Acta anaesthesiologica ScandinavicaWei Wei, Michael Seltz Kristensen, Wenxian Li, Anders Bo Nielsen, Lene Russell, Lars Konge, Yuan Han
BACKGROUND: Flexible laryngeal mask airway (fLMA) management is an important but technically demanding airway skill, particularly for novice physicians. Distance hands-on training may provide a scalable alternative when access to face-to-face simulation is limited, but evidence demonstrating transfer of such training to clinical performance in patients remains scarce. This trial aims to compare distance hands-on training with standardized video instruction for novice physicians learning fLMA management. METHODS: This is a prospective, multicenter, parallel-group, rater-blinded, superiority randomized controlled trial conducted at three tertiary hospitals in China. Thirty novice physician participants will be randomized 1:1 to either distance hands-on training or video instruction. Both groups will receive identical standardized theoretical materials, a procedural demonstration video, and a written checklist. The distance hands-on training group will additionally receive remotely supervised deliberate practice using an airway task trainer with real-time instructor feedback. After training, each physician participant will perform fLMA management in two adult patient participants under direct supervision by experienced anesthesiologists. Clinical performance will be video-recorded and independently assessed by two blinded raters using the Flexible LMA Airway Management Global Rating Scale, for which validity evidence was established in a preceding study. Secondary outcomes include procedural success, insertion time, ventilation parameters, fLMA positioning, procedure-related complications, learning progression, and physician participant satisfaction. DISCUSSION: This trial will evaluate whether structured distance hands-on training improves clinical procedural performance beyond standardized video instruction. By combining randomized educational intervention, multicenter clinical assessment, blinded video-based rating, and patient-based outcome evaluation, the study aims to provide evidence on the transfer of distance simulation training to clinical airway management. TRIAL REGISTRATION: Chinese Clinical Trial Registry (ChiCTR): ChiCTR2600121867.
Acta anaesthesiologica ScandinavicaConstance Thyra Giertz Christiansen, Mathilde Nellemann, Moritz Kilian German Denneborg, Sofie Amalie Bosholdt, Janus Christian Jakobsen, Rasmus Tingkær Hessel…
BACKGROUND: Pre-anaesthetic assessment is essential for safe perioperative care but is resource-intensive and applied variably in clinical practice. Increasing surgical demand and workforce constraints have prompted interest in alternative models, including digital approaches and AI-supported clinical decision support tools. However, limited knowledge exists on clinicians' perspectives, values, and perceived information requirements in relation to pre-anaesthetic assessment, including which information they consider essential prior to anaesthesia. METHODS: This international, cross-sectional survey aims to explore clinicians' perspectives on pre-anaesthetic assessment practices and their attitudes towards AI-supported clinical decision support tools. The survey will be conducted in two phases, beginning with Nordic centres before broader international expansion, using two linked instruments: an organisational survey completed once by a local site investigator at each site, and an individual survey completed by anaesthesia personnel recruited via site-based convenience sampling. The primary outcome is clinicians' perspectives on pre-anaesthetic assessment, comprising perceived importance, information not readily available from chart review considered essential prior to anaesthesia, and use in subsequent anaesthetic management; secondary outcomes include airway assessment practices and attitudes towards AI-supported tools, including willingness to adopt them. Exploratory outcomes include organisational workflows and time-use in pre-anaesthetic assessment. Responses will be summarised descriptively across clinician and organisational subgroups. CONCLUSION: This survey will provide an international overview of clinicians' perspectives on pre-anaesthetic assessment and its role in clinical decision-making. The findings may contribute to the development of more efficient, stratified, and clinically aligned assessment models while recognising that implementation of digital and AI-supported approaches will also depend on organisational, technical, ethical, legal, and patient-related factors.
Medical science monitor : international medical journal of experimental and clinical researchMarcin Rabijewski, Anna Kurek, Marcin Muża, Sławomir Paszek, Zofia Frączek, Michał Rabijewski
Adverse medical events remain a serious concern worldwide; deficiencies in human factors contribute to a large proportion of medical incidents and preventable patient harm. In anesthesiology, where high-risk and complex clinical events are routine, non-technical skills (NTS)-including communication, teamwork, situational awareness, and decision-making-play a critical role in preventing errors during high-acuity emergencies. Furthermore, Crisis Resource Management (CRM) provides a structured framework for enhancing team performance and coordination during medical crises by emphasizing effective leadership, task management, and the optimal allocation of resources within a multidisciplinary team. Simulation-based training is central to teaching and assessing NTS; validated tools such as the Anesthetists' Non-Technical Skills (ANTS) system allow objective evaluation of team performance and behaviors across various clinical scenarios. This modality is particularly effective because it incorporates structured debriefing, enabling practitioners to reflect on their clinical performance in a safe, risk-free environment. Evidence shows that CRM- and NTS-focused simulation training improves crisis response, adherence to protocols, and clinical outcomes. Despite widespread international adoption, CRM and NTS training in Poland remains limited because of institutional and systemic barriers. Implementation of structured CRM and NTS curricula in Poland could enhance patient safety, foster effective teamwork, and reduce medical errors, particularly in high-risk clinical settings, thus aligning Polish medical education with international standards. This narrative review synthesizes global evidence regarding the effectiveness of CRM and NTS training, highlights the benefits of simulation-based training and its impact on long-term skill decay, and identifies gaps in Polish residency programs.
The Journal of international medical researchRan An, Bingbing Meng, Xun Lu
Artificial intelligence is increasingly being investigated as a tool for supporting personalized perioperative care. This narrative review, guided by the Scale for the Assessment of Narrative Review Articles principles, examines clinically relevant applications of artificial intelligence and related automation in anesthesiology, with emphasis on three linked functions: risk prediction, clinical decision support, and supervised automation. We performed a structured search of PubMed, Web of Science, and Scopus databases and identified English-language literature published from January 2010 through June 2026. We prioritized clinical studies, guidelines, systematic reviews, and representative methodological studies. A central distinction of this review is that machine-learning prediction, rule-based decision support, conventional feedback control, and genuinely adaptive artificial intelligence-enabled control are not treated as equivalent technologies. Current evidence is strongest for selected prediction tasks and supervised systems that improve process measures such as time within a physiological target range. However, evidence that these technologies improve patient-centered outcomes remains less consistent. Model discrimination alone is insufficient; calibration, actionable thresholds, false-positive burden, external validation, clinical utility, dataset shift, missing data, and workflow consequences must also be considered. The proposed three-level framework is intended to provide anesthesiologists and perioperative teams with a practical way to connect technical capability with bedside action. Near-term implementation should prioritize interpretable prediction and low-burden decision support, whereas closed-loop systems should remain under clinician supervision. Future studies should emphasize prospective multicenter validation, clinically meaningful endpoints, transparent reporting, and post-deployment monitoring rather than technical performance alone.
