Emergency medicine is characterized by frequent exposure to critical situations that demand rapid and accurate decision-making, generating significant emotional and cognitive load for healthcare professionals (Boscolo et al., 2025; Janicki et al., 2020). Stress responses are mediated by prefrontal cortex functioning (Arnsten, 2009; Girotti et al., 2018; Plessow et al., 2011; Qin et al., 2009; Shields et al., 2016), and numerous reports highlight globally (Loeppke et al., 2017; Xiao et al., 2014) the high prevalence of occupational stress and burnout in emergency departments (EDs) worldwide—an issue also strongly evidenced among general medical students (Queirolo et al., 2024c; Quek et al., 2019). Among the most stressful events, cardiopulmonary resuscitation (CPR) represents one of the most demanding scenarios, requiring coordinated team effort under pressure, precise execution of resuscitation maneuvers, and effective communication management.

An effective team is essential during CPR to improve outcomes in patients experiencing cardiac arrest. Several studies debate about the optimal configuration, with some indicating that a four-member team represents an optimal configuration (Hunziker et al., 2011), others that teams from 4 up to 6 people are better (Tsai et al., 2020; Han et al., 2022).

Clear communication and effective task distribution among the roles are crucial for the success of the intervention (Nallamothu et al., 2018). Moreover, in cases of out-of-hospital cardiac arrest, teams of this size have been observed to perform more effectively in resuscitation maneuvers compared to larger or smaller teams, avoiding delays in the chain of survival (Martin-Gill et al., 2010). The adoption of structured protocols and the continuous training of team members remain key factors in optimizing emergency response and improving patient survival rates (Maconochie et al., 2020).

Studies have shown that negative emotions were inversely correlated with clinical performance (Xi et al., 2025), while acute stress can negatively impact clinical performance (LeBlanc, 2009; Russ et al., 2018), affecting both technical skills—such as the quality of chest compressions—and non-technical skills, including decision-making and leadership abilities (Hunziker et al., 2012).

To assess stress levels in healthcare providers involved in CPR, various psychometric tools have been employed, including the Spielberger State–Trait Anxiety Inventory (STAI-6) and the Perceived Stress Scale (PSS) (Şener et al., 2025), along with physiological parameters such as heart rate (HR), heart rate variability (HRV); a proxy of prefrontal cortex functions (Beauchaine and Thayer, 2015; Kim et al., 2018; Thayer et al., 2009), and electrodermal activity a proxy of the flight or fight system (EDA) (Queirolo et al., 2024a; Claverie et al., 2020).

EDA and HRV have shown interesting abilities to predict and monitor stress (Rahma et al., 2022; Dalmeida and Masala, 2021).

The literature suggests that the adoption of stress management strategies—such as post-event debriefing (Kam et al., 2022), high-fidelity simulation-based training, the use of breathing and visualization techniques (Lauria et al., 2017; Rudaz et al., 2017) and hypnosis—can improve the wellbeing (Fisch et al., 2020) of healthcare professionals.

The aim of this study is to analyze the relationship between stress and performance during CPR. To our knowledge, no previous study has simultaneously assessed psychological factors, HRV, EDA, and SCR in the context of simulated CPR, while also distinguishing between team roles. This integrative approach offers a novel lens on how physiological stress markers relate to performance under pressure that are essential to optimize the quality of care in critical settings by maximizing decision making and to reduce the risk of burnout among healthcare providers (Aronson et al., 2022; Sarmiento et al., 2024).

The sample consisted of 30 participants with a mean age of 30 years (SD = 2.74).

Moderate levels of stress were observed across participants at baseline, as indicated by physiological (SCR = 6.193 ± 3.23) and subjective measures (VAS-Stress = 3.89 ± 1.93; VAS-Anxiety = 3.93 ± 1.98; PSS-10 = 16.67 ± 6.64). No significant differences were found in VAS stress or anxiety scores between pre- and post-CPR conditions.

