MLMarc Lilot

Marc Lilot

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Marc Lilot has a total of 164 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A Personalized Recovery Intervention to Support Sleep Health in Emergency Healthcare Professionals

    by Laura Schmidt, Amandine E. Rey, Bérénice Valero, Sophie Schlatter, Eve Reynaud, Marc Lilot, Karim Tazarourte, Marion Douplat, Stéphanie Mazza

    Optimizing sleep health in healthcare professionals is a challenge, as misalignment between endogenous circadian rhythms and exogenous work schedules undermines recovery, well‐being, and performance. This study evaluated a tailored recovery intervention based on cognitive‐behavioural approaches for insomnia designed to support recovery to rotating night shifts. In a randomized controlled trial (REST), 72 emergency healthcare professionals were randomized to intervention or control. The intervention combined personalized sleep–wake feedback with stress management techniques. Sleep–wake rhythms were monitored via actigraphy for 3 weeks before and 1 week after the intervention. Outcomes included objective sleep duration, self‐reported sleepiness, insomnia severity, and subjective sleep quality, measured using the Epworth Sleepiness Scale, Insomnia Severity Index, and Leeds Sleep Evaluation Questionnaire. Fifty‐nine participants (28 intervention, 31 control) were included in the primary analyses. Even on free nights, average nocturnal sleep duration was 6 h and 21 min, below recommended levels for adults. Compared with controls, intervention participants increased nocturnal time in bed by 37 min, 95% CI [13, 62], p = 0.003, and total sleep time by 23 min, 95% CI [4, 42], p = 0.019. Perceived sleep quality also improved for the intervention group t(55) = −7.91, p < 0.001, d = 2.06, 95% CI [1.42, 2.70], and insomnia symptoms decreased ( β = −2.23, 95% CI [−4.26, −0.21], p = 0.031). These findings suggest that structured occupational recovery programs can enhance sleep health and facilitate recovery across rotating night work schedules. Integrating such interventions into training may reduce burnout, improve safety, and support workforce sustainability.

  • Open Access

    Influence of Personality Traits on Mental Health and Performance of Medical Residents and Physicians: A Systematic Review and Meta-analysis

    by Walaa Garoot, Edouard Leaune, Brice Canada, Christian Echevarría Colón, Marc Lilot, Gilles Rode, Sophie Schlatter

    Summary: Background: Medical residents and physicians face high rates of burnout, psychological distress, and impaired professional functioning. Big Five personality traits are stable characteristics that may explain variation in vulnerability and coping, but evidence in this population remains inconsistent. We aimed to synthesise associations between Big Five traits and stress-related outcomes, coping, and performance in medical residents and physicians. Methods: Systematic review and meta-analysis (PROSPERO CRD42023483408). PubMed, Embase, MEDLINE (Ovid), Cochrane Library, Scopus, and Web of Science were searched from inception to November 15, 2023, updated through January 2026. Eligible studies were primary research in English involving medical residents and/or practising physicians, assessing at least one Big Five trait and a validated measure of stress, coping, or performance. Two reviewers independently screened records; data were extracted by one reviewer and checked by a second. Risk of bias was assessed using the AXIS tool supplemented by JBI criteria and the Cochrane RoB 2. Pearson r was pooled using random-effects models (REML) when k≥3 studies used the same outcome instrument. Findings: Of 4967 records identified, 34 studies (21,379 participants; 17 countries) met inclusion criteria. Neuroticism was associated with greater emotional exhaustion (pooled r=0·42 [95% CI 0·22–0·62], k=11 studies), depersonalisation (r=0·30 [0·17–0·44], k=11), lower personal accomplishment (r=−0·24 [−0·39 to −0·09], k=11), and greater psychological distress (r=0·52 [0·18–0·87], k=3). Conscientiousness and, to a lesser extent, Extraversion and Agreeableness showed small protective associations with burnout dimensions (EE and DP: r=−0·11 to −0·36; PA r=0·11–0·31, k=11). Openness showed a weak, selective pattern. Heterogeneity was substantial (I²>75%); risk of bias was low-to-moderate for most studies. Interpretation: Big Five traits have modest associations with burnout, distress, coping, and performance in physicians and residents. These findings support personality assessment as a formative resource in medical education and workforce development. Funding: None.

