A subscription to JoVE is required to view this content. Sign in or start your free trial.

Review Article

Sleep Deprivation, Decision Fatigue, and Clinical Performance in Medical Training: A Narrative Review and Perspective on Neurophysiological Monitoring

87 views

⸱

DOI:

10.3791/71627

⸱

July 28th, 2026

 ,  ,  , 
1Department of Critical Care Medicine, The First Affiliated Hospital, Zhejiang University School of Medicine, 2Department of Education, The First Affiliated Hospital, Zhejiang University School of Medicine, 3Department of Orthopedics, The First Affiliated Hospital, Zhejiang University School of Medicine

* These authors contributed equally

In This Article

Summary

This narrative review and perspective examines how sleep deprivation and decision fatigue affect clinical performance and evaluates EEG- and fNIRS-based monitoring as conceptual tools for fatigue-aware interpretation and feedback in medical training.

Abstract

Sleep deprivation and decision fatigue are important but often under-recognized determinants of daytime clinical performance. In medical training environments, these state-dependent factors may impair attention, executive control, and decision-making, complicating the interpretation of observed competence during wakefulness. This narrative review and perspective synthesizes evidence from sleep science, neurophysiology, psychology, and medical education to examine how sleep loss and decision fatigue shape clinical performance and how non-invasive monitoring approaches may help contextualize this variability. Attention is given to electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) as candidate research tools for characterizing fatigue-related changes in cognitive workload and vigilance. The reviewed literature supports strong links between sleep loss, fatigue, and impaired neurocognitive functioning, but the use of real-time neurophysiological monitoring for actionable readiness assessment in medical training remains conceptual and requires empirical validation. We propose the CBME-BCI-Fatigue framework as a hypothesis-generating perspective model for integrating physiological state information with competency-based assessment, while emphasizing limitations related to construct validity, incremental value, individual variability, feasibility, and governance. This framework may inform future simulation-based and low-stakes research on fatigue-aware feedback and learner support in sleep-vulnerable clinical training environments.

Introduction

Medical education worldwide continues to move toward outcome-oriented models, most notably competency-based medical education (CBME), which emphasizes demonstrable abilities rather than time-dependent progression. The concept of professional competence has long been discussed in relation to observable capability, clinical reasoning, communication, professionalism, and safe patient care1,2. The Accreditation Council for Graduate Medical Education (ACGME) Outcomes Project, CanMEDS-informed approaches, and subsequent CBME theory have further encouraged educators to define training outcomes explicitly and to assess them longitudinally3,4. Recent discussions have emphasized CBME at the front lines of patient care5. Work on implementation across undergraduate, postgraduate, and continuing professional development settings also highlights persistent challenges in feasibility, assessment culture, and faculty development6,7,8,9,10.

Within CBME, entrustable professional activities (EPAs) and entrustment decisions have shifted assessment toward supervision-based judgments11,12,13,14,15. Learning curves and learning analytics have further supported longitudinal interpretation of learner progress16,17. Precision education, education passports, mastery-oriented learning environments, and coaching emphasize feedback, learner support, and adaptive supervision18,19,20,21. These approaches recognize that competence is inferred from patterns of performance across time and context rather than from a single test. However, a persistent blind spot remains: observed performance does not always reflect stable underlying competence. It is also shaped by temporary physiological and cognitive states, including insufficient sleep, circadian misalignment, workload, emotional stress, and decision fatigue22.

This problem is especially relevant in clinical environments where trainees work overnight, manage high patient volumes, perform procedures, and make repeated decisions under uncertainty. Sleep loss and fatigue have been associated with impaired vigilance, executive control, psychomotor speed, mood regulation, and medical-error risk23,24,25,26,27. Surgical and procedural performance may be particularly vulnerable because it depends on sustained attention, fine motor precision, and rapid situational judgment. Reviews and simulator studies link sleep deprivation with changes in surgical dexterity and crisis-management performance28,29,30. Patient-level and procedure-level studies show more heterogeneous findings, underscoring the need for cautious interpretation31,32,33. Duty-hour research, including iCOMPARE and related randomized trials, shows that sleep, handoffs, workload, patient safety, and educational outcomes are interdependent rather than reducible to a single scheduling variable34,35,36,37,38.

State-dependent performance variability is often transient, situational, and difficult to capture using conventional workplace-based assessment tools. A poor-performance episode after a night shift may reflect a durable learning need, a transient fatigue-related lapse, excessive workload, inadequate supervision, or a combination of these factors. Neurophysiological monitoring has been proposed as a complementary approach for characterizing cognitive state during learning and practice. Electroencephalography (EEG) can capture fast neural oscillatory and event-related responses, whereas functional near-infrared spectroscopy (fNIRS) can track cortical hemodynamic changes, particularly in prefrontal regions involved in executive control. These modalities may provide contextual indicators of workload, vigilance, and fatigue. However, they should not be interpreted as validated measures of competence or used as direct triggers for supervisory or entrustment decisions.