A&A practiceMatthew D Caldwell, Norah N Naughton, Sachin Kheterpal, Xinyi Zhao, Graciela Mentz, Lara Zisblatt, INSPECT Collaborators
Algorithmic analysis of electronic health record (EHR) data offers an objective, scalable approach to quantifying resident clinical experience. We tested whether Multicenter Perioperative Outcomes Group registry data could be used to accurately determine which pseudonymized IDs correspond to graduated anesthesiology residents. Algorithmic determination was compared with rosters of 338 graduated residents from seven residency programs. The algorithm demonstrated 91% sensitivity and 97% positive predictive value for determining which IDs are of graduated residents. The algorithm was applied across 29 institutions to analyze resident general anesthesia cases, demonstrating the feasibility of multicenter EHR registry data use in graduate medical education.
JMIR formative researchWilliam Gatenby, Sumeet R Gopwani, Glenn E Woodworth, McKenzie Hollon, Stacey Kaltman
BACKGROUND: Precision medical education (PME) is an emerging concept that aims to improve physician education by tailoring curricula specifically to each learner. The basic framework behind PME requires collecting data on each individual learner, identifying gaps in knowledge, and then applying direct educational interventions. However, logistical difficulties in the delivery of such interventions hinder the implementation of PME in existing academic programs. OBJECTIVE: We developed the Personalized Automated Reading Delivery System (PARDS) software application to support a potential PME intervention, aiming to increase the volume of relevant scientific literature read by resident physicians. The PARDS analyzes surgical case lists for the following day, selects relevant scientific articles according to predetermined keywords, and delivers each article via email to an anesthesiology resident physician. This technology may assist in the application of personalized educational interventions, and we evaluated the feasibility of implementing the PARDS in a population of medical learners. METHODS: From July 2018 to December 2018, a pilot prospective crossover study was performed with N=10 postgraduate year 2 (PGY2) anesthesiology residents in an academic medical center (response rate 10/10, 100%). Residents were randomly assigned to the PARDS intervention group for 2 of the 4 months of the study period. The primary outcomes were organizational viability (evaluated by successful delivery of selected articles by the PARDS, required faculty hours, and total expenses) and acceptability in the study population (assessed by the number of reported weekly reading hours). An analysis using a linear mixed model for repeating measures was performed to compare weekly reading hours between intervention and nonintervention conditions. RESULTS: Total reported faculty time involvement for the study period was 24 hours. The cost estimate for implementation ranged from US $2360 to US $8360. The mean difference of recorded reading hours by residents while using the PARDS vs not using the PARDS for months 1, 2, 3, and 4 of the study was 2.56 (95% CI -0.03 to 5.15; P=.053), 0.44 (95% CI -2.15 to 3.03; P=.74), 0.99 (95% CI -1.38 to 3.36; P=.41), and 3.7 (95% CI 1.34 to 6.08; P=.002), respectively. CONCLUSIONS: The PARDS was successfully implemented in an academic medical center. Calculated costs indicate that the PARDS is an inexpensive intervention compared to contemporary alternatives. The effect on resident reading requires further research.
BACKGROUND: Taiwan's declining birthrate has reduced obstetric case volume and may threaten obstetric anesthesia training. We examined training experiences, self-perceived competencies, and educational needs in Taiwan to inform feasible curricular strategies. METHODS: We conducted a cross-sectional nationwide anonymous online survey of anesthesiology residents and attending anesthesiologists in Taiwan (October 1-December 31, 2024). The 37-item questionnaire covered demographics, self-assessed competencies and confidence, satisfaction with current training, and future training preferences, with open-ended items on barriers to additional training. Descriptive statistics, independent-samples t tests, one-way ANOVA, Pearson correlations, and linear regression were performed. RESULTS: A total of 145 valid responses were analyzed (112 attendings, 33 residents), yielding an overall response rate of 8.3% (attendings 7.8%, residents 10.4%). Self-assessed competence increased with greater clinical seniority and longer obstetric anesthesia training duration. Highest-rated competencies were neuraxial anesthesia for cesarean delivery (mean 7.93) and airway management during general anesthesia in obstetric patients (7.82), whereas confidence was lower for managing cardiac parturients (6.21) and neonatal resuscitation (5.89). Attendings reported higher overall confidence in independently performing obstetric anesthesia than residents (8.00 vs 6.52, p < 0.001). Male respondents were significantly older than female respondents (mean 41.9 vs. 35.4 years, p < 0.001). Correspondingly, male respondents reported higher confidence than female respondents for independent practice (8.05 vs 7.08, p < 0.001) and for managing cardiac parturients (6.63 vs 5.59, p = 0.001). In multivariable regression, confidence in managing cardiac parturients (β = 0.61, 95% CI 0.217-0.997; p = 0.002) and competence in neuraxial anesthesia for cesarean delivery (β = 1.214, 95% CI 0.190-2.238; p = 0.020) independently predicted overall confidence. CONCLUSIONS: Obstetric anesthesia competence and confidence were associated with training exposure, yet gaps persisted in high-risk and emergency domains and differed by gender. Programs should prioritize simulation-based training for rare critical events and develop inter-institutional collaborations to strengthen competency acquisition amid declining obstetric volumes.
