ISP25: health crisis management training for public health residents in France, a pilot and feasibility study Authors: Aimé Nun, Abdullah Azaz, Abdullah Azaz, Francisco Mourao, Julien Letellier, Lilia Sfihi, Pierre Cachot, Xavier Maynadier, Alexis Ardoin Published: 2026 ABSTRACT IntroductionPublic health residents play a critical role in crisis management but operate under administrative and population-level constraints that differ fundamentally from traditional clinical “bedside” Crisis Resource Management (CRM). There is a need to adapt existing training models to this specific institutional context.ObjectiveThis study aims to (i) describe a simulation-based training model specifically adapted to the French administrative context for public health residents; (ii) assess the feasibility and acceptability of this pilot program; and (iii) provide a translation and cultural adaptation of two assessment tools (the Ottawa CRM Global Rating Scale and the Modified Simulation Effectiveness Tool [SET-M]) for the Francophone community.MethodsWe conducted a prospective pilot feasibility study with a convenience sample of six public health residents (n = 6) in the Île-de-France region. The intervention combined a theoretical session on French health administration with a high-fidelity simulation exercise involving role-play in a crisis room. We collected descriptive data on participant satisfaction and self-reported skill progression using the French-adapted questionnaires.ResultsParticipants reported high satisfaction and strong acceptance of the training format, with all residents agreeing on the utility of the prebriefing and debriefing components. Descriptive analysis of individual trajectories showed positive trends in self-assessed leadership and communication skills. Conversely, a stagnation or decrease in problem-solving scores for some participants suggested a pedagogical “recalibration” of self-efficacy (potential mitigation of the Dunning-Kruger effect) following the immersive simulation.ConclusionThis pilot study demonstrates the feasibility and high acceptability of adapting CRM training for public health residents acting in administrative capacities. It provides the first French-adapted assessment tools for this specific population. Due to the limited sample size, future research with larger cohorts and controlled designs is necessary to formally evaluate the effectiveness of this training model.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12909-026-09117-7. Introduction Effective management of health crises is a critical aspect of public health, necessitating specialized skills and knowledge to address complex and often unpredictable situations [1–3]. The unique challenges posed by health crises, ranging from natural disasters to pandemics, highlight the importance of robust training and preparation framework for public health professionals. While Crisis Resource Management (CRM) training is well-established in clinical healthcare settings since the 1990s, it has traditionally focused on interdisciplinary teams managing patients at the bedside [4]. However, public health residents operate in a fundamentally different context: they must make decisions at the population level, often under strict administrative and political constraints, rather than focusing on individual clinical care. There is currently a lack of data on the application of CRM principles specifically for public health administrative leadership [5, 6]. This creates a specific need to adapt existing CRM models to the institutional realities faced by future public health officials. In France, public health residents are medical doctors specializing in public health who are frequently integrated into public administrations, such as the Regional Health Agencies (ARS). These agencies play a pivotal role in coordinating health policies and emergency responses. Consequently, these residents require training that bridges theoretical knowledge with the practical application of non-technical skills—such as leadership, communication, and resource utilization—within this specific administrative framework. This study presents the development and pilot implementation of a health crisis management training program for public health residents, conducted in collaboration with the ARS Île-de-France. Given the exploratory nature of this work, our objectives are: (i) to describe a training model specifically adapted to the French administrative context and the profile of public health residents; (ii) to assess the feasibility and acceptability of this training through preliminary participant feedback rather than definitive effectiveness metrics; and (iii) to provide a translation and cultural adaptation of two recognized assessment tools (the Ottawa CRM Global Rating Scale and the Modified Simulation Effectiveness Tool) to facilitate future research in the Francophone community. Methods Study design and participants This study was designed as a prospective, single-group, pre-post intervention pilot feasibility study. Participants were recruited using convenience sampling from a cohort of public health residents in the Île-de-France region. In the French medical education system, public health residents are medical doctors undergoing specialized training in epidemiology, health policy, and administration; unlike clinical residents, they primarily operate within government agencies rather than at the patient’s bedside. Six residents voluntarily participated in the training session. The sample size was determined by the availability of residents within the department at the time of the study, rather than by an a priori power calculation, consistent with the exploratory nature of this pilot. Participants had a broad range of backgrounds and were in their first to fourth years of residency. The session was held at the Regional Health Agency (ARS) in the Île-de-France