This narrative review and perspective synthesizes evidence linking sleep deprivation and decision fatigue to clinical performance, summarize EEG and fNIRS features relevant to fatigue and cognitive workload, and propose a CBME-BCI-Fatigue framework. The goal is not to present an established assessment system, but to clarify a research agenda for state-aware interpretation of trainee performance in sleep-vulnerable medical training environments. Figure 1 presents the proposed framework, illustrating how state-dependent performance modulators, neurophysiological and multimodal monitoring, competency-based interpretation, AI-supported data integration, and adaptive feedback may be linked in future fatigue-aware medical training research.

Access restricted. Please log in or start a trial to view this content.

Review and Perspective

Search approach

This article was developed as a narrative review and perspective rather than a systematic review, scoping review, or meta-analysis. We searched PubMed, PsycINFO, Web of Science Core Collection, and ERIC from database inception to 31 December 2025. Search terms combined concepts related to sleep deprivation, sleep restriction, fatigue, decision fatigue, clinical performance, competency-based medical education (CBME), cognitive workload, electroencephalography (EEG), functional near-infrared spectroscopy (fNIRS), brain-computer interfaces (BCIs), and medical training. A targeted update was also performed to incorporate recent fNIRS literature relevant to sleep and fatigue monitoring.

Sources were selected and prioritized when they directly informed one of four areas: sleep- or fatigue-related changes in neurocognitive and clinical performance; EEG- or fNIRS-derived markers of fatigue, workload, vigilance, or sleep-related vulnerability; conceptual or applied work relevant to integrating physiological state information into CBME; and ethical or implementation issues relevant to educational use. Sources were not selected to provide exhaustive coverage of all sleep, BCI, or neurotechnology literature. We did not prioritize studies that were primarily basic neuroscience, device engineering, or therapeutic neuromodulation papers unless they directly informed fatigue-aware interpretation of trainee performance. No formal risk-of-bias assessment or meta-analytic pooling was performed; findings were synthesized descriptively and critically to support a conceptual, hypothesis-generating perspective framework.

Novel contribution and intended audience

The novelty of this review lies in connecting several related but often separately discussed bodies of literature—resident fatigue, sleep deprivation, decision fatigue, cognitive workload, neurophysiological monitoring, and CBME assessment—around a specific educational problem: how educators should interpret observed clinical performance when a trainee’s cognitive state may be temporarily altered by sleep loss, decision fatigue, or workload. The intended audience is medical educators, clinician-educators, program directors, and researchers interested in fatigue-aware interpretation of trainee performance. The proposed CBME-BCI-Fatigue framework is intended to organize future research questions and clarify validation requirements, not to validate an operational monitoring system or support current high-stakes assessment decisions.

Competence as dynamic performance in CBME

The original rationale for CBME was not simply to create new assessment forms, but to align medical education with the needs of patients and health systems. CBME emphasizes progression based on proficiency, individualized learning trajectories, and evidence that learners can perform meaningful clinical activities with appropriate supervision5,6. National and regional competency frameworks, including China-based residency competency frameworks and i-FIRST talent-development discussions, provide structured language for curriculum design, learner development, and educational accountability7,8. At the same time, literature on CBME implementation notes persistent concerns about feasibility, faculty development, assessment burden, and the risk of reducing complex professional practice to checklist behavior9,10.

EPAs and entrustment decisions have been particularly influential because they connect assessment to the clinical question of how much supervision a learner requires for a defined professional activity11,12,13,14. EPA implementation also highlights that entrustment is a judgment about a learner, a task, a context, and a level of supervision; it is not simply a fixed trait embedded in the trainee15. Similarly, learning curves and learning analytics can help educators identify trajectories over time rather than overinterpreting single data points16,17. Precision education, education passports, mastery-oriented learning environments, and coaching further reinforce the idea that assessment should support development, feedback, and adaptive support18,19,20,21.

A state-dependent perspective extends this logic. If a trainee performs poorly after prolonged wakefulness or a cognitively intense shift, the assessment question is not only whether the trainee lacks competence. It is also whether the observed behavior reflects fatigue-related vulnerability, excessive workload, poor supervision, system design, or a mismatch between task demands and cognitive state. A stable competence gap may require deliberate practice and coaching; a fatigue-related lapse may require recovery planning, workload redesign, supervision adjustment, or schedule-level intervention. State-aware assessment therefore does not redefine competence. Rather, it improves the context in which performance data are interpreted and reduces the risk of misclassifying transient impairment as durable inability.

Sleep deprivation, circadian disruption, and decision fatigue

Sleep deprivation and fatigue are pervasive yet underappreciated determinants of physician performance. In clinical training, extended-duration shifts, night work, unpredictable workload, and frequent handoffs expose trainees to conditions that can degrade alertness and decision-making. Studies of physician sleep and wellbeing have linked sleep and burnout-related distress with clinically significant medical errors23. Extended-duration shifts have also been associated with increased attentional failures, medical errors, and adverse events24. Resident fatigue and distress may further influence self-reported medical errors, suggesting that performance risk is partly behavioral, partly cognitive, and partly system-level25.