PloS oneMax Y Lu, Alexandra M Hajduk, Jeph Herrin, Mytien Nguyen, Dowin Boatright, Sarwat I Chaudhry
BACKGROUND: Physician workforce diversity improves access to care, outcomes, and workforce productivity. Yet, women and sexual minorities remain underrepresented in anesthesiology, while representation for low-income trainees remains poorly characterized. How interest in anesthesiology evolves among these groups during medical school remains unclear. METHODS: This retrospective cohort study of U.S. MD-granting medical school matriculants from 2014-2017 used Association of American Medical Colleges data on self-reported sociodemographic characteristics and anesthesiology interest. Students were grouped by interest evolution from matriculation to graduation: never interested (no-no), cultivated (no-yes), lost (yes-no), sustained (yes-yes). Chi-square tests characterized sociodemographic differences across interest paths. Firth logistic regression models assessed odds of cultivated vs never interested and sustained vs lost paths across sociodemographic groups. RESULTS: Among 39,910 students in the cohort, 295 (0.7%) sustained, 2,007 (5.0%) cultivated, 494 (1.2%) lost interest, and 37,114 (93.0%) were never interested in anesthesiology. Women were less likely than men to have initial interest (1.5% vs 2.5%, p < .001), whereas initial interest among low-income (2.1% vs 1.9%, p = .087) and sexual minority (1.8% vs 2.0%, p = .505) students did not significantly differ from majority counterparts. Sustained interest did not differ significantly between women and men (OR: 0.85, 95% CI: 0.63-1.14), according to low-income (OR: 0.89, 95% CI: 0.65-1.22), or sexual minority status (OR: 0.77, 95% CI: 0.41-1.45). Women (OR: 0.50, 95% CI: 0.45-0.55) and sexual minority students (OR: 0.79, 95% CI: 0.65-0.96) were less likely to cultivate interest than men and heterosexual students, respectively. Low-income students had greater (OR: 1.27, 95% CI: 1.16-1.40) odds of cultivating interest compared to non-low-income peers. CONCLUSIONS: Women's underrepresentation in anesthesiology is associated with lower initial and cultivated interest. Sexual minority students also demonstrated lower cultivated interest. These disparities identify time points in training where interest diverges by demographic group and may help prioritize where future research - including studies of targeted interventions - is most needed.
Canadian journal of anaesthesia = Journal canadien d'anesthesieWillem J M Schellekens, Lisanne A H M De Nijs, Lorenz Theiler, Susan A Braithwaite, Paul Blankman
PURPOSE: Hypoxemia is an important complication of general anesthesia and procedural sedation and is associated with cardiopulmonary complications, increased hospital length of stay, and increased mortality. Through this Continuing Professional Development module, we aim to guide anesthesiologists in the safe use of high-flow nasal oxygen (HFNO). It examines the potential advantages and limitations of HFNO across the spectrum of anesthesiology practice, from procedural sedation to airway management, highlighting its role in enhancing patient safety. PRINCIPAL FINDINGS: High-flow nasal oxygen is a noninvasive respiratory support modality capable of delivering high flow rates and oxygen concentrations to patients during spontaneous breathing and apneic periods. Its efficacy depends on airway patency and on whether a patient's mouth is open or closed. Although HFNO can improve oxygenation and potentially prolong safe apnea times during airway management, clinicians must be aware of risks, such as progressive hypercapnia during prolonged apneic periods. CONCLUSIONS: Key advantages of HFNO include ease of application, the delivery of high inspired oxygen fractions and flow rates, and the potential generation of continuous positive airway pressure. For specific patient populations, HFNO can enhance respiratory support and safety throughout diverse perioperative situations.
AnesthesiologyAnjali K Ghoshal, Robert S Kaplan, Angela M Bader, David L Hepner
Medicare's Transforming Episode Accountability Model (TEAM) represents a significant shift in surgical care reimbursement. TEAM's bundled payment introduces episode-level accountability for cost and outcomes, replacing fee-for-service for discrete activities. TEAM's risk-sharing payments will reshape the clinical and financial environment in which anesthesiologists practice, as their reimbursement is included within the bundle. Their role in perioperative decision-making can strongly influence the performance of hospitals operating under TEAM's risk-sharing arrangements. This article outlines TEAM's structure and examines how episode-based payment models elevate aspects of anesthesiology practice that have traditionally been undervalued under fee-for-service reimbursement. The article also describes the expanded roles for anesthesiologists to prevent complications and to optimize perioperative care. Finally, the article discusses the growing need for anesthesiologists to define, measure, and report outcomes that reflect their contribution to patient recovery and resource utilization. As payment models evolve, aligning anesthesiology practice with episode-level metrics will demonstrate its value for delivering accountable surgical care.
The Journal of international medical researchYu Zhu, Daoqing Xie, Renrui Liang, Jian-Jun Yang, Xueke Du, Cheng-Mao Zhou
This narrative review synthesizes evidence published from 2020 to 2026 on the implementation and educational impact of large language models in anesthesiology training, following the Scale for the Assessment of Narrative Review Articles guidelines. We outline core technical attributes of large language models and identify four validated anesthesiology-specific use cases: standardized learning resource generation, clinical scenario simulation, personalized remediation of knowledge gaps, and automated assessment tool development. Current evidence suggests that these applications improve trainee knowledge scores and reduce faculty workload. Key deployment barriers include the risk of hallucination in high-stakes anesthesia content and potential overreliance on large language models, which may impair independent clinical reasoning. We propose targeted mitigation strategies and a forward-looking research agenda for structured large language model integration. Our analysis confirms that large language models are high-value enabling tools for anesthesiology education and require intentional, guideline-aligned integration to maximize synergies.
