Effect of rescuer heart rate changes on the quality of infant CPR: a simulation-based study Authors: Da Hyun Kim, Seung Jun Choi Published: 2025 ABSTRACT AbstractBackgroundThis study aimed to evaluate whether increases in heart rate—used as an objective surrogate marker of rescuer fatigue—could influence CPR performance during infant chest compressions.MethodsThis study was a manikin-based simulation study that enrolled PALS-certified pediatric emergency nurses matched by clinical experience and randomly assigned to three groups. All participants performed three 2-min cycles of infant chest compressions on a manikin (Little Baby QCPR). Each group completed a 90-s exercise protocol at different time points to induce varying degrees of heart-rate elevation. The primary outcome was the difference in CPR quality across groups according to heart-rate variation (percentage increase from baseline).ResultsTwenty-seven nurses were enrolled and evenly allocated to Groups A, B, and C (n = 9 each). Heart rate increased immediately after exercise in all groups and gradually declined during subsequent compression cycles, with no significant between-group differences at any time point (baseline 85–100 bpm/cycle, peak 130–150 bpm/cycle). Across all cycles, CPR performance metrics—including total compression count, hand-placement accuracy, mean compression velocity, compression depth, and chest recoil—showed no significant within-group changes for any group. Similarly, no significant between-group differences were observed for any CPR parameter during any cycle. In post-hoc analyses, stratification by the median percentage increase in heart rate (> 67.7% vs. < 67.7%) revealed no statistically significant differences in CPR quality between groups.ConclusionWithin this short, three-cycle simulation, exercise-induced heart-rate elevation was not associated with measurable deterioration in infant CPR quality.Keywords: Cardiopulmonary resuscitation, Infant, Fatigue Introduction Effective chest compressions are a cornerstone of high-quality cardiopulmonary resuscitation (CPR) and are essential for improving survival outcomes [1, 2]. The American Heart Association (AHA) has endorsed hands-only CPR as a simplified, compression-only technique for lay rescuers, recognizing its role in increasing bystander intervention rates [1, 3]. However, chest compressions are physically demanding, and sustained exertion can lead to rescuer fatigue, potentially compromising compression quality. Fatigue and emotional stress are well-established determinants of impaired task performance, reduced accuracy, and diminished operational efficiency [4, 5]. Importantly, understanding fatigue-related physiological strain has implications beyond performance measurement alone. Evidence regarding when and how fatigue emerges during CPR may inform the structure of pediatric cardiac arrest training programs, including scenario duration, rest intervals, and rescuer rotation strategies. In particular, pediatric resuscitation training often involves nurses as first responders, whose performance may be influenced not only by physical exertion but also by occupational stress and emotional burden [6, 7] Heart rate is frequently used as an objective marker of fatigue across various disciplines [8, 9], as sympathetic activation leads to measurable increases in heart rate that reflect both emotional stress and physical strain [8]. Adult CPR studies have reported inconsistent conclusions regarding the timing and degree of deterioration in chest compression depth, a surrogate marker indirectly reflecting rescuer fatigue [10–16]; however, to our knowledge, no prior study has explicitly examined whether exercise-induced heart rate elevations in trained pediatric nurses translate into measurable deterioration in infant CPR performance. Given this gap, this study sought to evaluate CPR quality during infant chest compressions by experimentally inducing heart rate elevation as a proxy for sympathetic activation. We aimed to determine whether increases in heart rate—used as an indirect surrogate marker of fatigue—adversely affect the quality of infant CPR. Methods Participants Nurses working in the pediatric emergency department of a tertiary academic referral hospital who had completed Pediatric Advanced Life Support (PALS) training were recruited. Individuals who were pregnant or had health conditions that could interfere with participation were excluded. Institutional Review Board approval (#2025–0824) was obtained before study initiation, and written informed consent was obtained from all participants. After enrollment, participants were matched in blocks by years of clinical experience and then randomly assigned to one of three groups. Simulation protocol All participants completed three cycles of infant chest compressions on an infant manikin. The manikin was placed on a firm surface at standard bed height, and compressions were performed using either the