  • Techniques cognitives pour l’amélioration des acquisitions en simulation de situations médicales critiques

    by Marc Lilot

    L’éducation des sciences de la santé repose sur des bases conceptuelles solides et propose de multiples méthodes pédagogiques au service des acquisitions de compétences. La simulation en santé est un des outils à fort potentiel impact pédagogique. Par les diverses modalités, la simulation haute-fidélité permet de travailler les compétences requises lors de gestion de situation critiques médicales avec une forte implication des apprenants, facilitées par un très grand réalisme d’expérience immersive. Le niveau de stress des participants durant les séances de simulation haute-fidélité affecte leurs performances en simulation, leurs acquisitions, leurs mémorisations et leurs comportements ultérieurs en situation réelle.La thèse présentée ici s’est intéressée aux niveaux de stress ressentis des internes, apprenants en simulation haute-fidélité, ainsi qu’à certaines techniques cognitives développées dans le but d’amélioration des acquisitions et de diminution du niveau de stress. La première étude observationnelle relate du niveau de stress et d’anxiété ainsi que des différences observées parmi les internes lors de leur première session de simulation. La seconde étude montre, dans un essai randomisé, l’effet favorable d’une pause relaxation intégrée avant le débriefing du scénario, sur la mémorisation à trois mois des messages clés critiques du scénario. La troisième étude, ancillaire de la seconde, relate les différences de niveaux de stress et de mémorisation entre les apprenants impliqués activement et ceux observateurs du scénario. La quatrième étude est un essai randomisé qui montre l’intérêt d’une concertation pré-critique en équipe pour l’amélioration des performances techniques en simulation. La cinquième étude est un essai randomisé qui montre les effets favorables, sur les performances non-techniques et le stress en simulation, d’une formation à la gestion du stress préalable et d’une réactivation juste avant scénario. La sixième étude est le protocole de recherche d’un prochain essai randomisé qui visera à comparer les niveaux de stress et les performances des internes durant la simulation, après qu’ils aient pratiqué une respiration relaxante standardisée, guidée ou non par biofeedback de cohérence cardiaque et comparé à un groupe contrôle

  • Impact of a brief HRV-biofeedback intervention on emotion regulation following a real-life stressful event: A randomized controlled study

    by Jente Depoorter, Kristof Hoorelbeke, Théodore Guillaumée, Marion Cortet, Marc Lilot, Gilles Rode, Rudi De Raedt, Sophie Schlatter


  • Open Access

    Customisable digital cognitive aids cut total error by 75% defining a threshold for irreducible human error in clinical care: a pooled analysis of five randomised trials

    by Jean-Christophe Cejka, Clément Buléon, Marc Lilot, Antoine Duclos, François Lecomte, Karim Tazarourte, Baptiste Balanca, Thomas Rimmelé

