Fulden Cantaş Türkiş

OBJECTIVES: Pediatric pain is frequently under-recognized and undertreated in prehospital emergency care despite guideline-based recommendations, and practice variation may persist even within standardized emergency medical services (EMS) systems. METHODS: We conducted a prospective, cross-sectional survey of paramedics working in a nationally standardized EMS system in Türkiye to identify latent attitudinal (self-reported) profiles related to pediatric prehospital analgesia and to examine whether profile membership differed by demographic or professional characteristics. Attitudinal items underwent exploratory dimensionality assessment and were then modeled using latent class analysis. RESULTS: Among 129 paramedics, two profiles were identified: proactive/system-oriented (57.4%) and conservative/low-engagement (42.6%). The proactive/system-oriented profile reported higher endorsement of decision-support tools (mobile drug-dosing applications and in-ambulance analgesia algorithms), more consistent self-reported post-analgesia vital sign monitoring, and greater self-reported use of non-pharmacologic adjuncts. The conservative/low-engagement profile reported lower intended engagement with these self-reported practices and greater reliance on perceived barriers. Profile membership was not associated with age, sex, years of EMS experience, or prior training. CONCLUSİON: Distinct attitudinal (self-reported) profiles regarding pediatric prehospital analgesia were identified; whether these profiles translate into objectively observed analgesia delivery requires validation using routine documentation and, where feasible, linked ED outcomes.

Hyponatremia is a common electrolyte disorder in the emergency department (ED), but its clinical burden is not evenly distributed across the sodium spectrum. We conducted a retrospective observational cohort study of adult ED encounters at a tertiary referral hospital between July 2019 and July 2024 to identify a data‐driven threshold for extreme hyponatremia and evaluate its association with 28‐day mortality. Encounters with simultaneous sodium and glucose measurements were included, and sodium values were corrected for hyperglycemia. Kernel density estimation identified a distinct local minimum defining distribution‐derived extreme hyponatremia (DDEH). Propensity score matching based on age and sex was used to compare DDEH with nonextreme hyponatremia and normonatremia. Among 227,027 eligible ED patients, density modeling identified an inflection point at 110 mmol/L. Sixty‐one patients (0.027%) met criteria for DDEH. In matched analyses ( n = 61 per group), 28‐day mortality was highest in the DDEH group (36.1%), compared with nonextreme hyponatremia (18.0%) and normonatremia (6.6%) ( p < 0.001). Kaplan–Meier analysis showed significantly reduced survival in patients with DDEH (log‐rank p < 0.001). A distribution‐derived sodium threshold of 110 mmol/L identified a rare but high‐risk phenotype of extreme hyponatremia in ED patients. This cohort‐derived boundary identified a rare extreme‐tail phenotype associated with systemic physiological vulnerability and increased short‐term mortality risk, but external validation is required before broader clinical implementation.

Aims: This study aimed to evaluate Artificial Intelligence (AI)-related knowledge, healthcare-oriented AI perspectives, and multidimensional attitudes toward AI among medical students and faculty members in a Turkish medical school.Methods: This analytical cross-sectional study was conducted during the 2025-2026 academic year at a faculty of medicine. A total of 467 participants, including 356 medical students and 111 faculty members, were included. Data were collected using an online survey. Participants completed questionnaires regarding sociodemographic characteristics and AI usage habits, together with the AI Knowledge Scale, AI Perspective Scale, and AI Attitude Scale consisting of hope, anxiety, and adaptation subdimensions. Statistical analyses included Mann-Whitney U test, ANOVA/Kruskal-Wallis tests, and multiple linear regression analysis.Results: Faculty members demonstrated significantly higher hope and overall attitude scores than medical students, although the corresponding effect sizes were small (rank-biserial correlations=-0.154 and -0.166, respectively). Male participants had higher self-reported AI knowledge, AI perspective, and hope scores. Participants using premium AI tools and those with longer AI use duration generally demonstrated higher self-reported AI knowledge scores and more favorable scores in several attitude subdimensions. In the multivariable model, medical student status was associated with a lower overall AI attitude score (B=-1.856; 95% CI: -3.594 to -0.119), whereas higher self-reported AI knowledge (B=0.156; 95% CI: 0.017 to 0.296) and AI perspective scores (B=0.140; 95% CI: 0.032 to 0.249) were associated with more favorable overall AI attitudes. In an age-adjusted sensitivity analysis, the associations involving self-reported AI knowledge and AI perspective remained significant, whereas the association with participant group was attenuated.Conclusion: More favorable overall AI attitude scores were independently associated with higher self-reported AI knowledge and more positive healthcare-oriented AI perspectives, whereas AI use duration and premium AI use showed no independent association. The assessment of hope, anxiety, and adaptation highlights the multidimensional nature of AI-related attitudes. These findings may help inform hypotheses and priorities for future educational research rather than provide direct evidence for the effectiveness of specific curricular interventions.

