Comparing chest compression quality in 1-min vs. 2-min rescuer rotations with and without audiovisual feedback guidance during full adult advance life support simulated scenarios: a randomized simulation study Authors: Khrongwong Musikatavorn, Wittaya Namlimhemnatee, Jutamas Saoraya Published: 2026 ABSTRACT AbstractIntroductionGuidelines recommend 2-min rescuer rotations for high-quality cardiopulmonary resuscitation. However, 1-min cycles may improve performance. Their effectiveness in advanced cardiac life support, particularly with audiovisual feedback (AVF) devices, remains unclear. This study compares rotation strategies and AVF device use in a simulated advanced cardiac life support setting.MethodsIn this randomized simulation study, resuscitation teams performed a standardized scenario using either 1- or 2-min chest compression rotations, with or without AVF guidance. Manikin-derived data, including compression rate, depth, and percentage of target-compliant compressions, were compared among the four study groups. Chest compression fraction, interruption time, and rescuer fatigue were also assessed.ResultsA total of 31 participants performed chest compressions in this study. They were randomized to complete 10 simulated CPR sessions per group, resulting in data for 30 compressors per group. AVF use significantly improved compression depth and in-target performance in both 1- and 2-min rotation strategies (median 9.7% vs 52.8% in the 1-min no-AVF and with AVF, respectively; p < 0.001, and 6.7% vs 40.0% in the 2-min no-AVF and with AVF, respectively; p = 0.03). The 1-min groups had a lower chest compression fraction than the 2-min groups, although all groups maintained a chest compression fraction >80%. The 2-min with AVF group yielded the best overall performance.ConclusionWithout AVF, 1- and 2-min rotations showed comparable but suboptimal chest compression quality. AVF markedly improved performance across both strategies, with the 2-min rotation with AVF achieving the highest compression quality and chest compression fraction.Keywords: Chest compression quality, Rescuer rotation, Audiovisual feedback, Simulation Introduction The recent guideline of cardiopulmonary resuscitation (CPR) for patients in cardiac arrest recommend performing chest compressions at a depth of 5–6 cm and a rate of 100–120 compressions per minute with minimal interruption.1, 2 Adhering to these parameters improves survival rates and enhances defibrillation success.3, 4, 5 To maintain high-quality compressions, the guideline advises rotating rescuers every 2 min, which aligns with the need to assess pulse and electrocardiogram (ECG) rhythms during resuscitation. The guideline’s class 2a recommendation for 2-min rescuer rotations is supported by evidence showing that chest compression depth begins to decline after 90–120 s of continuous effort, even though the rate often remains unchanged.6, 7 However, previous studies have highlighted low compliance with the recommended 2 min of uninterrupted chest compressions, often resulting in a decline in performance due to rescuer fatigue during prolonged resuscitation, as well as factors such as the physical build and the challenges associated with wearing masks.8, 9, 10 Additionally, previous researches conducted in continuous chest compression CPR or Basic Life Support (BLS) scenarios suggested that a 1-min chest compression cycle may provide superior CPR quality compared to a 2-min cycle.11, 12, 13, 14 Limited research has evaluated CPR performance using a 1-min chest compression cycle in an extended advanced cardiac life support (ACLS) scenario. In such settings, the integration of other resuscitative interventions (e.g., intubation, defibrillation, medication administration) and assessments (e.g., pulse and rhythm checks), combined with more frequent compressor switches if 1-min rotation was applied, can divert attention from chest compressions and may increase the likelihood of compression pauses, compromising overall compression quality. Audiovisual feedback (AVF) devices are now widely implemented in resuscitation training and practice. They are designed to provide real-time guidance on compression quality, including depth, rate, and chest recoil. These devices, either standalone or integrated into modern defibrillators, display performance metrics and issue auditory or visual cues to help rescuers meet optimal compression standards. AVF devices and their associated technology also calculate the chest compression fraction (CCF), which represents the proportion of CPR time spent on chest compressions. Higher CCF values, accomplished by minimizing pauses in compressions, are associated with improved return of spontaneous circulation (ROSC) rate and survival outcomes.15, 16, 17 The feedback from devices may improve CPR performance by counteracting the decline in quality caused by rescuer fatigue and mitigating unintentional pauses when such devices are not utilized. This study aims to compare the quality metrics of chest compressions along with the CCF and interrupted time in advanced resuscitation of adult cardiac arrest scenarios using a simulated manikin. Specifically, it evaluated the impact of 1-min versus 2-min rescuer rotations on compression performance, with and without AVF guidance in the full and extended ACLS scenarios. The results of this study may provide additional insight into the optimal rotation strategy of chest compression cycles in general ACLS settings. Methods Study design and setting This randomized study was conducted at the Chulalongkorn Healthcare Advanced Multi-Profession Simulation Center (CHAMPS Center), an accredited facility by the Society for Simulation in Healthcare (SSH), located within an urban, 1500-bed, university-affiliated tertiary care hospital. This center regularly provides training in standard BLS and ACLS courses in accordance with updated guidelines. The study was conducted from February to April 2026 and received approval from the Institutional Review Board, Faculty of Medicine, Chulalongkorn University (Identifier: 0378/68, approval number 0995/2025) and was registered with Thai clinical Trials Registry (Identifier TCTR20260126007) on 26 January 2026. Participants We recruited board-certified emergency physicians, certified paramedics and emergency department (ED) registered nurses with a minimum of two years of ACLS and BLS experience to form resuscitation teams. Participation in the study was voluntary, and written informed consents were obtained. The study participants comprised 19 ED nurses and 12 certified paramedics, all of whom were assigned to be the role of chest compressor. To facilitate the simulation team, physicians and nurses from a pool of five board-certified emergency physicians and ten ED nurses were randomly selected to serve as code team leaders and other essential team members; these individuals were not study participants. Code team assembly The code team members were selected using computer-generated randomization and the team composition was established prior to the initiation of the study scenarios. Three chest compressors, designated as the study participants, were randomly selected from the test participant pool, which comprised 31 participants (12 paramedics and 19 registered nurses). The remaining team members, designated as non-test participants, included one team leader and four additional task members. The team leader was randomly selected from a pool of five emergency department staff physicians, while the four additional task members were randomly selected from a pool of 10 registered nurses. As a result, each team comprised 7 members: one emergency physician and six nurses and/or paramedics, including three assigned as chest compressors (study participants) and three serving in additional team roles. Computer-generated randomization of both study participants and non-test team members was performed by independent personnel. Each team member was assigned a specific role: team leader, chest compressor, airway manager, medication administrator or recorder/timer (details provided in the Supplementary Material). Study group allocation The seven-member code teams were then randomly assigned to one of the four study groups Group A: 1-min compression rotation without AVF guidance, Group B: 1-min compression rotation with AVF guidance, Group C: standard 2-min compression rotation without AVF guidance, and Group D: Standard 2-min compression rotation with AVF guidance. Group allocation was performed by independent personnel using computer-generated, block-of-4 randomization and concealed in sealed envelopes prior to the initiation of the study scenarios. The teams remained blinded to their assigned groups until the start of each scenario in the simulation theater. Each group was randomly assigned to perform ten rounds of the scenario, identified as Group A1, A2, A3, to A10, and similarly for Group B, C, and D. The procedure was repeated until a total of 40 sessions were completed, with 10 sessions for each group. The study operational diagram is summarized and presented in Fig. 1. The study operational diagram. Abbreviation: AVF: Audiovisual feedback, CPR: cardiopulmonary resuscitation. Team preparation and briefing Prior to beginning the assigned scenarios, all groups received a standard briefing emphasizing adherence to the standard metrics of high-quality chest compressions. In Group A and C, compression performance metrics