Observational and experimental studies in medical trainees and physicians reinforce this concern. Interns and residents working extended shifts show reduced sleep opportunity and altered alertness patterns26. Reducing intern work hours in intensive care units was associated with fewer serious medical errors in a landmark trial27. Surgical and procedural studies show that sleep deprivation may impair simulator dexterity and crisis-management performance28,29,30. Patient-level and procedure-level studies remain more heterogeneous and should not be oversimplified31,32,33. Differences across specialty, training level, case complexity, supervision, team structure, and outcome type may explain why fatigue-related decrements are clearer for vigilance, simulated performance, and process measures than for downstream patient outcomes.

Duty-hour trials illustrate why fatigue-aware education cannot rely only on shift length. The iCOMPARE trial found that flexible duty-hour policies were noninferior to standard policies for selected education outcomes and patient safety outcomes34,35. Companion analyses reported that sleep and alertness did not necessarily worsen under flexible duty-hour rules36. In contrast, a randomized trial of schedules without 24-hour shifts showed that eliminating extended shifts did not uniformly improve patient safety37. A related study of extended work shifts found neurobehavioral performance effects in resident physicians38. Taken together, these findings show that sleep opportunity, workload, continuity, handoffs, and supervision interact in complex ways.

The broader sleep science literature provides mechanistic context. Sleep disruption impairs sustained attention, working memory, response inhibition, emotion regulation, and complex task performance39,40,41,42. Irregular or unstable sleep schedules may also affect academic and cognitive outcomes, suggesting that sleep stability matters in addition to sleep duration43,44. Chronic sleep restriction with weekend recovery may leave residual effects on performance and wellbeing45. In medical education, these findings are important because training assessments are usually conducted during wakefulness but rarely document the learner state in which performance is expressed.

Decision fatigue adds a related but distinct mechanism. It refers to a deterioration in decision quality after repeated or effortful choices, particularly when decisions require self-control, uncertainty management, or value tradeoffs46. Clinical decisions often involve triage, prescribing, escalation of care, documentation, and handoffs under time pressure. Recent prescribing research suggests that time-on-task and accumulated decision demands can affect clinical choices47. Fatigue from overnight shifts can also alter search patterns and diagnostic performance in radiology48. However, decision fatigue should be interpreted cautiously because the literature is methodologically heterogeneous and effects may vary across tasks, environments, and clinicians.

EEG and fNIRS markers of fatigue and cognitive workload

EEG has been widely used to study vigilance, sleep pressure, fatigue, and mental workload. In physician studies, quantitative EEG after an overnight on-call duty has shown changes in theta, alpha, and beta activity consistent with mental fatigue49,50. These studies are important because they bring neurophysiological assessment closer to real clinical work, rather than relying solely on laboratory sleep-deprivation paradigms. However, they remain observational and cannot by themselves establish that EEG can determine whether a trainee is ready to perform a particular clinical task.

Feature-level EEG interpretation is broader than the commonly cited changes in theta or alpha/beta. Frontal EEG workload studies support the use of spectral and regional patterns for workload estimation51. Sleep-deprivation EEG studies have examined changes in spectral activity after sleep loss52. The theta band also has a dual interpretation: it may reflect sleep pressure, but can also participate in cognitive control, a point that complicates simple fatigue classification53. Real-time working-memory monitoring studies and driver-state classification work further show how EEG features can be embedded in BCI-style pipelines, but these pipelines remain sensitive to task design, preprocessing, and individual variability54,55. Reliability studies and neuroergonomic validation work emphasize that electrode configuration, preprocessing choices, subjective metrics, and task design influence workload indices56,57. More recent multilevel cognitive-load prediction work suggests that combining spatial, temporal, and spectral signatures can improve classification58.

Functional near-infrared spectroscopy fNIRS provides complementary information by measuring hemodynamic changes, including oxygenated hemoglobin, deoxygenated hemoglobin, and total hemoglobin. The fNIRS literature on fatigue and sleep deprivation supports its relevance for measuring cortical correlates of cognitive effort, reduced alertness, and social-cognitive changes59. Sleep-focused fNIRS reviews further show how optical hemodynamic signals may contribute to sleep research and fatigue detection60. A recent 2026 study on cardiac-band fNIRS biomarkers for non-rapid eye movement sleep differentiation with EEG validation illustrates the growing interest in frequency-domain and cardiac-band features beyond simple HbO amplitude61. Together, this literature supports a feature-level discussion that includes HbO, HbR, HbT, time-domain responses, connectivity, and frequency-domain biomarkers.

Prefrontal fNIRS studies are particularly relevant because executive control, response inhibition, and workload regulation depend heavily on frontal systems. Sleep deprivation has been associated with altered task-related functional connectivity in the frontal cortex62. Related work has found loss of frontal functional connectivity after sleep deprivation63. Studies in short-sleep young adults also suggest that prefrontal hemodynamic responses can vary with sleep behavior and physical activity64. Wearable fNIRS has been used for automatic cognitive-fatigue detection with machine learning65. Hybrid EEG-fNIRS systems can improve workload classification by integrating functional brain connectivity and complementary signal properties66,67. Collectively, these studies support the use of EEG and fNIRS as research tools and as candidate contextual indicators, but not yet as validated readiness measures in medical training.