two-finger or two-thumb encircling hands technique, in accordance with PALS guidelines. Of note, all participants ultimately performed chest compressions using only the two-thumb encircling hands technique. All participants were blinded to group allocation throughout the study. Before the simulation, participants received standardized instructions regarding the compression technique and study procedures. No feedback was provided during CPR performance to avoid introducing learning effects. Each group performed a brief period of physical exercise at a different time point to induce changes in heart rate. Group A exercised before the first cycle, Group B between the first and second cycles, and Group C between the second and third cycles. Each chest compression cycle lasted 2 min, followed by a 1-min rest period. For the group scheduled to perform exercise before the next cycle, participants engaged in 1 min and 30 s of jumping jacks, after which their heart rate was measured, and compressions began immediately. In the remaining groups, compressions were commenced immediately after the 1-min rest period. Briefly, the protocol comprised 2-min chest compression cycles separated by 1-min rest periods; for the designated groups, a 90-s jumping jack exercise was performed at predefined time points, followed immediately by the subsequent 2-min compression cycle, for a total protocol duration of approximately 10 min (Fig. 1). Inter-rater training was conducted among evaluators before the simulation to standardize assessments. Schematic diagram of the simulation Data collection Infant CPR was performed using the Laerdal Little Baby™ manikin (Laerdal Medical, Stavanger, Norway), which is equipped with a QCPR sensor capable of recording CPR performance metrics at the end of each cycle. The following parameters were collected for every cycle: total number of compressions, hand-placement accuracy, mean compression velocity (number of chest compressions per CPR cycle), compression depth, and degree of recoiling. Using the QCPR system, hand placement accuracy was defined as the percentage of compressions delivered at the correct hand position, and recoil was defined as the percentage of compressions achieving complete chest recoil. Heart rate was measured before and after each exercise and each compression cycle using a Nellcor™ pulse oximeter (Medtronic, Minneapolis, MN, USA). The primary outcome was the difference in CPR parameters among the three groups according to heart rate variation. The secondary outcome was the difference in CPR parameters according to the magnitude of heart rate change. Statistical analysis Continuous variables were summarized as mean ± standard deviation (SD), and categorical variables as frequencies and percentages. Differences in continuous variables, including heart rate and CPR parameters among the three groups, were evaluated using one-way analysis of variance (ANOVA) and independent t-tests, as appropriate. Post-hoc analysis was performed using Tuckey’s method. Given the small sample size and the exploratory nature of this simulation study, parametric tests were used based on the assumption of approximate normality. A p-value of < 0.05 was considered statistically significant. All analyses were performed using SPSS software (IBM SPSS Statistics, version 21.0; IBM Corp., Armonk, NY, USA). Results Baseline characteristics Of the 30 candidates, three were excluded due to health-related issues. A total of 27 PALS-certified pediatric emergency nurses were enrolled and evenly assigned to Groups A, B, and C (n = 9 per group) using matched randomization based on years of clinical experience. Most participants were female, and no significant differences were observed among the three groups regarding sex distribution or clinical experience (Table 1). Table 1Baseline characteristics of participants¶VariablesGroup A(N = 9)Group B(N = 9)Group C(N = 9)P-valueFemale, n (%)9 (100)8 (89)8 (89)0.583Clinical experience, n (%)1.0000–4 (yr)4 (44)4 (44)4 (44)5–94 (44)4 (44)4 (44) ≥ 101 (11)1 (11)1 (11)¶Participants were matched in blocks by years of clinical experience and randomly assigned to one of three groups Baseline characteristics of participants¶ ¶Participants were matched in blocks by years of clinical experience and randomly assigned to one of three groups Changes in heart rate before and after exercise and chest compressions Across all groups, heart rate increased immediately after the exercise protocol and gradually declined during subsequent CPR cycles (Fig. 2). Although transient heart rate elevations were observed after each exercise period, no significant between-group differences in heart rate were identified at any measurement point. Overall, participants demonstrated similar heart rate trajectories throughout the simulation, suggesting comparable levels of exertion. Heart rate variation across the three cycles Comparison of CPR parameters within groups Total compression