region and was led by the Agency’s Defense and Security Department. Training session and simulation exercise The training session aimed to introduce participants to Crisis Resource Management (CRM) principles specifically adapted to the context of public health administration. The program consisted of two distinct components: The morning session included a series of theoretical lectures addressing the fundamental principles of health crisis management and the specific organizational structure of the health administration system in France. The objective was to furnish participants with the essential conceptual knowledge and foundational tools necessary for effective crisis response at a regional level. The afternoon session featured a simulation exercise designed to replicate a realistic health crisis scenario requiring population-level decision-making rather than individual patient care. Participants were required to engage in real-time strategic planning, interagency coordination, and complex problem-solving tasks under administrative constraints. Each participant was assigned a specific functional role in the crisis room. To mimic the hierarchical reality of a crisis center, roles requiring higher-level management skills (e.g., Room Manager) were assigned to the most senior residents, while operational roles were assigned to less experienced residents. Assessment questionnaires We used two established Likert scales to assess participant performance and perceptions regarding the simulation experience. Following the ISPOR (Professional Society for Health Economics and Outcomes Research) principles of good practice, we performed a translation and cultural adaptation of both questionnaires [7–10]. The adaptation process was conducted by two medical doctors specialized in public health and one security officer to ensure the terminology accurately reflected the French administrative crisis management context rather than a standard clinical setting. Psychometric validation was not performed given the pilot nature of the study. The Ottawa Global Rating Scale for Crisis Resource Management (CRM) is an internationally recognized tool that measures essential non-technical skills involved in crisis response, including leadership, teamwork, communication, situational awareness, decision-making, and resource utilization [11]. Each criterion is measured using a seven-level Likert scale [12, 13]. We used our French-adapted version (see Supplementary S1.1) to assess participants’ self-perceived skills both before and after the simulation to observe potential trajectories in self-efficacy. The Modified Simulation Effectiveness Tool (SET-M) is a questionnaire designed to gather participants’ perspectives on the quality and educational impact of the simulation, specifically addressing the briefing, debriefing, and immersive aspects [14, 15]. Each criterion was measured using a three-level Likert scale, ranging from “disagree” to “totally agree.” A linguistically adapted French version was distributed for feedback collection (see Supplementary S2.1). Crucially, items related to clinical knowledge and patient communication were modified or omitted to align with the population-level decision-making focus of the exercise, as opposed to patient bedside management [8, 16]. To encourage honest feedback, questionnaire responses were de-identified. Participants completed the CRM scale before and after the practical session, and the SET-M immediately after the debriefing. While efforts were made to maintain confidentiality, we acknowledge that true anonymity is difficult to guarantee in a small group setting with specific assigned roles. Statistical analysis Data analysis Data management and visualization were performed using R software (version 4.4.2; R Core Team, 2024) and the Tidyverse package [17]. Consistent with the study’s design as a pilot feasibility assessment with a limited sample size (n = 6), no inferential statistical hypothesis testing was conducted. We relied exclusively on descriptive statistics to summarize participant characteristics and identify trends in the data. For the CRM scale, the analysis focused on: Individual trajectories: We examined the change in self-assessment scores for each participant between the pre- and post-simulation phases to avoid the limitations of reporting mean scores for such a small sample.Skill recalibration (Overestimation): We identified instances where a participant’s post-simulation score was lower than their pre-simulation score. This negative difference was operationally defined as initial “overestimation” of skills, indicating a recalibration of self-efficacy following the practical exercise. Individual trajectories: We examined the change in self-assessment scores for each participant between the pre- and post-simulation phases to avoid the limitations of reporting mean scores for such a small sample. Individual trajectories: We examined the change in self-assessment scores for each participant between the pre- and post-simulation phases to avoid the limitations of reporting mean scores for such a small sample. Skill recalibration (Overestimation): We identified instances where a participant’s post-simulation score was lower than their pre-simulation score. This negative difference was operationally defined as initial “overestimation” of skills, indicating a recalibration of self-efficacy following the practical exercise. Skill recalibration (Overestimation): We identified instances where a participant’s post-simulation score was lower than their pre-simulation score. This negative difference was operationally defined as initial “overestimation” of skills, indicating a recalibration of self-efficacy following the practical exercise. For the SET-M scale, we calculated the frequency distribution of agreement levels (Likert 1–3) to assess satisfaction and perceived utility of the training