Minerva anestesiologicaLinus H Kutup, Eike S Debus, Reinhart T Grundmann
BACKGROUND: The present study aimed to clarify how English-language publication activity in anesthesiology, intensive care medicine, and pain medicine has evolved in publishing countries over a 10-year period, considering not only publication volume but also scientific impact and shifting research methodologies. METHODS: A bibliometric analysis of 20,741 publications indexed in PubMed for the time periods 2008/09 and 2018/19 was performed across 30 journals. Journals were selected based on their impact factor (IF) in 2019. Publications were analyzed for total number, cumulative impact factors (CIF) - representing the sum of individual journal impact factors multiplied by the number of publications in each journal - study type, country of origin, and country output per million inhabitants. RESULTS: From 2008/09 to 2018/19, total publication volume declined by 6.9% from 10,740 to 10,001, while CIF increased by 25.4% from 42,441.2 to 53,219.1. Clinical (Observational) studies were most frequent; meta-analyses and registry studies rose markedly (216.1% and 125.3%, respectively), while experimental studies and randomized controlled trials (RCTs) declined (-51.9% and -29.7%, respectively). The United States led in both publication count (2008/09: 3,267; 2018/19: 4,250) and CIF (2008/09: 14,256.4; 2018/19: 24,421.9). When adjusted to population, smaller countries including Denmark (2008/09: 31.1; 2018/19: 80.7), Switzerland (2008/09: 23.6; 2018/19: 42.1), and the Netherlands (2008/2009: 26.3; 2018/2019: 34.9) ranked highest. Germany - the only leading country showing declining publication output at 7.0% - demonstrated a concerning trend, while China (+83.6%) and Canada (+69.4%) showed exceptional growth. CONCLUSIONS: This study identifies a decline in publication volume alongside increased citation activity in high impact anesthesiology and intensive care medicine journals, largely driven by the dominance of clinical studies and a growing prevalence of meta-analyses and registry studies, suggesting a shift in research priorities. Reduced publication numbers likely result from increasing journal selectivity and redistribution of research output toward open-access and non-specialty journals. The country ranking is dominated by the USA and other English-speaking nations, which can be attributed to the English-language literature analyzed. However, smaller nations like Denmark and Switzerland lead in population-based ranking, revealing critical insights into research infrastructure and institutional support.
Acta anaesthesiologica ScandinavicaWei Wei, Michael Seltz Kristensen, Wenxian Li, Anders Bo Nielsen, Yuan Han, Lars Konge, Lene Russell
BACKGROUND: Airway management is a core clinical competency, yet consistent and patient-safe hands-on training opportunities remain limited. Distance simulation, in which learners and facilitators are in different physical locations, may expand access to technical skills training. However, which airway management procedures are most suitable for this format remains unclear. METHODS: We conducted a three-round international Delphi study to identify and prioritize airway management procedures suitable for distance simulation. Fifty-four airway experts were invited, of whom 46 from 20 countries participated. In Round 1, experts proposed candidate procedures via free-text responses. In Round 2, procedures were prioritized using the Copenhagen Academy for Medical Education and Simulation Needs Assessment Formula, incorporating frequency, proportion of operators requiring competence, clinical impact, and feasibility for distance simulation. In Round 3, procedures achieving ≥ 70% agreement were retained. RESULTS: Response rates were 85%, 89%, and 95% across the three rounds. The 464 free-text responses were consolidated into 54 procedures, of which 22 reached ≥ 70% consensus in the final round. Highly ranked procedures included direct laryngoscopic intubation, video laryngoscopic intubation, face mask ventilation, and laryngeal mask airway insertion. Emergency front-of-neck access with scalpel and bougie also achieved consensus. Rankings between Rounds 2 and 3 were strongly correlated (Spearman's ρ = 0.78). CONCLUSIONS: This international Delphi study identified a prioritized set of airway management procedures suitable for distance simulation, including both high-frequency essential and low-frequency critical procedures. These findings support the development of standardized distance simulation curricula. EDITORIAL COMMENT: This Delphi process report presents an international group's expert views on process for how advanced airway management training in a simulator environment can be conducted where the instructor by necessity is in a remote location.
JMIR formative researchStefan Andrei, Thibault Giet, Alexis Belouard, Mihai Stefan, Mihai Popescu, Sébastien Tanaka, Philippe Montravers, Aurélie Gouel
BACKGROUND: Large language models (LLMs) have shown promising performance on medical examinations across specialties. However, comparative evaluations of current-generation LLMs across multiple European anesthesiology examinations, alongside structured assessment of hallucinations vs question-related confusion, remain lacking. OBJECTIVE: This study aimed to compare the performance of 4 state-of-the-art LLMs on anesthesiology and intensive medicine examination questions and assess their hallucination rates. METHODS: This computational comparative study analyzed 437 multiple-choice questions (1748 queries) from 3 sources: nurse anesthetist school examinations (infirmier anesthésiste diplômé d'État [registered nurse anesthetist]; n=100, 22.9%), European Diploma in Anaesthesiology and Intensive Care (EDAIC; n=219, 50.1%), and EDAIC On-Line Assessment (n=118, 27.0%). Each question was submitted to 4 LLMs (Claude Sonnet 4.5, Gemini 2.5 Pro, GPT-5, and Grok 4) using standardized prompts via default web interface settings. Responses were evaluated through structured consensus review by 2 examiners for accuracy, hallucinations, and question-related confusion. Statistical analysis included Friedman and Wilcoxon signed-rank tests with Holm-Bonferroni correction, the Cochran Q test, and generalized estimating equations. RESULTS: Average success rates ranged from 86% (SD 18%) to 94% (SD 10%) across LLMs and examination types, exceeding the EDAIC part I passing threshold, representing substantial improvement over previously reported GPT-3.5 performance. For the EDAIC, overall intermodel differences were significant (Friedman χ23=13.9; P=.003; W=0.02), with Gemini outperforming GPT-5 as the only pairwise difference. Hallucination rates ranged from 11% (11/100) to 20.1% (44/219) without significant intermodel differences. All models exceeded the EDAIC passing threshold. CONCLUSIONS: Current-generation LLMs demonstrated consistently high performance across multiple European anesthesiology examinations but continue to produce clinically relevant hallucinations, supporting their role as supervised educational tools rather than autonomous learning resources. These findings underscore the need for structured integration frameworks and systematic verification when deploying LLMs as learning tools in medical education.