were recorded, but the audiovisual feedback from the devices was deactivated and not displayed to the team. In Group B and D (performed with AVF guidance), chest compressors were instructed to closely follow the feedback provided by the device and to make every effort to align their performance with the guidance throughout the scenarios. Every team was emphasized to avoid spending more than 10 s or unnecessary pauses in chest compression during intubation, pulse or rhythm checks in all scenarios. During the test scenario, the three chest compressors in each study team were instructed to perform chest compressions independently and were not allowed to interact or communicate with one another to minimize potential clustering effects. The premature compressor switch before the assigned interval was not permitted or verbally indicated. Simulation scenario design and setup All teams were assigned to manage a scenario involving acute myocardial infarction leading to cardiac arrest with an initial shockable rhythm (ventricular fibrillation, VF), in accordance with the 2025 ACLS guidelines at the time of the study. Briefly, the scenario consisted of nine rounds of VF, requiring a total of eight chest compression cycles before achieving a return of spontaneous circulation (ROSC). As a result, the total CPR duration, defined as the interval from code initiation to termination, was standardized at 18–20 min across all teams and scenarios. Each chest compression session (1-min or 2-min, depending on the study group) was carried out with three rotating compressors for each scenario. The currently recommended CPR techniques aimed at minimizing chest compression pauses, including hand hovering during rescuer transitions, pre-charging the defibrillator, and pulse palpation prior to rhythm assessment at the end of 2-min CPR cycles, were consistently reinforced throughout all testing sessions. These measures were implemented to reduce peri-shock compression pauses, which have been associated with unsuccessful defibrillation and lower survival rates in patients with shockable cardiac arrest rhythms3, 18 and to ensure that chest compression interruptions occurring during the study scenarios primarily resulted from compressor switching rather than peri-shock pauses. A single advanced CPR manikin was used for all teams to ensure consistency and eliminate variability in compression quality measurement. The monitoring device and defibrillator used in all scenarios were the Zoll® R Series, equipped with Real CPR Help® audiovisual feedback technology. This approach minimized differences in material elasticity and sensor sensitivity that could affect the evaluation of chest compression performance. Detailed descriptions of the scenario setup and equipment are provided in the Supplementary materials. To minimize physical and mental fatigue and potential carryover effects, participants completed no more than one scenario per day, with a 2-day washout between sessions. After each study session, the temporary three-person rescuer team and remaining team members were all disbanded immediately, and all participants returned to their respective recruitment pools for subsequent randomization. Data collection and outcome measurement The individual performance of chest compressors was assessed by the measurement of quality metrics included compression rate, depth in centimeter, and percentage of in-target compression (percentage of total chest compressions delivered during CPR that fall within the recommended range for key parameters compression rate (100–120 compressions per minute, compression depth 5–6 cm). Overall team performance of the study groups was evaluated using CCF, defined as the proportion of time during a cardiac arrest resuscitation event that chest compressions are performed, and total duration of interruptions in chest compression All data were recorded and subsequently retrieved through the RescueNet® CaseReview program (Zoll®, MA, USA). An example of the dashboard displaying chest compression quality metrics is shown in Fig. S1. The primary objective of this study is to compare individual performance metrics (specifically measuring compression depth and rate) among the chest compressors in each study group, with 30 chest compressors assigned per group. The secondary objectives are to analyze team performance metrics, including overall interruption time and CCF. As an additional outcome, physical fatigue was assessed immediately after each CPR round using the 10-cm Visual Analog Scale (VAS), where 0 indicated “no fatigue at all” and 10 indicated “extreme fatigue.” The assessments were performed by independent personnel who were not involved in the study. The fatigue severity was