components. Ethics approval and consent to participate According to local regulations and institutional guidelines, formal ethics committee approval was not required for this educational training simulation. However, the study was conducted in strict accordance with the ethical standards of the Declaration of Helsinki. Participation was strictly voluntary. Residents were explicitly assured that their participation—or refusal to participate—would have no impact on their academic grading or evaluations, thereby ensuring that no coercion was involved. Written informed consent was obtained from all participants prior to the session, following a comprehensive explanation of the exercise’s nature, expectations, and potential psychological stressors. Participants were informed of their right to withdraw from the simulation at any time without penalty should they experience undue distress. Regarding data privacy, participants were informed that their data would be de-identified and used exclusively for publication purposes. Strict confidentiality was maintained during data processing and reporting. While every effort was made to ensure confidentiality, the authors acknowledge that complete anonymity could not be fully guaranteed given the small group size and the specific functional roles assigned to each resident. Results Participants A total of six public health residents participated in this pilot study. The cohort consisted of five male residents and one female resident. The group possessed varying levels of seniority: three participants were in their first year of residency, one was in their second year, and two were in their fourth year. Regarding prior exposure to the field, while the majority of participants (n = 4) reported having some prior experience in health crisis management, five out of the six residents had no prior professional experience working within a Regional Health Agency (ARS) (Table 1). Table 1Participants descriptionCharacteristicsN (%)Gender Female1 Male5Residency year 1st year3 2nd year1 3rd year0 4th year2Prior experience in health crisis management Yes4 No2Prior experience at the Regional Health Agency Yes1 No5 Table 1Participants descriptionCharacteristicsN (%)Gender Female1 Male5Residency year 1st year3 2nd year1 3rd year0 4th year2Prior experience in health crisis management Yes4 No2Prior experience at the Regional Health Agency Yes1 No5 Participants description CRM questionnaire CRM questionnaire Regarding overall crisis management skills, all six participants reported an improvement in their self-assessment following the simulation. The median self-rating increased from 1.5 (range: 1–4) to 3.5 (range: 2–6) on the 7-point scale (Table 2; Fig. 1). Table 2CRM questionnaire score evolutionSkill areasPre-simulationMedian[Min-Max]Post-simulationMedian[Min-Max]Median Difference[Min-Max]Improved(n)Unchanged(n)Overall1.5 [1–4]3.5 [2–6]+ 1.5 [+ 1 to + 3]60Leadership4.5 [2–5]4.5 [3–6]+ 0.5 [-1 to + 3]31Problemsolving4.5 [3–6]4.5 [3–5]0 [-2 to + 1]22Situationalawareness4.0 [4–5]4.5 [3–5]0 [-1 to + 1]14Resourceutilization4.0 [3–5]4.5 [4–6]+ 0.5 [0 to + 1]33Communication4.0 [3–5]5.0 [3–6]+ 0.5 [0 to + 2]33Due to the small sample size (n=6), data are presented as medians with ranges [Min-Max] rather than means to reduce sensitivity to outliers"Lower Post-Score" indicates participants who rated their performance lower after the simulation than their initial pre-simulation assessment Table 2CRM questionnaire score evolutionSkill areasPre-simulationMedian[Min-Max]Post-simulationMedian[Min-Max]Median Difference[Min-Max]Improved(n)Unchanged(n)Overall1.5 [1–4]3.5 [2–6]+ 1.5 [+ 1 to + 3]60Leadership4.5 [2–5]4.5 [3–6]+ 0.5 [-1 to + 3]31Problemsolving4.5 [3–6]4.5 [3–5]0 [-2 to + 1]22Situationalawareness4.0 [4–5]4.5 [3–5]0 [-1 to + 1]14Resourceutilization4.0 [3–5]4.5 [4–6]+ 0.5 [0 to + 1]33Communication4.0 [3–5]5.0 [3–6]+ 0.5 [0 to + 2]33Due to the small sample size (n=6), data are presented as medians with ranges [Min-Max] rather than means to reduce sensitivity to outliers"Lower Post-Score" indicates participants who rated their performance lower after the simulation than their initial pre-simulation assessment CRM questionnaire score evolution Problem solving Situational awareness Resource utilization Due to the small sample size (n=6), data are presented as medians with ranges [Min-Max] rather than means to reduce sensitivity to outliers "Lower Post-Score" indicates participants who rated their performance lower after the simulation than their initial pre-simulation assessment Fig. 1CRM questionnaire score likert-scale CRM questionnaire score likert-scale For specific non-technical skills, descriptive statistics using medians and individual trajectories revealed a nuanced picture: Communication and Resource Utilization: These areas showed a positive trend. For both skills, half of the participants (n = 3) reported an improvement, while the other half (n = 3) maintained their initial ratings. No participants reported a decrease in these skills.Leadership: Results were mixed. The median score remained stable at 4.5. However, individual trajectories showed that while three participants noted an improvement in their leadership capabilities, one remained unchanged, and two rated themselves lower after the exercise than before (indicating initial overestimation).Problem-solving: The median score remained stable at 4.5, but individual variations were notable. Specifically, two participants rated their performance lower after the simulation compared to their pre-simulation belief (with one participant dropping by 2 points), suggesting a recalibration of their perceived competence when faced with the complexity of the scenario. Two others improved, and two remained stable.Situational Awareness: This skill showed the most stability, with a median difference of 0. One participant improved, one rated themselves lower, and the majority (n = 4) retained their pre-simulation score. Communication and Resource Utilization: These areas showed a positive trend. For both skills, half of the participants (n = 3) reported an improvement, while the other half (n = 3) maintained their initial ratings. No participants reported a decrease in these skills. Communication and Resource Utilization: These areas showed a positive trend. For both skills, half of the participants (n = 3) reported an improvement, while the other half (n = 3) maintained their initial ratings. No participants reported a decrease in these skills. Leadership: Results were mixed. The median score remained stable at 4.5. However, individual trajectories showed that while three participants noted an improvement in their leadership capabilities, one remained unchanged, and two rated themselves lower after the exercise than before (indicating initial overestimation). Leadership: Results were mixed. The median score remained stable at 4.5. However, individual trajectories showed that while three participants noted an improvement in their leadership capabilities, one remained unchanged, and two rated themselves lower after the exercise than before (indicating initial overestimation). Problem-solving: The median score remained stable at 4.5, but individual variations were notable. Specifically, two participants rated their performance lower after the simulation compared to their pre-simulation belief (with one participant dropping by 2 points), suggesting a recalibration of their perceived competence when faced with the complexity of the scenario. Two others improved, and two remained stable. Problem-solving: The median score remained stable at 4.5, but individual variations were notable. Specifically, two participants rated their performance lower after the simulation compared to their pre-simulation belief (with one participant dropping by 2 points), suggesting a recalibration of their perceived competence when faced with the complexity of the scenario. Two others improved, and two remained stable. Situational Awareness: This skill showed the most stability, with a median difference of 0. One participant improved, one rated themselves lower, and the majority (n = 4) retained their pre-simulation score. Situational Awareness: This skill showed the most stability, with a median difference of 0. One participant improved, one rated themselves lower, and the majority (n = 4) retained their pre-simulation score. SET-M questionnaire Participant satisfaction with the training session was high. Notably, no participant disagreed with any of the statements in the SET-M questionnaire, indicating strong acceptance of the pilot format (Fig. 2). Fig. 2SET-M questionnaire score likert-scale SET-M questionnaire score likert-scale Regarding the prebriefing, four participants “totally agreed” that it was beneficial for their learning, while the remaining two “somewhat agreed.” However, regarding the impact on their confidence, the majority (n = 4) “somewhat agreed” that the prebriefing helped, with two participants stating they “totally agreed.” For the simulation scenario, participants expressed the highest confidence in their theoretical understanding and communication. Four participants “totally agreed” that they developed a better understanding of health crisis management and felt more confident communicating with colleagues. Conversely, regarding specific operational skills—such as prioritizing tasks, applying knowledge acquired during the morning session, and overall preparedness to respond—ratings were positive but moderate; for these items, five participants “somewhat agreed” and one “totally agreed.” The debriefing phase received the strongest endorsement from the group. All six participants “totally agreed” that the debriefing provided a constructive evaluation of the simulation. Similarly, five participants “totally agreed” that the debriefing contributed to their learning and allowed them to express their feelings before focusing on the analytical aspects of the scenario. Finally, qualitative feedback from the open-ended question highlighted a structural observation: some participants reported periods of inactivity toward the end of the session, noting that engagement levels varied depending on the specific functional role assigned in the crisis room. Discussion Main results and feasibility This pilot study demonstrates the feasibility of implementing a simulation-based Crisis Resource Management (CRM) training program specifically adapted for public health residents. Six residents with varying levels of seniority participated. The high acceptability of the training is evidenced by the SET-M results, where all participants agreed or totally agreed with the utility of the prebriefing, the scenario, and the debriefing. Regarding skill acquisition, the analysis of individual trajectories reveals a nuanced picture. While overall confidence in crisis management increased for all participants—suggesting the training successfully empowered residents to engage in crisis response—specific complex domains showed different trends. Notably, while leadership and communication scores generally improved, scores for problem-solving and situational awareness remained stable or even decreased for some participants. This divergence may reflect a “recalibration” of self-efficacy. One possible explanation is that the theoretical session in the morning may have initially inflated participants’ confidence (providing conceptual knowledge), while the immersive simulation in the afternoon confronted them with the practical difficulty of applying these concepts under pressure. The decrease in self-rating for problem-solving observed in some participants could potentially be interpreted through the lens of the Dunning–Kruger effect [18, 19]. This cognitive bias suggests that individuals with limited experience may initially overestimate