The Journal of international medical researchMuzeyyen Beldagli, Ahmet Ozan Aydin, Mehmet Alperen Avci, Selma Kahyaoglu, Taner Tunc, Yavuz Yigit, Serkan Tulgar
ObjectiveTo evaluate the demographic and hierarchical representation biases in artificial intelligence-generated depictions of healthcare professionals in anesthesiology, pain medicine, and intensive care, with a focus on sex, age group, skin tone, ethnicity, and professional role hierarchy.MethodsThis cross-sectional comparative study analyzed 4400 artificial intelligence-generated images produced by 4 text-to-image models (DALL·E 3, Midjourney, Leonardo AI, and Gemini). Ten professional roles in anesthesiology, pain medicine, and intensive care, ranging from trainees to department heads, were evaluated. Two independent anesthesiologists with clinical experience in anesthesiology and intensive care assessed each image using a structured digital evaluation form, recording sex, age group, skin tone, ethnicity, and professional role hierarchy. Statistical comparisons were performed using chi-square test with Bonferroni-adjusted post hoc proportion analyses.ResultsMale representation predominated across roles and models (mean: 68.5%), with leadership positions showing the highest male proportion (up to 90.0%). Light skin tones (mean: 70.6%) and Caucasian ethnicity (mean: 68.0%) were the most commonly depicted categories. Younger individuals (age: <40 years) were overrepresented (mean: 53.1%), whereas individuals aged >60 years were rarely depicted (<3.0%) and appeared mainly in leadership roles. Significant differences in sex representation were observed in 9 out of 10 professional roles across artificial intelligence models, with the exception of the anesthesia specialist role (p = 0.051).ConclusionsText-to-image artificial intelligence systems consistently reproduced results with demographic and hierarchical biases in depictions of healthcare professionals in anesthesiology, pain medicine, and intensive care. The predominance of male, light skin tone, and Caucasian ethnicity, particularly in senior roles, highlights the need for more diverse training datasets and greater transparency in the development of artificial intelligence systems.
Israel journal of health policy researchAriel Wimpfheimer, Yehuda Ginosar, Daniela Quesada, Shai Fein, Charles Weissman, Israel Anesthesiology Workforce Study Group
BACKGROUND: The smooth and efficient operation of a country's healthcare system is highly dependent on the size and specialty composition of its physician workforce. Therefore, it is incumbent upon a country's healthcare leadership to continually examine the workforce situation in each medical specialty, especially critical specialties like anesthesiology, so as to detect issues that could lead to shortfalls and surpluses and then apply effective corrective actions. The aim was to assess the accuracy of the different predictive methods and identify possible ways of improving their predictive capabilities. Our assumption was that the accuracy of workforce predictions is primarily affected by unforeseen and disruptive events. METHODS: Cross-sectional surveys of the Israeli anesthesiology workforce were performed in 2005 (n = 711 anesthesiologists) and 2021 (n = 1330 anesthesiologists). This permitted comparing needs and demand-based workforce forecasts made in 2005 with the actual 2021 workforce. Reasons for discrepancies between the 2005 forecasts and the reality in 2021 were sought. RESULTS: The 2005 needs-based projections underestimated the actual increase in anesthesiologists because the population grew at a rate faster than that estimated in 2005. Furthermore, there was an increase in the proportion of the population > 65 years old, an increase in life expectancy and a higher birth rate requiring a greater number of caesarean sections and epidural catheter insertions. Moreover, the 2011 union contract between the government and the Israel Medical Association increased anesthesiologist vacation time by two weeks per year. The demand-based projection underestimations were attributed to the significant increase in anesthetics performed outside operating rooms, such as, in cardiac catheterization laboratories and invasive radiology and endoscopy suites. Comparing anesthesiologist workforce projections made in 2005 to the workforce realities 16 years later revealed that many unpredictable (greater population growth) and disruptive (union contract) factors that ensued during that time period led to underestimating future workforce needs. CONCLUSIONS: Forecasting the future Israeli physician anesthesiologist workforce is an endeavor fraught with unpredictable challenges. To facilitate frequently updated forecasts of future workforce needs, healthcare systems require robust, long-term forecasting capabilities that use near real-time data obtained from the ubiquitous electronic medical and hospital record systems. These data should then be entered into sophisticated data analytic tools, rather than simple needs and demand-based methods. Such up-to-date forecasts should help healthcare system leaders and policy makers make more informed workforce decisions.
Current opinion in anaesthesiologyAstri M V Luoma, L Jane Easdown
PURPOSE OF REVIEW: This article aimed to examine the value of education and training within perioperative neuroscience and its development for supporting the delivery of high-quality, safe patient care. RECENT FINDINGS: Perioperative neuroscience is a rapidly evolving field in anesthesiology, with current research demonstrating the potential for neurological injury in patients undergoing nonneurological and neurological surgery. Expertise is essential to mitigate risk. Current evidence in perioperative neuroscience suggests a patient benefit with subspecialty training. Advanced training programs have continued to evolve with the development of global accreditation programs for fellowship training. SUMMARY: Advanced training within perioperative neuroscience is necessary to improve healthcare outcomes. Accreditation of training programs offers standardization and quality assurance.
JMIR medical informaticsJonas Roth, Verena Voigt, Michael Schauwinhold, Andreas Follmann, Michael Czaplik
BACKGROUND: Telemedicine may improve access to specialized care, but its use for supervision during critical anesthesia situations remains underexplored. A standardized tele-supervision (TSV) solution for operating rooms (ORs) is lacking. OBJECTIVE: This study aimed to evaluate a novel telemedical supervision system for critical anesthetic scenarios in a simulated OR environment and compare it to traditional on-site supervision. Specifically, adherence to standard operating procedures (SOPs), the number and modality of senior physician contacts, workload, and user perceptions were assessed. METHODS: In this randomized controlled simulation study, 16 anesthesiology residents in their first 2 years of training at the Uniklinik Rheinisch-Westfälische Technische Hochschule (RWTH) Aachen (Germany) were randomized using block randomization into 2 groups. The intervention group received remote support exclusively via a TSV system, while the control group used a conventional phone with on-site support. The telemedical system comprised an anesthesia workstation that integrated data from the patient monitor, anesthesia device, and syringe pumps using the Institute of Electrical and Electronics Engineers (IEEE) 11073 Service-Oriented Device Connectivity (SDC) standard and a mobile supervision workstation that enabled the senior physician to monitor multiple ORs and communicate via text, audio, and video. The simulated scenario involved a male patient aged 51 years undergoing an appendectomy who developed an anaphylactic reaction 3 minutes after receiving cefuroxime. Primary outcomes focused on the completion rate of necessary SOP measures. Secondary outcomes included workload, measured using the NASA Task Load Index (NASA-TLX), and participants' postscenario questionnaire responses. Statistical comparisons were performed using a Welch t test. RESULTS: All participants in both groups contacted the senior physician at least once. The control group had a mean of 6.44 (SD 4.80) SOP measures supported by the senior physician, whereas the intervention group had a mean of 5.00 (SD 3.06). The mean SOP completion rate was 92.5% (SD 0.04%) in the control group and 91.6% (SD 0.05%) in the intervention group, with no significant difference (t11.94=0.439; P=.67). NASA-TLX scores revealed that compared to the control group, there was lower mental demand in the TSV group but higher temporal demand. Subjective evaluations indicated mixed preferences regarding on-site support; however, most participants acknowledged the TSV system as a viable alternative when on-site support was not feasible. CONCLUSIONS: This study demonstrated no statistically significant differences between the groups, indicating similar performance with high adherence to SOPs and comparable clinical decision-making in a simulated high-stakes environment. Despite increased temporal workload, user feedback was positive, underscoring the system's potential to address staffing shortages and resource limitations. Further research in real clinical settings is needed to optimize usability and validate these findings.