compared among the three rotating compressors within each group. Accordingly, both compression performance and physical fatigue were evaluated for all 30 chest compressors in each study group. Sample size estimation Sample size estimation was based on prior data from the previous study which reported mean chest compression depths between 1-min and 2-min rescuer rotations.14 If only two groups had been assessed, the mean and standard deviation of adequate compression depth would have been used to calculate the required sample size using a two-sample t-test. Accounting for a potential 10% rate of incomplete data, the minimum sample size per group based on two group comparison was determined to be 30 participants. However, the present study was designed as a randomized four-group trial. For multiple-group comparisons, the sample size was calculated using one-way ANOVA, assuming four groups and an effect size of 0.3 (medium). With a two-sided α of 0.05 and 80% power, the estimated sample size was at least 28 participants per group. Accordingly, a target sample size of 30 participants per group was set to ensure sufficient power for both two-group and multiple-group analyses. Statistical analysis Data from all study participants (chest compressors) and code teams were included in the analysis. Statistical analyses were performed using SPSS software, version 17.0 for Windows (Chicago, IL, USA). Shapiro-Wilk test was used for normality testing. Given that the primary analysis involved multiple group comparisons, normally distributed and non-normally distributed continuous data (e.g., compression rate, compression depth, CCF) were analyzed using ANOVA and the Kruskal-Wallis test, respectively. For pairwise comparisons between two groups, normally distributed and non-normally distributed continuous data were analyzed using the independent t-test and the Mann-Whitney U test, respectively. Since six pairwise comparisons were conducted (Group A vs B, A vs C, A vs D, B vs C, B vs D, and C vs D), the family-wise error rate was estimated to be 26.5%. To account for this and reduce the error rate, the alpha levels for statistical significance were adjusted using the Bonferroni method in the pairwise analyses among the study groups. Results A total of 31 participants (nurses and paramedics) performed chest compressions in this study. They were randomized to complete 10 simulated CPR sessions per group, resulting in data for 30 participants per group. In each study group, the chest compression performance metrics and physical fatigue were assessed from 30 chest compressors from 10 CPR sessions. Total CPR duration, from code initiation to termination, ranged from 18 to 20 min across all teams and scenarios. The characteristics of the chest compressors were shown in Table S1. Focusing on the individual performance metrics, a significant difference in mean compression depths, proportion of in-target depths, and compressions was found among the study groups. However, there was no statistical difference in the median compression rate and in-target rate among all groups. The comparisons among the four study groups were shown in Table 1. In pairwise comparison among the groups, the performance metrics in depth and rate were not statistically different between the 1-min or 2-min rotation group when AVF device was not used. However, the performance in compression depth was substantially improved in both 1-min AVF and 2-min AVF compared with the no-AVF groups (20.8% [IQR 1.9, 55.5] vs 70.5% [IQR 34.6, 81.9] in 1-min no-AVF and with AVF, respectively; p-value < 0.001, and 8.8% [IQR 0.3, 43.7] vs 53.8% [IQR 13.0, 84.9] in 2-min no-AVF and with AVF, respectively; p-value = 0.003) Moreover, chest compression with AVF guidance make the mean depth fall within the standard recommendation (5–6 cm) than doing it without, as reflected with the large improvement in the proportion of in-target compressions (9.7% [IQR 1.4, 33.1] vs 52.8% [IQR 29.0, 72.3] in 1-min no-AVF and with AVF, respectively; p-value < 0.001, and 6.7% [IQR 0.0, 33.6] vs 40.0% [IQR 13.0, 73.4] in 2-min no-AVF and with AVF, respectively; p-value < 0.008). The individual performance metrics among the groups and the pairwise comparison are shown in Table 2. Table 1Primary outcomes: comparison of individual and team performance metrics among the study groups.Individual performance metricsGroup A1-min(n = 30)Group B1-min/AVF(n = 30)Group C2-min(n = 30)Group D2-min/AVF(n = 30)P valueDepth (cm)a4.7 ± 0.75.2 ± 0.34.6 ± 1.05.0 ± 0.60.004*In-target depth (%)b20.8(1.9, 55.5)70.5(34.6, 81.9)8.8(0.3, 43.7)53.8(13.0, 84.9)<0.001*Compression rate (compressions/min)b114.5(109.9, 120.2)114.0(111.1, 117.9)111.4(108.3, 116.7)114.7(110.0, 116.6)0.578In-target