count, hand-placement accuracy, mean compression velocity, compression depth, and chest recoiling across the first, second, and third cycles were analyzed. The average chest compression depth was 43–44 mm, consistent with current infant CPR guideline recommendations [17]. In Group A, no significant differences were observed in total compression count, hand-placement accuracy, mean compression velocity, compression depth, or chest recoiling across the first, second, and third cycles. The same pattern was observed in Groups B and C, where none of the CPR performance metrics demonstrated significant within-group variation (Table 2). One-way ANOVA showed no significant differences in CPR quality metrics across cycles (all p > 0.05). Tukey post-hoc analyses demonstrated small between-cycle mean differences, with all 95% confidence intervals crossing zero. Table 2Comparison of CPR parameters within each group1st2nd3rdp-valueA Number of compression216.67 ± 18.01209.00 ± 24.30210.11 ± 24.120.737 Accuracy (%)90.22 ± 20.5283.22 ± 33.4785.33 ± 29.420.866 Mean velocity†109.78 ± 9.28106.56 ± 12.75106.56 ± 12.220.793 Depth (mm)43.33 ± 3.4643.22 ± 3.3143.22 ± 3.070.997 Recoiling (%)41.44 ± 42.2942.00 ± 40.4332.67 ± 39.180.860B Number of compression216.00 ± 23.74220.11 ± 23.26219.89 ± 24.150.918 Accuracy (%)77.56 ± 34.9566.33 ± 38.7872.67 ± 39.880.822 Mean velocity109.89 ± 12.13111.78 ± 11.6111.22 ± 12.220.943 Depth (mm)44.56 ± 1.1344.11 ± 1.6943.89 ± 1.540.625 Recoiling (%)34.44 ± 36.3938.89 ± 37.1627.89 ± 35.740.814C Number of compression224.33 ± 20.39224.44 ± 20.19226.78 ± 21.530.961 Accuracy (%)58.89 ± 44.6764.56 ± 42.1758.78 ± 38.540.946 Mean velocity114.00 ± 10.71113.33 ± 9.89115.11 ± 10.960.937 Depth (mm)43.56 ± 2.9644.44 ± 1.7444.33 ± 1.660.652 Recoiling (%)59.67 ± 39.3961.89 ± 38.5866.78 ± 30.310.914Values are presented as mean ± SD. Normally distributed variables were compared using one-way ANOVAAbbreviation: CPR cardiopulmonary resuscitation†Mean velocity: number of chest compression per CPR cycle Comparison of CPR parameters within each group Values are presented as mean ± SD. Normally distributed variables were compared using one-way ANOVA Abbreviation: CPR cardiopulmonary resuscitation †Mean velocity: number of chest compression per CPR cycle Comparison of CPR parameters among groups When comparing Groups A, B, and C at each cycle, there were no statistically significant differences in total number of compressions, hand-placement accuracy, mean compression velocity, compression depth, or chest recoiling percentage (Table 3). These findings indicate that CPR quality remained comparable across the three groups regardless of the timing of exercise-induced fatigue. Tukey post-hoc analyses demonstrated minimal between-cycle differences, with all corresponding 95% confidence intervals crossing zero. Table 3Comparison of CPR parameters among groupsABCP-value1st Number of compression216.67 ± 18.01216.00 ± 23.74224.33 ± 20.390.647 Accuracy (%)90.22 ± 20.5277.56 ± 34.9558.89 ± 44.670.180 Mean velocity†109.78 ± 9.28109.89 ± 12.13114.00 ± 10.710.643 Depth (mm)43.33 ± 3.4644.56 ± 1.1343.56 ± 2.960.602 Recoiling (%)41.44 ± 42.2934.44 ± 36.3959.67 ± 39.390.3892nd Number of compression209 ± 24.30220.11 ± 23.26224.44 ± 20.190.345 Accuracy (%)83.22 ± 33.4766.33 ± 38.7864.56 ± 42.170.530 Mean velocity106.56 ± 12.75111.78 ± 11.6113.33 ± 9.890.435 Depth (mm)43.22 ± 3.3144.11 ± 1.6944.44 ± 1.740.536 Recoiling (%)42.00 ± 40.4338.89 ± 37.1661.89 ± 38.580.4073rd Number of compression210.11 ± 24.12219.89 ± 24.15226.78 ± 21.530.330 Accuracy (%)85.33 ± 29.4272.67 ± 39.8858.78 ± 38.540.316 Mean velocity106.56 ± 12.22111.22 ± 12.22115.11 ± 10.960.323 Depth (mm)43.22 ± 3.0743.89 ± 1.5444.33 ± 1.660.567 Recoiling (%)32.67 ± 39.1827.89 ± 35.7466.78 ± 30.310.056Values are presented as mean ± SD. Normally distributed variables were compared using one-way ANOVAAbbreviation: CPR cardiopulmonary resuscitation†Mean velocity: number of chest compression per CPR cycle Comparison of CPR parameters among groups Values are presented as mean ± SD. Normally distributed variables were compared using one-way ANOVA Abbreviation: CPR cardiopulmonary resuscitation †Mean velocity: number of chest compression per CPR cycle Comparison of CPR parameters stratified by heart rate change Participants were additionally stratified into two groups based on the median percentage change in heart rate before and after exercise (cut-off: 67.7%). Thirteen participants were assigned to the high heart rate-change group (> 67.7%) and thirteen to the low heart rate-change group (< 67.7%). No significant differences were observed between the two groups for any CPR quality metric, including compression count, accuracy, mean velocity, compression depth, or chest recoiling percentage (Table 4). Post-hoc Tukey analyses indicated minimal between-cycle differences, with all 95% confidence intervals crossing zero. Table 4Comparison of CPR parameters among groups based heart rate