Table 1 is therefore intended as a feature-level map rather than a hierarchy of validated biomarkers. It distinguishes EEG spectral, event-related, temporal, and connectivity features from fNIRS amplitude-based, time-domain, connectivity, and frequency-domain or cardiac-band features. This distinction is important because each feature family captures different dimensions of the fatigue workload construct and carries different artifact and interpretation risks. For example, a higher theta/alpha ratio, reduced P300 amplitude, or lower prefrontal HbO may be compatible with fatigue, but may also reflect task difficulty, compensatory effort, stress, sensor configuration, or individual baseline differences. The table is intended to support study design, construct alignment, and cautious interpretation, not to define operational thresholds for competence, entrustment, or clinical readiness.

BCI, multimodal fusion, and educational applications

Non-invasive BCIs and wearable neurophysiological technologies have expanded rapidly across medical, engineering, and educational domains. Reviews of wearable EEG-based BCI devices highlight progress in portability and medical applications68. Broader state-of-the-art reviews emphasize that non-invasive BCI systems are improving but still face limitations in signal quality, usability, calibration, and robustness69,70,71,72. These constraints matter for medical education because an educational tool must be not only accurate in a controlled setting, but also feasible, acceptable, interpretable, and aligned with educational constructs.

The most defensible educational applications are simulation, research, and low stakes feedback. VR laparoscopic training can synchronize performance data with workload related signals to study how cognitive demand relates to procedural learning73. Simulated neurosurgical tasks have used AI and EEG to distinguish expertise levels, suggesting a pathway for studying skill acquisition rather than replacing faculty judgment74. Serious-game and mixed-reality environments have used biosensor analytics to estimate affective engagement75. Passive BCI systems that adapt learning speed to cognitive load illustrate how state-sensitive feedback might operate in lower-risk educational contexts76. Ambulatory sleep detection work shows how multimodal time-series data can support longitudinal monitoring outside the laboratory77. These examples are best interpreted as supporting evidence or proof-of-concept, not as validation for high-stakes entrustment or remediation decisions.

AI-assisted sleep staging studies show how machine-learning methods can classify sleep-related physiological time series78,79,80. Targeted memory reactivation, wearable neuromodulation, fatigue interventions, and wireless fNIRS systems further illustrate how physiological signals can be tracked or used in adaptive feedback architectures81,82,83,84,85. These studies are technically informative, but their direct educational relevance is limited unless they are linked to daytime performance, feedback timing, supervision decisions, or learner support in training contexts. They should therefore be interpreted as methodological bridges rather than as evidence that fatigue-aware assessment within CBME is ready for routine use.

Workplace translation requires greater caution than simulation. Authentic clinical environments introduce motion artifacts, interruptions, emotional stress, teamwork demands, patient acuity, variable task goals, and implicit power dynamics between trainees and supervisors. Physiological data in workplace-based assessment should therefore be voluntary, low-stakes, and contextual. Reasonable applications include aggregate fatigue-risk mapping, schedule evaluation, simulation debriefing, or learner-controlled feedback. Unreasonable applications would include using EEG, fNIRS, or multimodal thresholds as standalone evidence for entrustment, promotion, remediation, or discipline.

Established evidence versus conceptual extensions

A central distinction in this review is between established evidence and conceptual extension. The established evidence is that sleep deprivation, circadian disruption, prolonged wakefulness, and fatigue impair neurocognitive domains that matter for clinical work23,24,25,26,27. Meta-analytic and experimental work further supports effects on attention, executive control, mood regulation, and complex task performance39,40,41,42. Medical and procedural studies indicate that fatigue can affect error risk, diagnostic search, crisis management, and technical performance28,29,30,48. However, procedure-level and outcome studies also show heterogeneity, which argues against simple one-to-one interpretation of fatigue and clinical harm31,32,33. Duty-hour trials demonstrate that fatigue-aware policy must consider work-hour design together with sleep opportunity, workload, handoffs, continuity, and educational consequences34,35,36,37,38. EEG studies in physicians and workload paradigms support neural sensitivity to fatigue and cognitive load49,50,51, while fNIRS, wearable, and hybrid EEG-fNIRS studies provide complementary hemodynamic and multimodal evidence59,60,61,62,63,64,65,66,67.

The conceptual extension is the proposal that physiological state indicators might help contextualize trainee performance within CBME. This extension is plausible but unvalidated. Evidence that sleep loss changes neural activity does not automatically imply that real-time monitoring can generate actionable supervisory or educational decisions. At present, no generalizable real-time pattern of neural or hemodynamic activity has been shown to reliably identify cognitive instability or clinical readiness across individuals, tasks, specialties, and operational settings. The same physiological pattern may reflect fatigue, cognitive effort, anxiety, motivation, task difficulty, compensatory engagement, sensor artifact, environmental stress, or individual baseline differences.