Baseline physiological values were within normative ranges: HR (79.77 ± 13.92 bpm), HRV (RMSSD = 45.51 ± 18.89 ms) and EDA (Median = 1.4800 [0.72–3.03] 95% CI bound Bootstrap median).

Friedman’s ANOVA showed significant differences in EDA across conditions (χ² = 6.20, p < 0.05); more specifically an increase from baseline to CPR (p < 0.001) and an increase from baseline to Post CPR (p < 0.001) (see Figure 3). ANOVA also revealed a significant effect of condition on HRV (p < 0.05), with TL showing higher HRV than TM post CPR (see Figure 4). Friedman’s ANOVA showed a significant effect of condition on HR (χ² = 22.30, p < 0.001), with post hoc comparisons showing HR increased during CPR (p < 0.001) and decreased post-CPR (p < 0.001) (see Figure 5).

No significant group differences emerged between team leaders (TL) and team members (TM) in any physiological measure, with the solely exception for HRV (see Figure 4).

However, a significant direct correlation was found between TL’s pre-CPR EDA and TM’s post-CPR SCR (R = 0.62, df = 2, 13, p < 0.05).

Regression models showed that no linear regression significantly predicted CPR performance. A binomial logistic regression predicting high performance (>7.5) using sympathetic data showed a good fit (TM EDA Pre, TL EDA Pre and TL SCR pre simulation R² McFadden = 0.39, AIC = 19.80, p < 0.05, accuracy = 86.7%).

More notably, a nonlinear (quadratic) regression using SD1 significantly predicted CPR performance (R² Mc Fadden = 0.56), displaying a clear inverted U-shaped curve. Peak predicted performance (CPR = 9.17) was observed at SD1 = 0.0201. Both the quadratic and linear terms were statistically significant (a = −15748.2, p < 0.01; b = 631.81, p < 0.01; intercept = 2.88), indicating that performance decreased when SD1 deviated above or below this optimal point (see Figure 4).

The present study investigated psychophysiological stress responses and performance during high-fidelity CPR simulations in emergency medicine, with a focus on the role within the team (leader vs. member) and the potential for predicting performance through physiological markers.

Stress is commonly reported in emergency medicine (Janicki et al., 2020) and is often linked to decreased performance (Harvey et al., 2012), but the relationship between stress and performance is complex and not yet fully understood (Groombridge et al., 2022; Vincent et al., 2021).

Despite moderate baseline stress levels reported by participants, no significant changes were observed in subjective stress or anxiety ratings before and after the simulation. This dissociation between subjective and physiological stress is consistent with previous findings suggesting that individuals may not consciously perceive or report physiological arousal during acute stress, especially in high-demand clinical contexts (Boucher and Plusquellec, 2019; Qi et al., 2016).

Physiological data, however, revealed significant changes in HR, HRV and EDA during CPR, followed by partial recovery post-simulation (with the exception for EDA). These results confirm the validity of the emotional regulation processes described by the biopsychosocial model of arousal regulation that postulates a rise in both sympathetic and parasympathetic tone in response to challenges perceived as manageable (Blascovich and Tomaka, 1996). The activation of the autonomic nervous system in response to simulated resuscitation scenarios, was marked and supports the ecological validity of simulation-based stress induction. The persistence of elevated EDA post-simulation may indicate sustained sympathetic arousal even after the critical task has ended (claiming space for more training and experience needed to fully master the simulation with excellent emotional regulation).

Interestingly, no significant differences were observed between TL and TM in physiological stress responses, suggesting that both roles entail comparable cognitive and emotional demands within the structured simulation setting. The exception was HRV, which may function as a modulator of the broader context, potentially reflecting the pivotal role of the TL in the clinical scenario as a modulator. This finding challenges common assumptions that leadership roles inherently entail greater stress (in magnitude). Notably, differentiates the two roles the pre-simulation SCR in team leaders, which negatively correlated with CPR performance—indicating that higher fast sympathetic activation was associated with poorer outcomes, possibly reflecting a difference based on the quality of stress response rather than quantity. This pattern suggests a predominant role of rapid, phasic sympathetic reactivity in Team Leaders in shaping performance outcomes, beyond baseline tonic arousal.