Background/objectives: Artificial intelligence-based clinical decision support systems have emerged as promising tools for improving early sepsis recognition in emergency departments. This study aimed to develop and evaluate interpretable machine learning models for identifying Sepsis-3-defined sepsis using routinely available admission-stage variables and to compare their performance with conventional biomarkers and the Systemic Inflammatory Response Syndrome (SIRS). Methods: This secondary analysis used a publicly available prospective cohort of 1,572 adult emergency department patients with suspected community-onset sepsis. The outcome was Sepsis-3-defined sepsis. Thirteen demographic, physiological, and laboratory admission variables were evaluated. Following a stratified 70:30 train–test split, missing data were handled using multivariate imputation by chained equations, and feature selection was performed using least absolute shrinkage and selection operator logistic regression. Penalized logistic regression (LR), Explainable Boosting Machine (EBM), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM) models were developed using five-fold cross-validation. Performance was evaluated by discrimination, classification, calibration, decision curve analysis, and explainability. Results: Among 1,572 adult patients, 560 met the Sepsis-3 criteria. In the internal holdout test set, AUROC values were 0.753 (LR), 0.763 (XGBoost), 0.767 (LightGBM), and 0.766 (EBM). Conventional biomarkers and SIRS showed lower discrimination (AUROC 0.599–0.693), whereas a biomarker-only logistic regression model achieved an AUROC of 0.710. No significant AUROC differences were observed among the machine learning models. LightGBM achieved the numerically highest AUROC and accuracy, whereas EBM achieved the highest sensitivity (0.845). Age, procalcitonin, oxygen saturation, respiratory rate, and neutrophil-to-lymphocyte ratio were the most influential predictors. Conclusions: Interpretable machine learning models based on routinely available emergency department admission variables provided moderate discrimination, clinically relevant sensitivity, and transparent predictor effects for identifying Sepsis-3-defined sepsis. Compared with conventional biomarkers, SIRS, and a biomarker-only logistic regression model, the machine learning models generally showed higher discrimination, although the magnitude and statistical significance of these differences varied across comparators. External validation is required before clinical implementation.

Objectives: Emergency medicine (EM) internships offer high-acuity, real-time learning environments that can significantly influence the academic self-efficacy and clinical confidence of final-year medical students. This study aimed to evaluate the effect of a two-month structured EM internship on students’ academic self-efficacy and perceived competence in core EM domains.Methods: This prospective observational pre-post study was conducted between September 1, 2022, and August 31, 2023, in a university-affiliated tertiary care hospital in Türkiye. Final-year medical students, who are referred to as medical interns within the Turkish medical education system, completed the Medical Achievement Self-Efficacy Scale (MASS) and a 30-item self-assessment aligned with the National Core Medical Curriculum before and after the internship. Statistical analyses included nonparametric tests and Cronbach’s alpha for reliability.Results: Of 126 eligible final-year medical students, 99 completed the pre-internship assessment and 66 completed both pre- and post-internship assessments. MASS scores significantly improved (pre: 68.0 [18–90] vs. post: 77.0 [32–90]; Z = –4.482, p < 0.001). Self-efficacy increased in all five EM domains, particularly in patient assessment and triage. Limited gains were observed in burnout recognition and complex procedures.Conclusions: Structured EM internships may enhance self-efficacy and clinical readiness among final-year medical students. Curriculum development should include immersive EM rotations and targeted support for stress resilience and procedural skills.