Accordingly, the CBME-BCI-Fatigue framework should be interpreted as a research model rather than as an implemented assessment system. Its purpose is to clarify how state-dependent modulators, physiological and behavioral monitoring, and competency-based interpretation of observed performance might be studied together. In simulation or other low-stakes research settings, EEG or fNIRS features could be collected alongside task errors, response time, workload ratings, sleep history, psychomotor vigilance testing, and supervisor feedback to test whether state indicators help explain performance variability. In workplace assessment contexts, physiological data require even greater caution. In the current evidence base, such data may be most appropriate for aggregate fatigue-risk research, schedule design, or voluntary learner support. They should not be used as a standalone basis for entrustment, promotion, remediation, or disciplinary decisions. Table 2 summarizes these distinctions by separating established evidence, conceptual extensions, validation requirements, and representative supporting references within the CBME-BCI-Fatigue framework.

Construct validity, incremental value, individual baselines, and governance

Construct validity is the central methodological challenge for fatigue-aware neurophysiological monitoring. EEG and fNIRS signals are not specific to fatigue or readiness; they may also reflect task difficulty, anxiety, motivation, cognitive effort, compensatory engagement, environmental stress, sensor artifacts, or individual baseline differences. Future studies should therefore predefine the educational construct being measured, specify the expected relation between physiological state and performance, and triangulate physiological signals with behavioral, contextual, and educational data rather than interpreting a single signal as a direct marker of competence.

Incremental value must also be demonstrated. EEG/fNIRS monitoring should be evaluated against simpler and more feasible comparators, including sleep logs, actigraphy, psychomotor vigilance testing, brief cognitive screening, workload ratings, performance telemetry, and expert observation. A neurophysiological approach would be justified only if it improves explanation, prediction, feedback timing, or learner support beyond these measures. Without such evidence, multimodal monitoring may add complexity without improving educational decision-making.

Individual baseline modeling is essential. Interindividual variation in neural activity, sleep vulnerability, chronotype, stress response, clinical experience, and sensor fit limits the value of universal thresholds. A trainee-specific baseline may help distinguish unusual fatigue-related deviation from normal physiological variation, but this requires repeated measurements, reliable sensor placement, standardized task anchors, and transparent analytic models. Machine learning may help integrate multimodal signals, but it cannot solve invalid constructs, biased data, poor signal quality, or unclear causal assumptions. Domain adaptation, privacy-preserving algorithms, and generative modeling may reduce some calibration burden, but they require broader validation before being used to support educational decisions86,87.

Technical and ecological barriers remain substantial. EEG is vulnerable to motion artifact, electrode impedance, environmental noise, montage dependence, and preprocessing choices. fNIRS is affected by optode placement, scalp blood flow, motion, hair, limited depth penetration, and lower temporal resolution. Wearable systems are improving, including compact wireless fNIRS devices for cognitive activity monitoring85, but comfort and feasibility do not establish educational validity. Deployment also requires workflow-compatible setup, interpretable reporting, data security, and end-user acceptance. Social acceptance studies remind us that user trust and perceived purpose are central to adoption88.

Ethical governance should be treated as a boundary condition of the framework rather than as an afterthought. Neurophysiological data can reveal or imply information about fatigue, attention, emotion, or health status. Responsible development of neurotechnology requires purpose limitation, data minimization, secure storage, transparency, access controls, and careful attention to autonomy and consent89,90. Networked BCI systems also raise cybersecurity concerns, especially if decoded outputs influence feedback, training opportunities, or safety-relevant decisions91. In education, learners may feel implicit pressure to participate, and physiological metrics could unintentionally amplify bias if models perform unevenly across populations or contexts. Future research should prioritize voluntary participation where possible, clear separation between supportive feedback and high stakes judgment, appeal mechanisms, and stakeholder-informed governance.

Access restricted. Please log in or start a trial to view this content.

Conclusions

Sleep deprivation and decision fatigue are established contributors to state-dependent variability in attention, executive control, psychomotor performance, and clinical decision-making. In medical training, these fluctuations may complicate the interpretation of observed competence, particularly in high-demand environments such as emergency medicine, critical care, anesthesia, and surgery. Competency-based medical education may therefore benefit from more explicit attention to physiological and cognitive state when perf...

Access restricted. Please log in or start a trial to view this content.

Disclosures

The authors declare no conflict of interest.

Acknowledgements

The authors gratefully acknowledge colleagues from the Department of Critical Care Medicine and the Department of Education at our institution for their support and discussions during preparation of this review. This work was supported by the General Scientific Research Project of the Zhejiang Provincial Department of Education (Grant No. Y202557476).

Access restricted. Please log in or start a trial to view this content.