BMC anesthesiologyKirsten C Odegard, Erin A Gottlieb
Congenital cardiac anesthesia has evolved substantially over recent decades, contributing to improving outcomes and survival for patients with congenital heart disease (CHD). BMC Anesthesiology has initiated a new Collection, Recent Advances in Anesthesia for Congenital Cardiac Diseases, to highlight emerging research and encourage large-scale studies addressing knowledge gaps, implementation barriers, and future opportunities in the field.
BACKGROUND: Perioperative anaphylaxis is a rare but complex clinical crisis, with outcomes heavily dependent on prompt and effective management by anaesthesiologists. However, evidence supporting the efficacy of scenario-based simulation (SBS) training in perioperative anaphylaxis management remains limited. This study aimed to evaluate the impact of SBS versus video-based training (VBT) on learning outcomes and satisfaction among undergraduate anaesthesia students. METHODS: In this prospective randomised study, 84 fourth-year anaesthesia undergraduate students from Anhui Medical University were assigned to either SBS or VBT for perioperative anaphylaxis education. Knowledge assessments were conducted before (pretest) and after (post-test) training. Outcomes included post-test scores, student satisfaction (assessed via a 5-point Likert scale) and performance in a simulated clinical scenario 1 week later. Scenario performance was evaluated using a technical checklist and the Anesthesiologists' Non-Technical Skills scoring system. RESULTS: Baseline characteristics (age, sex and pretest scores) were comparable between groups. The post-test scores were higher in the SBS group than those of the VBT group (80.0 (73.7-85.0) vs 70.0 (65.0-75.0), respectively, p<0.001). Both groups improved in post-test scores after training. The SBS group's post-test score increased significantly more than that of the VBT group (39.2% (30.7%-44.6%) vs 16.6% (8.3%-20.0%), respectively, p<0.001). In the simulation scenario test, the SBS group outperformed the VBT group in both technical (33.5 (31.0-36.0) vs 28.0 (26.0-30.0), p<0.001) and non-technical skills (47.5 (44.0-51.0) vs 43.0 (42.0-48.0), p=0.002). CONCLUSIONS: SBS demonstrates greater efficacy than VBT in improving anaesthesia students' learning outcomes, including technical and non-technical skills in managing perioperative anaphylaxis, with higher student satisfaction.
Canadian journal of surgery. Journal canadien de chirurgieSally J Bird, Mathew B Kiberd
SummaryAnesthesia workforce shortages are an increasingly important contributor to operating room closures and surgical delays. These gaps disproportionately affect rural and regional hospitals, limiting equitable access to surgery. This commentary outlines how Canada can modernize the anesthesia care team (ACT) model to sustain surgical capacity while maintaining patient safety. Anesthesia assistants (AAs), trained professionals working under anesthesiologist supervision, already support perioperative care in many centres but remain underutilized. Standardizing AA credentials, defining supervision based on case conditions rather than fixed ratios, and implementing transparent safety reporting could expand access responsibly. Surgeons, as procedural leaders whose productivity depends on anesthesia workforce availability, should understand the ACT framework and its safeguards. Strengthening collaboration between surgical and anesthesia leadership offers a practical path to preserve access to surgery and reduce wait times without compromising safety.
Die AnaesthesiologieTimo Kummerow, Pablo Cavalié, Felix Reinecke, Melissa Barroux, Matthias Göpfert, P Scheiermann, R Tomasi
BACKGROUND: Focused cardiac ultrasound (FOCUS) is a targeted point-of-care echocardiographic approach for rapid assessment of cardiac function and morphology. It is established in emergency and intensive care medicine and increasingly more relevant for making perioperative decisions. Data on its integration into perioperative workflows by anesthesiologists and on the role of training and supervision, remain limited. OBJECTIVE: This nationwide survey assessed FOCUS training, supervision and clinical application among anesthesiologists in Germany and examined how training and supervision were associated with self-reported confidence in performing FOCUS and the extent of its clinical use. MATERIAL AND METHODS: We conducted an anonymized digital cross-sectional survey among physician members of the German Society of Anesthesiology and Intensive Care Medicine (DGAI). The standardized questionnaire comprised 21 items on professional background, echocardiographic training, supervised examinations, clinical integration of FOCUS and self-assessed skills. Training formats were grouped as no course, basic course, or basic plus advanced course. Supervision was recorded as the number of supervised FOCUS examinations after course attendance but the survey did not distinguish between bedside instruction, joint interpretation or image review. Complete datasets were analyzed using Pearson's χ2-test and Wilson 95% confidence intervals. RESULTS: Overall, 645 physicians participated and 469 completed the questionnaire. Senior physicians were overrepresented compared with the DGAI membership structure, whereas residents were underrepresented. More than 80% had completed at least one basic FOCUS course, yet more than 40% reported no supervised examinations before independently performing FOCUS in patients. Approximately 40% used FOCUS regularly, another 40% irregularly and 20% did not use it in practice. The FOCUS examination was used most often in emergency and intensive care situations, followed by intraoperative monitoring and preoperative evaluation. For elective preoperative assessment, cardiology was named most frequently, whereas before emergency surgery anesthesiologists were more often involved. Course participation was significantly associated with more frequent FOCUS use and supervised examinations showed an even stronger association with clinical use. Self-reported confidence in qualitative left ventricular function assessment and quantitative ejection fraction measurement increased with course level and supervision. Lack of qualification, insufficient supervision and time constraints were the main barriers. CONCLUSION: Despite high course participation and broad acceptance of FOCUS, the integration into perioperative practice remains limited. Courses alone appear insufficient for sustainable implementation. Supervision, continuous practice and structured longitudinal training pathways seem central to build confidence and support safe patient use. Beyond image interpretation, training should include correct device handling, suitable presets and exposure to normal and pathological findings. The use of FOCUS should be regarded as an adjunct to clinical assessment, not as a replacement for comprehensive cardiological echocardiography; it requires appropriate indications, technical proficiency and awareness of individual limitations. Interpretation is limited by the cross-sectional design, self-reported outcomes and selection bias towards a more experienced and FOCUS-affine subgroup.