Surgical informed consent documents represent an important interface between anatomical knowledge and patient‐facing clinical information; however, no structured approach currently exists for evaluating how anatomy is represented and communicated within these materials. This study aimed to develop and apply an exploratory Anatomical Clarity Framework for evaluating anatomical communication in surgical informed consent documents. A cross‐sectional document analysis was conducted on 10 General Surgery informed consent templates (five emergency and five elective/major procedures) obtained from a tertiary‐care hospital in Türkiye. Readability was assessed using the Ateşman, Bezirci–Yılmaz, and Çetinkaya–Uzun indices. Document characteristics and anatomical communication features were evaluated using a structured assessment framework, and inter‐rater agreement was examined using intraclass correlation coefficients. Anatomical information was present in all analyzed documents. Anatomical structures and procedural targets were generally identified; however, substantial variability was observed in explanatory features supporting anatomical communication, including anatomy–risk linkages, plain‐language anatomical explanations, and visual representations. A recurring pattern across documents was the combination of high anatomical specificity with limited explanatory support and minimal visual representation. Only one document included a schematic anatomical illustration. All documents demonstrated high readability demands, corresponding to difficult, very difficult, or academic‐level categories across the applied readability indices. Readability, document design, and anatomical communication described complementary rather than interchangeable characteristics of the analyzed documents. Communicative and anatomical domain scores appeared broadly similar across emergency and elective templates. The Anatomical Clarity Framework demonstrated good inter‐rater agreement, with the strongest agreement observed in the Anatomical Clarity domain (intraclass correlation coefficient = 0.858). These findings suggest that the presence of anatomical information alone does not necessarily correspond to a high level of anatomical clarity. The Anatomical Clarity Framework provides a structured approach for evaluating how anatomical information is represented, contextualized, and communicated within surgical consent documents and may complement conventional readability assessments in clinical anatomy research. Because this study represents an initial methodological application of the framework, the findings should be interpreted as preliminary and require further validation in broader clinical settings.

We examined the validity and reliability of the DSM-5-Based Childhood ADHD Self Report Scale (DSM-5-CASRS), a self-report questionnaire developed to assess Attention-Deficit/Hyperactivity Disorder (ADHD) symptoms in elementary school children. Two-hundred twelve elementary school second-grade children (130 children with ADHD and 82 non-ADHD) derived from an epidemiological sample completed our scale. Parents completed Turgay DSM-IV Disruptive Behavior Disorders Rating Scale. Children completed the DSM-5-CASRS. We also applied a semi-structured psychiatric interview to all the participants and their parents. The DSM-5-CASRS demonstrated satisfactory internal consistency (McDonald's ω = 0.92) and strong item-total correlations (0.51 - 0.74). Confirmatory Factor Analysis confirmed a two-factor structure aligning with DSM-5 symptom domains. The scale showed significant and moderate positive correlations with parent-reported ADHD symptoms (r = .44, p < .001), supporting its convergent validity. Machine learning classification using a Random Forest model yielded high sensitivity (89.7%) and moderate specificity (62.5%), suggesting the scale effectively identifies ADHD cases. Findings indicate that 8- and 9-year-old children can reliably report their ADHD symptoms, challenging previous assumptions regarding young children's self-awareness. The DSM-5-CASRS fills a critical gap in ADHD assessment by providing a structured and developmentally appropriate self-report tool. Future research should expand validation across different age groups and cultural contexts while exploring alternative response formats to enhance usability.

High‐dimensional datasets with missing values present substantial challenges for statistical modeling and inference because conventional multiple imputation (MI) methods may become unstable or nonidentifiable under severe dimensional imbalance. Regularized regression techniques, including lasso, ridge, and related Bayesian formulations, offer a promising way to improve feasibility by enabling variable selection and stabilizing estimation in high‐dimensional settings. This overview provides a structured synthesis of regularized regression‐based MI methods, organized into direct and indirect integration strategies, Bayesian sparsity‐inducing approaches, optimization‐based and sequential penalization methods, and post‐imputation resampling or adjustment techniques. For each category, we summarize the underlying methodological logic, review empirical findings from simulation studies, and highlight practical considerations for implementation. Particular attention is given to methods developed for missing not at random (MNAR) and to the inferential challenges that arise after variable selection. Overall, this overview aims to help researchers choose regularized imputation strategies that are methodologically appropriate, practically feasible, and aligned with the analytical goals of high‐dimensional incomplete‐data studies. This article is categorized under: Technologies > Machine Learning Algorithmic Development > Statistics