    Background: Human error is a leading cause of preventable harm in clinical care, driven by factors such as cognitive overload and staff turnover. While customisable digital cognitive aids (cDCAs) have emerged as real-time protocol-tailored support, a framework to quantify their total impact on human error and its fundamental limits has been lacking. Methods: We conducted a pooled analysis of five randomised high-fidelity simulation trials with consistent methodologies, including 370 healthcare professionals across diverse clinical settings and levels of experience. Using bootstrap resampling, we modelled the distributions of technical and non-technical skills (TS and NTS) to quantify the impact of cDCAs on clinical competence and on Total Human Error (THE)—defined as the sum of systematic deviation from standards (bias 2 ) and inter-individual variability (variance). Findings: The use of cDCAs reduced THE by 75% in a novice user context. This effect was driven by concurrent, drastic reductions in systematic deviation from standards ( bias 2 : 328.1 vs 1377.1) and inter-individual variability ( variance : 124.3 vs 210.0). This was reflected in significant improvements in both technical (81.9±0.8 vs 62.9±1.0, presidual error threshold , quantifying an empirical upper bound of ∼25% for what we define as Irreducible Human Error (IHE). Interpretation: By reducing both systematic bias and performance variability, cDCAs demonstrably harmonise clinical practices and enhance competence robustness. Our analysis provides the first empirical framework to quantify this beneficial effect, revealing in turn a ∼25% residual error that delineates the fundamental limits of performance enhancement achievable via procedural workflow support. We posit this threshold serves as an empirical upper bound for Irreducible Human Error (IHE)—errors arising not from procedural flaws but from higher-order cognitive or interpersonal factors. Establishing IHE as a new benchmark for patient safety thus provides foundational evidence for a model of augmented practice, where complementary innovations such as AI may be required to move beyond procedural support, augment clinical judgement, and uphold human-centred care. Research in Context: Evidence before this study: Cognitive support has long been used to mitigate human error in high-stakes environments. ‘Traditional’ cognitive support tools (e.g., paper checklists) improve adherence to technical protocols during medical crises, but their impact on non-technical skills—such as communication and decision-making—is limited and their clinical adoption remains low. Early digital aids largely replicated these static tools, often lacking the ergonomic integration needed for real-time clinical use and failing to leverage the full potential of digital platforms. Consequently, no prior study had systematically assessed the impact of customisable digital cognitive aids (cDCAs) on both technical and non-technical skills across diverse clinical domains to establish their overall efficiency and robustness . Furthermore, no prior study has established a comprehensive framework for evaluating their effects on ‘Total Human Error’ (THE)—encompassing both bias and variability in practice—or introduced a measurable threshold for an ‘Irreducible Human Error’ (IHE) in clinical care. What this study adds: This study is the first to quantify competence—defined as the integration of procedural proficiency (TS) and cognitive-behavioural dimensions (NTS)—and measured the impact of cDCAs on TS and NTS performance and variability, across multiple clinical settings and experience levels. We pooled five randomised controlled trials with consistent methodologies and used a resampling technique (bootstrap analysis), which allowed us to model distributions of performance and assess robustness across diverse professional backgrounds. Unlike traditional digital checklists, which often impose cognitive effort or disrupt team workflows, cDCAs are adaptive interfaces, integrating seamlessly into decision-making with minimal cognitive load. Our work systematically quantifies clinical competence and THE across both TS and NTS. We demonstrated that cDCAs significantly reduced systematic bias—defined as deviation from expected standards—along with inter-individual and inter-situation variability, thereby harmonising clinical practices and improving reliability—even in high-stakes scenarios. The reduction in inter-situation variability is particularly relevant, as it strengthens the robustness of clinical care delivery, ensuring consistent performance across diverse settings. By defining and quantifying IHE, we establish a benchmark for understanding the limits of human performance—beyond which cognitive aids alone may be insufficient. Implications of all the available evidence: Integrating customisable, ergonomically optimised cDCAs into routine clinical workflows provides a scalable solution to reduce THE by harmonising practices and improving adherence to care standards. This is particularly valuable in resource-limited settings, where cDCAs could help exchange expertise. By establishing IHE as a measurable threshold, this study provides a foundation for further innovations in cognitive support. While current cDCAs provide substantial reductions in THE, these findings raise the question of whether AI could further complement cognitive aids while preserving human expertise at the core of decision-making. Future research should explore how AI-enhanced cognitive aids might ethically and transparently address residual human error, ensuring these technologies reinforce—rather than undermine—the foundational principles of trust, accountability, and patient-centred care. Finally, beyond individual performance, cDCAs may contribute to global equity by promoting robust, harmonised care and enhancing knowledge retention and application across healthcare systems. In an interconnected world, where no system operates in isolation, ensuring safe, high-quality care on a global scale is imperative.

  • Open Access

    The Cognitive Safety Net: Comparing Human and AI Diagnostic Reasoning during Complex Clinical Situations

    by Antonin Audion, Mathieu Henkeme, Baptiste Balanca, Marc Lilot, Thomas Rimmelé, Ismail Abaakil, Jean-Christophe Cejka