On the other hand a significant association emerged between TL tonic sympathetic activation; TL EDA pre-CPR and TM post-CPR EDA, suggesting a possible mechanism of emotional transmission or synchronization within the team, supporting a complementary interpretation, whereby tonic sympathetic readiness in leaders may interact with fast phasic responses to influence subsequent team-level autonomic regulation. Overall Omnibus likelihood ratio tests indicated that Team Leader phasic sympathetic activation was the most significant individual predictor of CPR performance, whereas Team Leader and Team Member tonic sympathetic activity (EDA pre-CPR) showed marginal associations, suggesting that post-CPR TM SCR reflected the TL overall modulation.

Together, these findings indicate that TL SCR may act as a proximal physiological driver of performance, while tonic sympathetic indices contribute to a broader regulatory context, potentially mediating interpersonal synchronization within the team. This interpersonal physiological linkage may reflect the qualitative nature of stress regulation and coordination, rather than a mere difference in stress magnitude. The unique predictive value of TL SCR could thus represent the added cognitive and emotional load borne by leaders due to their need for rapid decision-making under pressure, highlighting the unique physiological cost of leadership during team-based resuscitation. The phasic sympathetic reactivity of team leaders may shape the subsequent tonic sympathetic level in team members, suggesting a form of emotional transference or contagion within the team dynamic. This unique marker may reflect the equal accountability embedded in collaborative team-based simulation and the most prominent relevance of fast action accountability for TL.

Notably, a quadratic relationship between SD1 and CPR performance was identified and its relevance in a very narrow amount of time (90 s) as previously highlighted by the scientific literature (Hunziker et al., 2011). This nonlinear model suggests that both very low and very high autonomic variability are associated with poorer performance, with peak performance observed at SD1 = 0.0201, confirming the existence of an optimal HRV window for task and adaptive regulation. This aligns with the neurovisceral integration model (Thayer and Lane, 2000), and the Yerkens and Dodson’s model (Yerkes and Dodson, 1908) (the reversed U shape between performance and anxiety here highlighted by SD1) which joint together propose that efficient physiological regulation supports cognitive flexibility and executive function—key components in managing dynamic clinical scenario as well.

The absence of predictive value in linear models further emphasizes the importance of considering nonlinear, binomial and individualized physiological patterns as already previously underlined by scientific literature, where experience is not enough (Krage et al., 2014) and subjective perception is relevant (Vincent et al., 2021); training it is necessary when dealing with performance in complex tasks. These findings underscore the potential for HRV and EDA as biomarkers of functional readiness, though further validation is needed in real-world clinical settings.

To our knowledge, this is the first study to simultaneously examine multiple physiological stress markers (HR, HRV, SD1, EDA, SCR) and psychological one, within role-differentiated CPR teams, highlighting the unique predictive contribution of parasympathetic and sympathetic reactivity in explaining CPR outcomes variability revealing their relevance as meaningful predictors of performance.

This study demonstrates that high-fidelity CPR simulation elicits robust psychophysiological responses that are not fully reflected in subjective stress reports, highlighting the value of physiological measures in high-demand clinical tasks. Although Team Leaders and Team Members showed comparable levels of autonomic activation, differences in autonomic regulation, rather than stress magnitude, were critical for performance. Phasic sympathetic reactivity in Team Leaders (SCR) emerged as the strongest physiological predictor of CPR outcomes, underscoring the importance of rapid autonomic dynamics during time-critical decision-making. Tonic sympathetic activity contributed to a broader regulatory context and was associated with subsequent autonomic modulation within the team. In addition, performance followed a nonlinear relationship with parasympathetic regulation, with optimal outcomes observed within a narrow SD1 window. Together, these findings suggest that effective team-based resuscitation depends on the timing and coordination of autonomic responses, providing novel insights into the physiological mechanisms underlying leadership and performance in emergency care.