References

  1. Epstein RM, Hundert EM. Defining and assessing professional competence. JAMA. 2002;287(2):226–235.
  2. McClelland DC. Testing for competence rather than for “intelligence”. Am Psychol. 1973;28(1):1–14.
  3. Swing SR. The ACGME Outcome Project: retrospective and prospective. Med Teach. 2007;29(7):648–654.
  4. Frank JR et al. Competency-based medical education: theory to practice. Med Teach. 2010;32(8):638–645.
  5. Cooper D, Holmboe ES. Competency-based medical education at the front lines of patient care. N Engl J Med. 2025;393(4):376–388.
  6. Caccia N, Nakajima A, Kent N. Competency-based medical education: the wave of the future. J Obstet Gynaecol Can. 2015;37(4):349–353.
  7. China Consortium of Elite Teaching Hospitals for Residency Education. Consensus on core competency framework for residency education among China Consortium of Elite Teaching Hospitals for Residency Education [in Chinese]. Med J Peking Union Med Coll Hosp. 2022;13(1):17–23.
  8. Chen S et al. Competency-based i-FIRST talent development: leadership. China Higher Med Educ. 2025;(1):20–21, 45. doi:10.3969/j.issn.1002-1701.2025.01.006.
  9. Danilovich N et al. Implementing competency-based medical education in family medicine: a narrative review of current trends in assessment. Fam Med. 2021;53(1):9–22.
  10. Hawkins RE et al. Implementation of competency-based medical education: are we addressing the concerns and challenges. Med Educ. 2015;49(11):1086–1102.
  11. Ten Cate O et al. Entrustment decision making: extending Miller’s pyramid. Acad Med. 2021;96(2):199–204.
  12. Dine CJ, Conforti LN, Holmboe ES, Kogan JR. Accuracy of entrustment-based assessment: implications for programs and patients. J Grad Med Educ. 2024;16(1):30–36.
  13. Tekian A et al. Entrustment decisions: implications for curriculum development and assessment. Med Teach. 2020;42(6):698–704.
  14. Ten Cate O. Entrustment as assessment: recognizing the ability, the right, and the duty to act. J Grad Med Educ. 2016;8(2):261–262.
  15. Carraccio C, Martini A, Van Melle E, Schumacher DJ. Identifying core components of EPA implementation: a path to knowing if a complex intervention is being implemented as intended. Acad Med. 2021;96(9):1332–1336.
  16. Pusic MV et al. Learning curves in health professions education. Acad Med. 2015;90(8):1034–1042.
  17. Smith BK et al. The use of learning analytics to enable detection of underperforming trainees: an analysis of national vascular surgery trainee ACGME milestones assessment data. Ann Surg. 2023;277(4)–e977.
  18. Desai SV et al. Precision education: the future of lifelong learning in medicine. Acad Med. 2024;99(4S Suppl 1)–S20.
  19. Warm EJ et al. The education passport: connecting programmatic assessment across learning and practice. Can Med Educ J. 2022;13(4):82–91.
  20. Ross S, Pirraglia C, Aquilina AM, Zulla R. Effective competency-based medical education requires learning environments that promote a mastery goal orientation: a narrative review. Med Teach. 2022;44(5):527–534.
  21. Baenziger K, Chan M, Colman S. Coaching in postgraduate competency-based medical education: a qualitative exploration of three models. Acad Psychiatry. 2023;47(1):10–17.
  22. Taylor TS et al. Principles of fatigue in residency education: a qualitative study. CMAJ Open. 2016;4(2)–E204.
  23. Trockel MT et al. Assessment of physician sleep and wellness, burnout, and clinically significant medical errors. JAMA Netw Open. 2020;3(12).
  24. Barger LK et al. Impact of extended-duration shifts on medical errors, adverse events, and attentional failures. PLoS Med. 2006;3(12).
  25. West CP et al. Association of resident fatigue and distress with perceived medical errors. JAMA. 2009;302(12):1294–1300.
  26. Basner M et al. Sleep and alertness in medical interns and residents: an observational study on the role of extended shifts. Sleep. 2017;40(4). doi:10.1093/sleep/zsx027.
  27. Landrigan CP et al. Effect of reducing interns’ work hours on serious medical errors in intensive care units. N Engl J Med. 2004;351(18):1838–1848.
  28. Sugden C, Athanasiou T, Darzi A. What are the effects of sleep deprivation and fatigue in surgical practice. Semin Thorac Cardiovasc Surg. 2012;24(3):166–175.
  29. Arzalier-Daret S et al. Effect of sleep deprivation after a night shift duty on simulated crisis management by residents in anaesthesia: a randomised crossover study. Anaesth Crit Care Pain Med. 2018;37(2):161–166.
  30. Maltese F et al. Night shift decreases cognitive performance of ICU physicians. Intensive Care Med. 2016;42(3):393–400.
  31. Rothschild JM et al. Risks of complications by attending physicians after performing nighttime procedures. JAMA. 2009;302(14):1565–1572.
  32. Taffinder NJ et al. Effect of sleep deprivation on surgeons’ dexterity on laparoscopy simulator. Lancet. 1998;352(9135):1191.
  33. Howard SK et al. Simulation study of rested versus sleep-deprived anesthesiologists. Anesthesiology. 2003;98(6):1345–1355.