    Background: Diagnostic error in high-stakes clinical environments remains a significant cause of preventable harm. While a new generation of customisable digital cognitive aids (cDCAs) has shown a capacity to improve performance, achieve robust competence, and double learning retention, the potential for artificial intelligence (AI) to augment the foundational, anticipatory reasoning that precedes action is not well understood. This study aims to compare the diagnostic reasoning strategies of experienced anaesthesiology residents with those of a large language model (LLM) during a simulated, complex and realistic anaesthesiology scenario. Methods: We conducted a comparative analysis within a high-fidelity simulation randomised controlled trial (Anticipamax, NCT06487208 ). Thirty-four experienced anaesthesiology residents and a conversational LLM (ChatGPT-4) managed a perioperative shock of deliberately multifactorial aetiology. Diagnostic lotteries—sets of hypotheses with assigned plausibility scores—were collected before and after the simulation. We implemented a novel analytical framework based on the social choice Condorcet method , to rank not only individual hypotheses but also to compare the complete diagnostic strategies as the case evolved. Results: The AI and residents demonstrated distinct reasoning profiles. Initially, the AI produced an exhaustive, non-hierarchical analysis, correctly identifying septic shock among its top, similarly-scored hypotheses. Residents, in contrast, employed a pragmatic, focused strategy, prioritising immediate surgical risks and unanimously identifying an experience-based risk (gas embolism) that the AI systematically overlooked, and consistently reserved a portion of their reasoning for uncertainty, termed ‘Place for Doubt’. After the clinical evolution, both converged on septic shock. A ‘complex scrutiny’ analysis of the overall strategies revealed that the residents’ focused and adaptive reasoning was consistently ranked as strategically superior to the AI’s exhaustive but diluted approach. Conclusions: AI demonstrates a powerful capacity for broad diagnostic anticipation, acting as a potential safeguard against premature diagnostic closure . Experienced residents exhibit a strategically superior reasoning process in its focus and adaptation. Our findings support a powerful synergy where the AI serves as a ‘Cognitive Safety Net’ to augment, not replace, the contextualised judgment of the human practitioner. Research in Context: What is already known on this topic: Human error in healthcare is a global prominent cause of death. ‘Traditional’ cognitive support tools (e.g., paper checklists) have been shown to improve technical skills during medical crises, but their impact on non-technical skills is limited and their clinical adoption remains low. A new generation of customisable digital cognitive aids ( cDCAs ) can significantly improve both technical and non-technical performance, fostering better team management and crisis resolution. Information on how clinicians deliver the best anticipatory clinical reasoning is scarce. Recent work comparing machine-learning models to clinicians in trauma triage found comparable accuracy but only moderate agreement, suggesting a collaborative paradigm and motivating deeper analyses of the reasoning process itself. However, a critical gap remains in understanding the underlying nature of the diagnostic reasoning strategies that lead to these outcomes. The ‘how’ of human and AI reasoning, especially in dynamic, anticipatory clinical tasks, is not well understood. What this study adds: This is the first study to directly compare in action the diagnostic reasoning strategies o f clinicians and a large language model (AI). It introduces a novel analytical framework based on the Condorcet social choice method to move beyond simple performance scores and rigorously model and rank the overall quality of diagnostic strategies in a simulated daily complex situation. The findings support a model of human-AI complementarity , where the AI excels at broad, exhaustive analysis, while clinicians demonstrate a superior, focused, and adaptive strategic reasoning, suggesting the human’s role as a meta-cognitive supervisor of AI-driven exhaustive but ‘diluted’ insights.


  • Open Access

    Determining the influence of an intervention of stress management on medical students’ levels of psychophysiological stress: the protocol of the PROMESS-Stress clinical trial

    by Angèle Métais, Mohammad Omarjee, Bérénice Valero, Alexia Gleich, Aya Mekki, Amélie Henry, Antoine Duclos, Marc Lilot, Gilles Rode, Sophie Schlatter

    Medical students face highly competitive stressful situations throughout their curriculum, which can lead to elevated stress levels and a major decline in quality of life, well-being, learning abilities, and health. It is crucial to assist medical students in coping with these stressful situations during their curriculum. The PROMESS-Stress clinical trial aims to support future healthcare professionals by enhancing their abilities to manage stressful situations. The support will be provided through a 3-session stress management program. Each session will include an individual meeting between a PROMESS-Stress expert and a medical student. To reduce stress levels and enhance coping mechanisms, these sessions will focus on establishing personalized advice and goals. The present protocol is designed to assess the influence of this program on forty-five undergraduate medical students (4th and 5th-year) of the Lyon-Est Faculty of Medicine (Claude Bernard University Lyon 1, France). Assessments of psychological and physiological stress variables will be conducted before and during the intervention. At the end of the 3rd session the student’s levels of satisfaction will be assessed. The primary outcome will be changes in scores on the Perceived Stress Scale (PSS), while secondary outcomes will provide a detailed characterization of the intervention’s effects on stress coping behaviors, psychological and physiological stress variables. Exploratory outcomes will provide information regarding the student’s level of satisfaction and will determine the moderators of the program’s efficacy. Data will be analyzed according to the intention-to-treat principle and presented in accordance with the CONSORT Guidelines. Ethical approval has been obtained by the Institutional Review Board (IRB: 2023–07-04–02) and all the procedures will be performed in adherence to the Helsinki declaration. Results from this study will be presented at scientific conferences and in peer-reviewed scientific journals. Results will provide valuable insights into the program's efficacy in reducing stress and improving coping abilities. If its efficacy is proven, PROMESS-Stress could become an integral and sustainable part of medical education, fostering a healthier and more resilient future for healthcare professionals. This manuscript follows the SPIRIT guidelines (Additional files 1 & 6). ClinicalTrials.gov: NCT06295133; retrospectively registered.