changesGroup 1(n = 13)Group 2(n = 13)P-value1st Number of compression221.08 ± 19.92221.08 ± 15.881.000 Accuracy (%)72.92 ± 38.5379.08 ± 35.790.677 Mean velocity†112.31 ± 10.4112.23 ± 8.270.984 Depth (mm)43.62 ± 3.444.00 ± 1.920.726 Recoiling (%)55.31 ± 37.7737.92 ± 40.660.2702nd Number of compression215.31 ± 26.88223.31 ± 15.490.364 Accuracy (%)68.46 ± 40.2776.15 ± 37.260.618 Mean velocity108.92 ± 13.26113.62 ± 8.050.288 Depth (mm)43.92 ± 2.9643.92 ± 1.751.000 Recoiling (%)59 ± 34.6234.62 ± 37.820.1143rd Number of compression214.31 ± 25.89226.92 ± 16.350.153 Accuracy (%)70.31 ± 38.7779.54 ± 30.450.506 Mean velocity108.69 ± 13.11114.92 ± 8.490.165 Depth (mm)43.92 ± 2.6043.69 ± 1.840.796 Recoiling (%)53.92 ± 37.5929.85 ± 37.640.116Values are presented as mean ± SD. Normally distributed variables were compared using one-way ANOVAAbbreviation: CPR cardiopulmonary resuscitation†Mean velocity: number of chest compression per CPR cycle Comparison of CPR parameters among groups based heart rate changes Values are presented as mean ± SD. Normally distributed variables were compared using one-way ANOVA Abbreviation: CPR cardiopulmonary resuscitation †Mean velocity: number of chest compression per CPR cycle Discussion This study examined the association between exercise-related increases in heart rate and CPR performance during three cycles of infant chest compressions in a simulation setting. Across all groups, CPR quality remained consistent: compression rate, hand-placement accuracy, mean compression velocity, depth, and recoiling showed no significant within-group differences across cycles. Similarly, no significant between-group differences were observed regardless of when exercise was performed. Furthermore, stratifying participants by the magnitude of heart rate elevation revealed no significant differences in CPR performance. These findings suggest that heart rate elevation, used as an indirect surrogate of physical fatigue, did not meaningfully affect key CPR parameters during infant chest compressions. Both physical and cognitive fatigue are known to impair task accuracy and psychomotor performance [4, 18]. We hypothesized that exercise-induced increases in heart rate—serving as an indirect marker of physical exertion and sympathetic activation—might impair infant CPR quality. However, our findings contrast with many reports from adult simulation and clinical studies. Prior research, primarily involving adult CPR models and adult rescuers, has demonstrated that fatigue following physical activity can diminish CPR quality, prompting efforts to reduce fatigue during resuscitation [19–21]. Several factors may explain this discrepancy. Notably, the present study used an infant model, which requires substantially less compression depth than adult CPR. As a result, the physical demand imposed on rescuers is considerably lower, potentially mitigating fatigue-related decline. This interpretation aligns with findings from a previous comparative study examining heart rate responses during adult, pediatric, and infant chest compressions under varying compression-to-ventilation ratios [15]. In that study, heart rate fluctuations were markedly smaller in the infant model, a pattern consistent with the heart rate trajectories observed in our analysis. Another factor that may explain the differences between our findings and those of previous studies is the inclusion of a structured 1-min rest period between compression cycles. Earlier investigations often required participants to perform continuous ventilation and compressions for 5–10 min [16, 21, 22]. In contrast, brief recovery periods in our protocol may have mitigated fatigue sufficiently to preserve compression quality across cycles. Several previous studies evaluating rescuer fatigue have frequently relied on subjective measures such as the visual analog scale, Borg Rating of Perceived Exertion, or Numerical Rating Scale. Although widely used, these tools have inherent limitations when used as objective indicators of physiological fatigue [21, 23, 24]. Considering the reported associations between heart rate changes and CPR quality in earlier investigations, we adopted heart rate as an objective surrogate marker of sympathetic activation and fatigue in this study [15, 19, 20]. In addition, prior research examining fatigue and CPR quality has shown considerable variability regarding the duration of a single compression cycle, with reported intervals ranging from 2 to 10 min, and the length of rest periods provided between cycles [16, 19, 21–23, 25, 26]. To reflect real-world resuscitation environments, particularly those with limited personnel in which rescuers frequently rotate roles, we implemented a protocol of 2 min of compressions followed by a 1-min rest. This design aimed to balance methodological control with clinical relevance and enhance the applicability of our findings to practical pediatric resuscitation conditions. This study has several