  34. Desai SV et al. Education outcomes in a duty-hour flexibility trial in internal medicine. N Engl J Med. 2018;378(16):1494–1508.
  35. Silber JH et al. Patient safety outcomes under flexible and standard resident duty-hour rules. N Engl J Med. 2019;380(10):905–914.
  36. Basner M et al. Sleep and alertness in a duty-hour flexibility trial in internal medicine. N Engl J Med. 2019;380(10):915–923.
  37. Landrigan CP et al. Effect on patient safety of a resident physician schedule without 24-hour shifts. N Engl J Med. 2020;382(26):2514–2523.
  38. Rahman SA et al. Extended work shifts and neurobehavioral performance in resident physicians. Pediatrics. 2021;147(3). doi:10.1542/peds.2020-009936.
  39. Wickens CD, Hutchins SD, Laux L, Sebok A. The impact of sleep disruption on complex cognitive tasks: a meta-analysis. Hum Factors. 2015;57(6):930–946.
  40. Lowe CJ, Safati A, Hall PA. The neurocognitive consequences of sleep restriction: a meta-analytic review. Neurosci Biobehav Rev. 2017;80:586–604.
  41. Lim J, Dinges DF. A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychol Bull. 2010;136(3):375–389.
  42. Tomaso CC, Johnson AB, Nelson TD. The effect of sleep deprivation and restriction on mood, emotion, and emotion regulation: three meta-analyses in one. Sleep. 2021;44(6). doi:10.1093/sleep/zsaa289.
  43. Phillips AJK et al. Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing. Sci Rep. 2017;7(1):3216.
  44. Koa TB, Lo JC. Neurobehavioural functions during variable and stable short sleep schedules. J Sleep Res. 2021;30(4).
  45. Smith MG et al. Effects of six weeks of chronic sleep restriction with weekend recovery on cognitive performance and wellbeing in high-performing adults. Sleep. 2021;44(8). doi:10.1093/sleep/zsab051.
  46. Pignatiello GA, Martin RJ, Hickman RL Jr. Decision fatigue: a conceptual analysis. J Health Psychol. 2020;25(1):123–135.
  47. Maier M et al. Assessing decision fatigue in general practitioners’ prescribing decisions using the Australian BEACH data set. Med Decis Making. 2024;44(6):627–640.
  48. Hanna TN et al. The effects of fatigue from overnight shifts on radiology search patterns and diagnostic performance. J Am Coll Radiol. 2018;15(12):1709–1716.
  49. Su AT, Xavier G, Kuan JW. The measurement of mental fatigue following an overnight on-call duty among doctors using electroencephalogram. PLoS One. 2023;18(7).
  50. Xavier G, Su Ting A, Fauzan N. Exploratory study of brain waves and corresponding brain regions of fatigue on-call doctors using quantitative electroencephalogram. J Occup Health. 2020;62(1).
  51. So WKY, Wong SWH, Mak JN, Chan RHM. An evaluation of mental workload with frontal EEG. PLoS One. 2017;12(4).
  52. Forest G, Godbout R. Effects of sleep deprivation on performance and EEG spectral analysis in young adults. Brain Cogn. 2000;43(1–3):195–200.
  53. Snipes S, Krugliakova E, Meier E, Huber R. The theta paradox: 4–8 Hz EEG oscillations reflect both sleep pressure and cognitive control. J Neurosci. 2022;42(45):8569–8586.
  54. Mora-Sanchez A et al. A brain-computer interface for the continuous, real-time monitoring of working memory load in real-world environments. Cogn Neurodyn. 2020;14(3):301–321.
  55. Zeng H et al. A LightGBM-based EEG analysis method for driver mental states classification. Comput Intell Neurosci. 2019;2019:3761203.
  56. Mastropietro A et al. Reliability of mental workload index assessed by EEG with different electrode configurations and signal pre-processing pipelines. Sensors (Basel). 2023;23(3):1367.
  57. Bjegojevic B et al. Neuroergonomic attention assessment in safety-critical tasks: EEG indices and subjective metrics validation in a novel task-embedded reaction time paradigm. Brain Sci. 2024;14(10):1009.
  58. Liu Y et al. Fusion of spatial, temporal, and spectral EEG signatures improves multilevel cognitive load prediction. IEEE Trans Hum Mach Syst. 2023;53(2):357–366.
  59. Pan Y, Borragán G, Peigneux P. Applications of functional near-infrared spectroscopy in fatigue, sleep deprivation, and social cognition. Brain Topogr. 2019;32(6):998–1012.
  60. Huang M et al. An overview on sleep research based on functional near-infrared spectroscopy [in Chinese]. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi. 2021;38(6):1211–1218.
  61. Hasim YS et al. Cardiac-band fNIRS biomarkers for non-REM sleep stage differentiation: a frequency-domain evaluation with EEG validation. J Innov Opt Health Sci. 2026. doi:10.1142/S1793545826500100.
  62. Mukli P et al. Sleep deprivation alters task-related changes in functional connectivity of the frontal cortex: a near-infrared spectroscopy study. Brain Behav. 2021;11(8).
  63. Verweij IM et al. Sleep deprivation leads to a loss of functional connectivity in frontal brain regions. BMC Neurosci. 2014;15(1):88.
  64. You Y et al. Cognitive performance in short sleep young adults with different physical activity levels: a cross-sectional fNIRS study. Brain Sci. 2023;13(2):171.