  • Effect of coping interventions on performance of medical students during objective structured clinical examination: A randomized controlled trial

    by Sophie Schlatter, Alexandre Berland, Antoine Lutz, Rebecca Shankland, Noémie Barret, Theodore Guillaumée, Antoine Duclos, Marion Cortet, Thomas Rimmelé, Gilles Rode, et al.

    Objective Structured Clinical Examination (OSCE) is a stressful exam assessing medical competencies. Stress coping strategies are expected to enhance students' performance during OSCE. The objective was to determine the effect of short preventive coping interventions on performance of medical students. Double-blinded, randomized controlled trial with multiple arms and a superiority hypothesis. Enrolment was proposed to each fourth-year undergraduate medical student convened to the Lyon Est University OSCE in 2022. There was no exclusion criterion. Students were randomized to one of four groups: standardized breathing with cardiac biofeedback (BFB), mindfulness-based intervention (MBI), positive psychology intervention (PPI), or control (CTRL). Each intervention was video-guided, lasted six minutes, and occurred just before starting the OSCE. The primary outcome was the academic OSCE score, assessed through specific grids by university examiners blinded to the interventions. Secondary outcomes included specific performance scores, and student perception of the influence of the intervention on their performance. A total of 482 students were included. No difference was found between BFB (-0.17 [95%CI, -1.20 to 0.86], p = .749), MBI (0.32 [95%CI, -0.71 to 1.36], p = .540), or PPI groups (-0.25 [95%CI, -1.29 to 0.79], p = .637) on the academic OSCE score compared to the control group, nor regarding the specific performance scores. Compared to the control group, the students perceived that the intervention influenced more positively their performance (BFB +3 [95%CI, 0-8]), p < .001; MBI +4 [95%CI, 1-9], p = .040; PPI +1 [95%CI, 0-4], p = .040]). A single six-minute cardiac biofeedback, mindfulness, or positive psychology intervention performed by fourth-year medical students just before an OSCE did not improve their following academic performance. Still, students reported that the interventions helped them to enhance their performance. Future research should aim to further explore the perception of intervention on performance and potential long-term effects for students. The study protocol was registered on ClinicalTrials.gov Identifier: NCT05393219.


  • Open Access

    Association of Personality Traits With the Efficacy of Stress Management Interventions for Medical Students Taking Objective Structured Clinical Examinations

    by Olivia Le Saux, Brice Canada, Ursula Debarnot, Nour El Houda MsC Haouhache, Jean Jacques Lehot, Marion Msc Binay, Marion Cortet, Thomas Rimmelé, Antoine Duclos, Gilles Rode, et al.

    Purpose Personality traits are associated with psychophysiological stress, but few studies focus on medical students. This study aimed to better understand the association of personality traits with the efficacy of stress management interventions for medical students. Method A randomized controlled trial was conducted with fourth-year students who took the objective structured clinical examination at Bernard University Lyon 1 in December 2021. Students were randomized in cardiac biofeedback, mindfulness, and control groups. Each intervention was implemented for 6 minutes before the examination. Physiological stress levels were collected during the intervention. Psychological stress levels were rated by students at baseline and after the intervention. Personality traits were assessed via the Big-Five Inventory. Interactions between personality traits and the efficacy of the interventions were analyzed using multivariable linear regression models. Results Four hundred eighty-one students participated. Higher baseline psychological stress levels were associated with higher neuroticism and agreeableness (β = 10.27, 95% confidence interval (CI) [7.40, 13.13], P < .001 and β = 3.42, 95% CI [0.98, 5.85], P = .006, respectively) and lower openness (β = –4.95, 95% CI [–7.40, –2.49], P < .001). As compared to the control intervention, both stress management interventions led to lower levels of psychological (P < .001 for both) and physiological stress levels (biofeedback: P < .001 and mindfulness: P = .009). Biofeedback efficacy varied by extraversion score for psychological (β = –5.66, 95% CI [–10.83, –0.50], P = .03) and physiological stress reduction (β = –0.002, 95% CI [–0.003, –0.00004], P = .045). Mindfulness efficacy varied by agreeableness score for psychological stress reduction (β = –7.87, 95% CI [–13.05, –2.68], P = .003). Conclusions Students with a high score in extraversion may benefit more from biofeedback interventions, while students with high scores in agreeableness may benefit more from mindfulness interventions.