limitations. First, the sample size was relatively small, and the protocol included only three cycles of chest compressions; thus, results may differ in studies incorporating larger cohorts or longer, more demanding CPR durations. Nevertheless, the use of matched randomization likely helped reduce variability and strengthen the consistency of our observations. Second, because this was a manikin-based simulation, the study cannot fully replicate the complexities of actual clinical resuscitation. Real-world scenarios involve dynamic and unpredictable elements, including spatial constraints, heterogeneous patient anatomy, and the need to coordinate multiple simultaneous interventions, which can influence CPR quality. Patient-specific factors, such as chest wall compliance or progressive physiologic decline, may further alter rescuer mechanics and performance in ways that cannot be captured in controlled simulations. In addition, emotional stress, which can elevate heart rate during real cardiac arrest events, could not be adequately captured in this simulation-based study. Furthermore, heart rate elevation alone is insufficient to define fatigue; therefore, our findings reflect an indirect assessment of fatigue-related physiological burden rather than a direct measurement of clinical fatigue, which should be considered a limitation. Next, we did not adjust for individual differences in physical tolerance or baseline physical activity. However, given the relatively low physical burden associated with infant chest compressions, it is reasonable to assume that inter-individual variability in physical condition had minimal influence on the performance metrics observed. Finally, previous studies have identified strategies to reduce rescuer fatigue in infant CPR, reporting that the two-thumb encircling technique improves both compression efficacy and fatigue resistance compared with the two-finger technique [23, 27, 28]. Although we did not explicitly standardize hand positioning in our protocol, all participants naturally used the two-thumb technique, which may have contributed to the preservation of CPR quality. Therefore, the generalizability of our findings may be limited. In summary, the key limitations include that this study was conducted at a single center with a limited sample size and used a manikin-based simulation model; therefore, its conclusions cannot be directly extrapolated to real-world clinical practice or used to inform changes in clinical guidelines. Instead, the results should be viewed as hypothesis-generating for future clinical studies. Despite these limitations, this study is, to our knowledge, the first to examine the potential association between elevation of heart rate and CPR parameters using an infant manikin model. Our findings tentatively suggest that infant chest compressions may be less susceptible to fatigue-related deterioration than adult compressions. These findings may contribute to evidence-based determination of rescuer rotation intervals in training and to the design of longer simulation scenarios aimed at identifying the onset of fatigue. If confirmed in future studies, this could indicate that the interval for switching rescuers during infant CPR may not need to be as frequent as currently practiced. Furthermore, it raises the possibility that, in infant CPR scenarios, fewer providers may need to be dedicated solely to chest compressions, potentially allowing additional team members to focus on other critical tasks such as airway management, medication preparation, and establishment of intravenous or intraosseous access. However, these implications remain speculative and should be interpreted with considerable caution until validated through larger clinical studies. In conclusion, we found that change of heart rate did not have a significant impact on CPR quality during simulated infant chest compressions. Further investigations in real-world clinical environments are required to confirm these findings and to explore their implications for optimizing team allocation during pediatric resuscitation. Acknowledgements None. Authors’ contributions DH Kim: Conceptualization, formal analysis, writing, investigation, data curation, writing–original draft, visualization. SJ Choi: Conceptualization, methodology, writing– review & editing, supervision. All authors approved the final manuscript. Funding None. Data availability The data that support the findings of this study are available from the corresponding author upon reasonable requestion; however, restrictions apply to the availability of these data, which were used under license for the current study. Declarations Ethical approval and consent to participate IRB approval (#2025–0824) and written informed consent were obtained. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable requestion; however, restrictions apply to the availability of these data, which were used under license for the current study.