BACKGROUND: Open-source, mid-scale large language models (LLMs) have emerged as scalable, privacy-preserving alternatives to ultra-large foundation models (eg, GPT-4) in health care systems. Techniques such as retrieval-augmented generation (RAG) enable sub-100-billion-parameter models to address highly specialized medical domains such as anesthesiology. However, studies evaluating RAG architectures on complex medical examinations remain scarce, highlighting the need for rigorous benchmarking to bridge the gap between raw parametric knowledge and clinically relevant application. OBJECTIVE: This study aimed to systematically evaluate RAG pipelines for answering anesthesiology board-style questions, quantify the effects of key design choices including hyperparameter settings, embedding models, source complexity, and chunking strategies, and compare the performance of reasoning-oriented models with that of conventional LLMs. METHODS: We conducted large-scale benchmarking using American Board of Anesthesiology-style multiple-choice questions to compare multiple RAG-enabled configurations with matched standalone LLM baselines. Configurations were first optimized on a 46-item diagnostic set and then validated on a 350-item corpus. Additional experiments on three 100-question subsets derived from the 350-item corpus were used to assess the effects of source selection, source complexity, information density, and chunking strategy on answer accuracy. Models including Llama-3-8B-Instruct, Llama-3.1-8B-Instruct, Llama-3.2-3B-Instruct, Llama-3.3-70B-Instruct, Qwen2.5-7B and Qwen2.5-72B, and Qwen3-8B and Qwen3-32B reasoning models were evaluated under this framework. Self-reflective RAG (self-RAG) with adaptive retrieval techniques was also implemented and evaluated. Cochran Q and McNemar tests were used to assess performance differences across configurations and model pairs. RESULTS: The RAG framework increased the number of correct answers. System stability peaked under highly deterministic sampling configurations (temperature=0.1, top-p [nucleus sampling]=0.1). High-capacity general-text embeddings and applying context-preserving semantic chunking further improved accuracy. Standard RAG provided only modest gains over nonaugmented baselines, improving accuracy from 50.29% to 56.57%, and self-RAG yielded similarly limited gains of up to 4.85 percentage points. Overall, the Qwen family outperformed the Llama series. The 32-billion-parameter reasoning model Qwen-3-32B achieved an 89% correct ratio under complex distractor-heavy retrieval conditions and up to 96% with direct context, significantly outperforming the much larger 72-billion-parameter conventional model Qwen-2.5-72B-Instruct (84%). Smaller reasoning models also showed greater robustness to noise or suboptimal retrieved documents than larger conventional LLMs. Within the Llama family, increasing parameter size to 70 billion did not produce proportional performance gains on this benchmark. CONCLUSIONS: RAG-based LLM systems improved performance on anesthesiology board-style questions, but gains depended strongly on retrieval design. Careful optimization of retrieval settings, embeddings, and chunking strategies improved robustness and answer accuracy. Reasoning-oriented models demonstrated that multistep reasoning can, in some settings, compensate for larger parameter scale. These findings provide a methodological foundation for developing locally deployable LLM systems for anesthesiology education within structured examination settings.
Anasthesiologie, Intensivmedizin, Notfallmedizin, Schmerztherapie : AINSFlorian Bubser, Karina Jakobsen, Başak Ceyda Meço, Anne Lena Sacher, Gorki Sacher, Vanessa Moll, Finn M Radtke
Preoperative fasting is among the most routine measures in perioperative medicine. While fasting from solid food remains well justified, current evidence suggests that prolonged fluid fasting - particularly for clear liquids - places an unnecessary burden on patients without proven benefit and may be associated with unfavorable clinical outcomes. This review examines preoperative fluid fasting within the context of the Safe Brain Initiative (SBI) and the concept of Patient-centered Precision Care - (PC)2. Using data from a multicenter registry analysis (n = 15837 patients), adherence to recommended fluid fasting intervals was shown to be associated with a lower incidence of postoperative delirium and a shorter hospital length of stay. At the same time, substantial implementation challenges are addressed transparently: despite considerable efforts, the median duration of fluid fasting remained 5 hours, and nearly 60% of patients fasted for more than 4 hours. A differentiated discussion of various fluid fasting regimens - from the conventional 2-hour rule to more liberal approaches such as "Sip-til-send" - is integrated into the (PC)2 framework. For a more detailed discussion of current guideline recommendations and their practical implementation, readers are referred to the 2026 international consensus statement (DOI: 10.1111/anae.70130).