  65. Varandas R et al. Automatic cognitive fatigue detection using wearable fNIRS and machine learning. Sensors (Basel). 2022;22(11):4010.
  66. Cao J, Garro EM, Zhao Y. EEG/fNIRS-based workload classification using functional brain connectivity and machine learning. Sensors (Basel). 2022;22(19):7623.
  67. Hong KS, Khan MJ, Hong MJ. Feature extraction and classification methods for hybrid fNIRS-EEG brain-computer interfaces. Front Hum Neurosci. 2018;12:246.
  68. Zhang J et al. Recent progress in wearable brain-computer interface devices based on electroencephalogram for medical applications: a review. Health Data Sci. 2023;3:0096.
  69. Edelman BJ et al. Non-invasive brain-computer interfaces: state of the art and trends. IEEE Rev Biomed Eng. 2025;18:26–49.
  70. Rashid M et al. Current status, challenges, and possible solutions of EEG-based brain-computer interface: a comprehensive review. Front Neurorobot. 2020;14:25.
  71. Xu M et al. Current challenges for the practical application of electroencephalography-based brain-computer interfaces. Engineering. 2021;7(12):1710–1712.
  72. Maiseli B et al. Brain-computer interface: trend, challenges, and threats. Brain Inform. 2023;10(1):20.
  73. Yu P et al. Quantitative influence and performance analysis of virtual reality laparoscopic surgical training system. BMC Med Educ. 2022;22(1):92.
  74. Natheir S et al. Utilizing artificial intelligence and electroencephalography to assess expertise on a simulated neurosurgical task. Comput Biol Med. 2023;152:106286.
  75. Antoniou PE et al. Biosensor real-time affective analytics in virtual and mixed reality medical education serious games: cohort study. JMIR Serious Games. 2020;8(3).
  76. Beauchemin N et al. Enhancing learning experiences: EEG-based passive BCI system adapts learning speed to cognitive load in real time, with motivation as catalyst. Front Hum Neurosci. 2024;18:1416683.
  77. Sano A, Chen W, Lopez-Martinez D, Taylor S, Picard RW. Multimodal ambulatory sleep detection using LSTM recurrent neural networks. IEEE J Biomed Health Inform. 2019;23(4):1607–1617.
  78. Kong G, Li C, Peng H, Han Z, Qiao H. EEG-based sleep stage classification via neural architecture search. IEEE Trans Neural Syst Rehabil Eng. 2023;31:1075–1085.
  79. Yao H, Liu T, Zou R, Ding S, Xu Y. A spatial-temporal transformer architecture using multi-channel signals for sleep stage classification. IEEE Trans Neural Syst Rehabil Eng. 2023;31:3353–3362.
  80. Eldele E et al. Self-supervised learning for label-efficient sleep stage classification: a comprehensive evaluation. IEEE Trans Neural Syst Rehabil Eng. 2023;31:1333–1342.
  81. Ngo HV, Staresina BP. Shaping overnight consolidation via slow-oscillation closed-loop targeted memory reactivation. Proc Natl Acad Sci U S A. 2022;119(44).
  82. Bressler S, Neely R, Yost RM, Wang D, Read HL. A wearable EEG system for closed-loop neuromodulation of sleep-related oscillations. J Neural Eng. 2023;20(5). doi:10.1088/1741-2552/acfb3b.
  83. Gervais C, Hjeij D, Fernandez-Puerta L, Arbour C. Non-pharmacological interventions for sleep disruptions and fatigue after traumatic brain injury: a scoping review. Brain Inj. 2024;38(6):403–416.
  84. Herrero Babiloni A et al. The effects of non-invasive brain stimulation on sleep disturbances among different neurological and neuropsychiatric conditions: a systematic review. Sleep Med Rev. 2021;55:101381.
  85. Victorio M et al. Wearable wireless functional near-infrared spectroscopy system for cognitive activity monitoring. Biosensors (Basel). 2025;15(2):92.
  86. Wen S et al. Rapid adaptation of brain-computer interfaces to new neuronal ensembles or participants via generative modelling. Nat Biomed Eng. 2023;7(4):546–558.
  87. Xia K, Deng L, Duch W, Wu D. Privacy-preserving domain adaptation for motor imagery-based brain-computer interfaces. IEEE Trans Biomed Eng. 2022;69(11):3365–3376.
  88. Xia R, Yang S. Factors influencing the social acceptance of brain-computer interface technology among Chinese general public: an exploratory study. Front Hum Neurosci. 2024;18:1423382.
  89. Goering S et al. Recommendations for responsible development and application of neurotechnologies. Neuroethics. 2021;14(3):365–386.
  90. Ienca M, Andorno R. Towards new human rights in the age of neuroscience and neurotechnology. Life Sci Soc Policy. 2017;13(1):5.
  91. Ajrawi S, Rao R, Sarkar M. Cybersecurity in brain-computer interfaces: RFID-based design-theoretical framework. Inform Med Unlocked. 2021;22. doi:10.1016/j.imu.2020.100489.

Access restricted. Please log in or start a trial to view this content.

Reprints and Permissions

Tags

Executive ControlCognitive WorkloadElectroencephalography EEGFunctional Near-Infrared SpectroscopyCompetency-Based Assessment
Contact Us
Recommend to Library

Privacy

Terms of Use

Policies

Copyright © 2026 MyJoVE Corporation. All rights reserved.