their ability (unconscious incompetence). The simulation likely facilitated a transition to “conscious incompetence,” where participants recognized the true complexity of the task and their own limitations. If this is the case, this realization, while numerically represented as a lower score or stagnation, would represent a critical pedagogical step in professional development, as noted by similar studies in medical education. Limitations This study is subject to several methodological limitations inherent to its design as a pilot feasibility project. Sample Size and Generalizability: With a small convenience sample () from a single region, our results are descriptive and cannot be generalized to the broader population of public health residents. The homogeneity of the training conditions (same day, same instructors) further limits external validity.Bias: Unlike larger randomized trials, this study is susceptible to selection bias, as participants were volunteers and likely more motivated than the average resident. Furthermore, given the small group size and assigned roles, complete anonymity was impossible to guarantee. This introduces a potential social desirability bias, where participants may have rated the training or their own improvement more favorably to please the instructors, although the presence of decreased scores in some domains suggests this bias was not absolute.Assessment limitations: The pre- and post-assessments occurred on the same day. Consequently, we measured immediate reaction and short-term self-perception rather than long-term retention or behavioral change in real-world settings. Future research must include longitudinal follow-up to assess whether these “lessons learned” translate into professional practice. Sample Size and Generalizability: With a small convenience sample () from a single region, our results are descriptive and cannot be generalized to the broader population of public health residents. The homogeneity of the training conditions (same day, same instructors) further limits external validity. Sample Size and Generalizability: With a small convenience sample () from a single region, our results are descriptive and cannot be generalized to the broader population of public health residents. The homogeneity of the training conditions (same day, same instructors) further limits external validity. Bias: Unlike larger randomized trials, this study is susceptible to selection bias, as participants were volunteers and likely more motivated than the average resident. Furthermore, given the small group size and assigned roles, complete anonymity was impossible to guarantee. This introduces a potential social desirability bias, where participants may have rated the training or their own improvement more favorably to please the instructors, although the presence of decreased scores in some domains suggests this bias was not absolute. Bias: Unlike larger randomized trials, this study is susceptible to selection bias, as participants were volunteers and likely more motivated than the average resident. Furthermore, given the small group size and assigned roles, complete anonymity was impossible to guarantee. This introduces a potential social desirability bias, where participants may have rated the training or their own improvement more favorably to please the instructors, although the presence of decreased scores in some domains suggests this bias was not absolute. Assessment limitations: The pre- and post-assessments occurred on the same day. Consequently, we measured immediate reaction and short-term self-perception rather than long-term retention or behavioral change in real-world settings. Future research must include longitudinal follow-up to assess whether these “lessons learned” translate into professional practice. Assessment limitations: The pre- and post-assessments occurred on the same day. Consequently, we measured immediate reaction and short-term self-perception rather than long-term retention or behavioral change in real-world settings. Future research must include longitudinal follow-up to assess whether these “lessons learned” translate into professional practice. Perspectives Despite these limitations, this study addresses a critical gap identified in the literature: the lack of CRM training specifically tailored to the administrative and population-based decision-making context of public health, as opposed to the clinical “bedside” focus of traditional CRM. By providing a translated and culturally adapted version of the Ottawa CRM Scale and SET-M, this work offers a methodological foundation for future Francophone research. Future iterations should aim for larger, multi-centric cohorts and utilize objective external observers alongside self-assessment to provide a more rigorous evaluation of training effectiveness. Transferability and contextualization Transferability of the training model To date, this initiative represents a novel proposal for crisis management training specifically tailored to public health residents acting in administrative capacities. While Crisis Resource Management (CRM) is well-established in clinical settings, its application to the specific constraints of the French health administration—where decisions are made at a population level rather than at the bedside—constitutes a specific pedagogical niche [3]. Consequently, the specific roles and scenarios designed for this study may not be directly transferable to countries with different health system architectures without local adaptation. Nevertheless, the core pedagogical approach—combining theoretical instruction with high-fidelity simulation—aligns with the broader literature. Established models, such as the Health Crisis Management Framework, consistently show that simulation enhances non-technical skills like decision-making and teamwork. Although our sample size (n = 6) prevents generalizing