A&A practiceMegan L Allen, Janette Wright, Kara Allen
BACKGROUND: Clinical critical incidents in anesthesiology practice may adversely affect clinician well-being. Post-critical incident debriefing may support staff and patient safety. However, implementing programs in high-acuity environments remains challenging. We aimed to determine baseline post-clinical critical incident practices at our institution, develop and implement a context-adapted program, and evaluate its integration into practice and its impact. METHODS: We undertook a mixed-methods study. The baseline survey measured psychological safety, prior experiences with critical incidents, and debriefing practices. Focus groups deepened our understanding of key stakeholder needs. These findings informed the design of an adapted post-incident support program. A post-implementation evaluation using surveys and focus groups examined perceived impacts. Quantitative data were analyzed descriptively, and qualitative data were analyzed thematically to identify determinants of feasibility and acceptability, as well as contextual barriers. RESULTS: The baseline survey (N = 55/111; 50% response rate) generally demonstrated high psychological safety. Debriefing was a highly emotional experience, logistical barriers were common, and the purpose of debrief sessions was sometimes ambiguous. Following program refinement, duty relief after critical incidents increased (N = 10/37; 27% at baseline, N = 27/42; 64% post-program, χ2(1) = 10.97, P < .001) and most respondents reported feeling supported despite persistent logistical barriers to post-incident responses. Observed vulnerability in senior staff in post-incident support sessions was welcomed by trainees, but sometimes challenging for senior clinicians. Although critical incidents had negative impacts on respondents (N = 28/54, 52% professional and N = 20/54, 37% personal adverse impact), we also identified potential for post-traumatic professional growth (N = 26/54, 48% positive professional impact). CONCLUSIONS: Our stakeholder-informed, contextually adapted debriefing program was acceptable, feasible, and may mitigate the adverse effects of critical incidents while fostering professional growth. Persisting challenges included logistics, the tension between a well-being and clinical review focus, and differential risks and benefits by career stage.
AANA journalMaria van Pelt, Rachel Epstein, Daniel King, Connie L Lorette, Darleen M Lessard, Claudia Garcia, Rigo Garcia
Substance misuse and drug diversion among anesthesiology professionals may begin during training and span decades of practice. Because treatment for substance use disorder (SUD) is often delayed, early recognition and structured education are essential. This national quality improvement study evaluated the effect on knowledge, perceptions, and the likelihood of reporting suspected substance misuse. A descriptive, comparative pre-post survey was conducted with resident registered nurse anesthetists and support people recruited from 32 randomly selected U.S. nurse anesthesia programs. The intervention was a dual-perspective presentation that integrated clinical content and lived experience. Pre-post perceptions, knowledge, and reporting intentions were analyzed among matched respondents using the Wilcoxon signed-rank test (SAS 9.4; P < .05). Of 1,397 attendees, 868 completed the presurvey and 678 the postsurvey (78.1% retention); 425 were matched for analysis. Significant postintervention improvements were observed in perceptions of SUD as a disease (P = .003), knowledge of safety measures (P < .001), confidence in recognizing symptoms (P < .001), and the likelihood of reporting suspected misuse (P < .001). This national educational intervention was associated with immediate improvements in awareness and reporting intentions, offering a reproducible model for anesthesia education and early prevention.
BMC medical educationXiaocheng Zhu, Anliu Tang, Munle Chin, Huan Chang
BACKGROUND: Early clinical exposure allows undergraduate medical students to encounter authentic clinical practice, but novice learners may not know what to observe or how to interpret complex clinical events. This challenge is particularly relevant in anesthesiology, where many professional activities are embedded in perioperative workflow and are not immediately visible to beginners. This study examined whether a structured perioperative observation model was associated with improved learning objective clarity and self-reported purposeful observation during early anesthesiology clinical exposure. METHODS: This prospective cluster-randomized educational study included 154 s-year undergraduate medical students from 12 clerkship groups. Six groups were assigned to a conventional lecture-observation-question-and-answer clerkship, and six groups were assigned to a structured perioperative observation model organized according to the Bridge-in, Objectives, Pre-assessment, Participatory learning, Post-assessment, and Summary (BOPPPS) framework. The primary outcome was the post-clerkship score for learning objective clarity and self-reported purposeful observation ability. Secondary outcomes included anesthesia-related awareness, knowledge test scores, post-clerkship learning experience, and open-ended feedback. A linear mixed-effects model was used to account for clerkship group clustering. RESULTS: Seventy-six students were assigned to the conventional group and 78 to the structured observation group. Academic track distribution differed between groups, while other baseline characteristics were comparable. The structured observation group had a higher primary outcome score than the conventional group (22.55 ± 1.17 vs. 16.74 ± 1.36, P < 0.001). The association remained significant after adjustment for age, sex, academic track, prior operating room exposure, and baseline interest in anesthesiology. The structured group also showed greater gains in anesthesia-related awareness and knowledge and higher post-clerkship learning experience scores. Findings were consistent across academic track and prior operating room exposure strata. CONCLUSIONS: Structured perioperative observation was associated with higher learning objective clarity and self-reported purposeful observation during early anesthesiology clinical exposure. By guiding students to observe perioperative workflow, anesthesiologists' responsibilities, patient safety, and team collaboration, this model may help novice medical students transform passive clinical exposure into more goal-directed learning.
Best practice & research. Clinical anaesthesiologyDirk Vrancken, Björn Stessel, Kristof Nijs, Jasperina Dubois
Point-of-Care Ultrasound (POCUS) has become a core competency for the anaesthesiologist managing patients in the postoperative intensive care unit. By providing immediate bedside imaging without consultative delay, it enhances diagnostic accuracy and directly informs haemodynamic and respiratory decision-making. This review addresses five clinical domains: lung and diaphragm assessment, cardiac evaluation and shock classification, fluid management, abdominal screening and deep vein thrombosis assessment, and procedural guidance. For each domain, sonographic findings, clinical interpretation, and management implications are described. A structured multi-organ approach with mandatory re-evaluation after every intervention is emphasised throughout. Governance requirements, including competency assessment, image archiving, and peer review are discussed alongside their medico-legal implications. Diagnostic accuracy is well established across all domains; randomised outcome data remain the principal evidence gap. Artificial intelligence may reduce operator dependency but requires prospective perioperative validation before routine implementation.
Anesthesiology clinicsEmily E Sharpe, Mark D Rollins
Administration of anesthetic care in the peripartum period represents numerous areas where vigilance in clinical care and implementation of systematic safety measures are needed to improve maternal outcomes. Obstetric anesthesia care often necessitates expeditious intervention in an environment where both maternal and fetal conditions change rapidly. A thorough understanding of the risks of each intervention as well as best practices to improve maternal safety are critical to optimize patient outcomes. Continued reassessment of protocols and implementation of new evidence and consensus-based guidelines are needed to continually enhance maternal safety in this complex health care environment.