of subclinical atherosclerosis. This study aimed to investigate the association between PA levels and carotid intima-media thickness (CIMT), the presence of carotid atherosclerotic plaque (CAP), and the amount of epicardial adipose tissue (EAT) in patients with chronic CD.

METHODS: This cross-sectional study included patients with chronic CD. The PA level was determined using the short version of the International Physical Activity Questionnaire (IPAQ-short). CIMT and CAP were assessed using Doppler ultrasound of the carotid arteries. The amount of EAT was assessed using transthoracic echocardiography. Linear and binomial logistic regression models were used.

RESULTS: The median age of the 349 participants was 62 years (54.0-69.0), 56.5% were women, and 79.5% were non-white, with 67.6% having <9 years of schooling. The most common clinical form of CD was the cardiac form without heart failure (HF) (53.9%). Median CIMT values were 0.65 mm (left) and 0.66 mm (right). CAP was present in 46.4% of participants, and the median EAT was 5.0 mm. No statistically significant association was observed between PA levels and CIMT, the presence of CAP, or the amount of EAT.

CONCLUSION: PA levels were not associated with the markers of subclinical atherosclerosis in individuals with chronic CD.

PMID:42221074 | PMC:PMC13219020 | DOI:10.3389/fmed.2026.1793058

Front Med (Lausanne). 2026 May 14;13:1786137. doi: 10.3389/fmed.2026.1786137. eCollection 2026.

ABSTRACT

OBJECTIVE: Fentanyl-induced cough (FIC) is a common adverse event during anesthesia induction with a high incidence and may result in serious clinical complications. Although our clinical observations suggest that oliceridine attenuates FIC, the available evidence remains limited. This study was designed to assess the prophylactic efficacy of oliceridine against FIC and to characterize its peri-induction safety profile.

METHODS: A total of 168 adult surgical patients with American Society of Anesthesiologists physical status I-III scheduled for general anesthesia were randomized to receive either oliceridine 2 mg (OF group) or normal saline (SF group) prior to fentanyl administration. The primary outcome was the incidence of cough within 1 min following fentanyl injection. Secondary outcomes included cough severity, vital sign changes, and the incidence of adverse events.

RESULTS: No patients in the OF group experienced FIC, compared to 58.33% in the SF group (p < 0.001). Cough severity in the SF group was classified as mild (17.86%), moderate (20.24%), and severe (20.24%). Vital signs did not show statistically significant changes from before to after oliceridine injection within each group, and no significant between-group differences were identified. The incidence of adverse events was low in both groups, with no significant between-group differences.

CONCLUSION: Pretreatment with 2 mg oliceridine effectively reduces the incidence of FIC without increasing significant additional risks, providing a rationale and safe approach for anesthetic induction.

CLINICAL TRIAL REGISTRATION: https://www.chictr.org.cn/showproj.html?proj=272947, identifier ChiCTR2500105221.

PMID:42221070 | PMC:PMC13215988 | DOI:10.3389/fmed.2026.1786137