our findings on effectiveness, our observations of high participant satisfaction mirror results from larger studies. For instance, Escribano et al. demonstrated significant improvements in resilience and self-efficacy among health science students following CRM training, and Gartland et al. found similar benefits for leadership skills in emergency medicine residents [20, 21]. These parallels suggest that the feasibility and acceptability observed in our French pilot could potentially be replicated in other contexts, provided the scenario is adapted to local institutional needs [2]. Looking forward, the ongoing global health crisis has underscored the need for flexible educational models. As noted by a study from Shiraz University of Medical Sciences, personalized, technology-enhanced education is essential for meeting pandemic-related challenges. Integrating virtual reality or remote simulation components into future iterations of this training could not only enhance its accessibility but also improve its external validity by facilitating multi-center studies with larger cohorts. Relevance of the adapted evaluation tools A key contribution of this study is the translation and cultural adaptation of the CRM scale and SET-M for the Francophone public health context [7]. While the original English versions are internationally recognized and validated for clinical environments, their application to public health leadership remains limited [9, 14, 16]. Unlike clinicians managing individual patients, public health residents operate under distinct administrative and political constraints that require a specific lexicon and skill set [22]. By adapting these tools to exclude purely clinical items (e.g., patient communication) and focusing on population-level coordination, we provide a resource for the broader community. The SET-M, in particular, proved valuable for collecting structured feedback on the simulation design, identifying specific areas for pedagogical refinement [23–26]. This combined approach—adapting established instruments to a new linguistic and professional context—serves as a model for other competency-based public health training programs seeking to evaluate non-technical skills rigorously [8, 14–16]. Perspectives The preliminary results of this pilot study suggest that a simulation-based training model specifically adapted for public health residents in an administrative context is both feasible and highly acceptable to learners. While the small sample size precludes definitive conclusions regarding effectiveness, these findings generate several hypotheses to guide future research and development efforts. Expanding study design and sample size To address the limitations of generalizability and statistical power identified in this pilot, future research must transition from descriptive analysis to rigorous evaluation. Expanding the sample size through multi-centric collaborations with other Regional Health Agencies would mitigate the homogeneity bias inherent in a single-center study. Furthermore, as recommended by the reviewers, future studies should employ a controlled design with a comparison group to isolate the specific impact of the training from general professional development or social desirability biases. Technological integration Incorporating advanced technology and virtual components into training sessions represents a key avenue for development, as highlighted by recent literature on pandemic preparedness. Virtual simulations and e-learning modules could offer flexible training options, overcome geographical constraints and allowing for the standardization of scenarios across different regions. This would also facilitate the implementation of large-scale studies necessary for robust statistical analysis. Longitudinal assessment A major limitation of the current study was the immediate post-training assessment. To determine the true pedagogical value of the program, future iterations should incorporate long-term follow-ups. Assessing participants several months after the training would allow researchers to evaluate skill retention and, crucially, the transfer of these non-technical skills to actual professional practice during real health crises. Psychometric validation of tools Finally, this study provided the initial step of translating and culturally adapting the Ottawa CRM scale and SET-M for the French administrative context. However, full psychometric validation (assessing internal consistency, factor analysis, and inter-rater reliability) was not possible with the current sample size. A dedicated methodological study with a larger cohort is now required to formally validate these French versions, thereby providing the Francophone public health community with robust metrics for future educational research. Conclusion In conclusion, this study describes the development and pilot implementation of a health crisis management training model specifically tailored to public health residents acting in administrative capacities—a population distinct from the clinical teams traditionally targeted by Crisis Resource Management (CRM). The initial findings demonstrate that this pedagogical approach is feasible and meets the felt needs of residents for practical, decision-based training. While the descriptive data suggests positive trends in learner confidence and engagement, the exploratory nature of this work requires caution in interpreting efficacy. This pilot serves as a foundational step, providing adapted assessment tools and a tested scenario design. Future research must now build on this groundwork by conducting larger, controlled studies with longitudinal follow-up to definitively establish the effectiveness of this training in improving the resilience and response capabilities of the future public health workforce. Supplementary Information Supplementary Material 1. Supplementary Material 1.