rate (%)b74.3(45.7, 86.3)79.3(60.4, 88.4)75.6(57.0, 88.6)84.8(69.4, 91.6)0.487In-target compression (%)b9.7(1.4, 33.1)52.8(29.0, 72.3)6.7(0.0, 33.6)40.0(13.0, 73.4)<0.001*Physical fatigue10-cm VAS (cm)b4.7(3.0, 5.6)5.3(4.7, 6.1)5.6(4.0, 6.4)6.3(5.7, 7.5)0.001* Team performance metrics†(n = 10)(n = 10)(n = 10)(n = 10)Total interrupt time (s)b153.3 (143.1, 177.2)151.0 (145.3, 160.4)132.3(121.9, 141.7)119.6(114.7, 121.3)<0.001*CCF (%)b86.9(84.5, 87.7)86.9(86.1, 87.5)88.8 (87.9, 89.6)89.8(88.7, 90.3)<0.001*Notes: AVF = Audiovisual feedback, 10-cm VAS = 10-centimeter Visual Analog Scale.Total interrupt time = Total time that chest compression was interrupted from code started to code stopped.CCF = Chest compression fraction from code started to code stopped.aMean ± SD; One-way ANOVA.bMedian (interquartile range, IQR); Kruskal-Wallis test.†Total interruption time and chest compression fraction (CCF) were used to assess the continuality of chest compression and team smoothness in compressor switch across the 10 CPR sessions in each study group (n = 10).*P-value < 0.05. Primary outcomes: comparison of individual and team performance metrics among the study groups. Notes: AVF = Audiovisual feedback, 10-cm VAS = 10-centimeter Visual Analog Scale. Total interrupt time = Total time that chest compression was interrupted from code started to code stopped. CCF = Chest compression fraction from code started to code stopped. Mean ± SD; One-way ANOVA. Median (interquartile range, IQR); Kruskal-Wallis test. Total interruption time and chest compression fraction (CCF) were used to assess the continuality of chest compression and team smoothness in compressor switch across the 10 CPR sessions in each study group (n = 10). P-value < 0.05. Table 2Pairwise comparisons among the study groups.OutcomesNMean ± SD/Median (IQR)P-value1-min1-min/AVF2-min2-min/AVFIndividual performance metricsDepth (cm)a1-min304.7 ± 0.7––––1-min/AVF305.2 ± 0.30.089–––2-min304.6 ± 1.01.0000.005*––2-min/AVF305.0 ± 0.60.8131.0000.088–In-target depth (%)b1-min3020.8(1.9, 55.5)––––1-min/AVF3070.5(34.6, 81.9)<0.001**–––2-min308.8(0.3, 43.7)0.482<0.001**––2-min/AVF3053.8(13.0, 84.9)0.0290.1390.003**–Compression rate (times/min)b1-min30114.5(109.9, 120.2)––––1-min/AVF30114.0(111.1, 117.9)0.701–––2-min30111.4(108.3, 116.7)0.2040.287––2-min/AVF30114.7(110.0, 116.6)0.7560.9060.337–In-target rate (%)b1-min3074.3(45.7, 86.3)––––1-min/AVF3079.3(60.4, 88.4)0.344–––2-min3075.6(57.0, 88.6)0.8590.460––2-min/AVF3084.8(69.4, 91.6)0.1790.5840.261–In-target compression (%)b1-min309.7(1.4, 33.1)––––1-min/AVF3052.8(29.0, 72.3)<0.001**–––2-min306.7(0.0, 33.6)0.625<0.001**––2-min/AVF3040.0(13.0, 73.4)0.0140.2840.007**–Physical fatigue10-cm VAS (cm)b1-min304.7(3.0, 5.6)––––1-min/AVF305.3(4.7, 6.1)0.064–––2-min305.6(4.0, 6.4)0.0590.646––2-min/AVF306.3(5.7, 7.5)<0.001**0.006**0.024– Team performance metrics†Total interrupt time (s)b1-min10153.3(143.1, 177.2)––––1-min/AVF10151.0(145.3, 160.4)0.539–––2-min10132.3(121.9, 141.7)0.002**0.003**––2-min/AVF10119.6(114.7, 121.3)<0.001**<0.001**0.089–CCF (%)b1-min1086.9(84.5, 87.7)––––1-min/AVF1086.9(86.1, 87.5)0.351–––2-min1088.8(87.9, 89.6)<0.001**<0.001**––2-min/AVF1089.8(88.7, 90.3)<0.001**<0.001**0.170–Notes: AVF = Audiovisual feedback.In-target depth = Proportion of total chest compressions achieving a depth of 5–6 cm, in accordance with current guideline recommendations.In-target rate = Proportion of total chest compressions delivered at a rate of 100–120 compressions per minute, in accordance with guideline recommendations.In-target compression = Proportion of total chest compressions simultaneously achieving both the recommended depth (5–6 cm) and rate (100–120 compressions per minute).Total interrupt time = Cumulative duration of pauses in chest compressions from the initiation to the termination of the resuscitation event.CCF (Chest compression fraction) = Proportion of total resuscitation time (from event initiation to termination) during which chest compressions were actively delivered.aMean ± SD (standard deviation); t-test (parametric).bMedian (interquartile range, IQR); Mann-Whitney U test (non-parametric).*In parametric tests, the presented P-values were corrected with Bonferroni test and P < 0.05 is considered statistical significant.**In non-parametric tests, P-value < 0.008 (after Bonferroni correction) is considered statistical significant.†Total interruption time and CCF were used to assess team smoothness of chest compression across the 10 CPR sessions in each study group (n = 10). Pairwise comparisons among the study groups. Notes: AVF = Audiovisual feedback. In-target depth = Proportion of total chest compressions achieving a depth of 5–6 cm, in accordance with current guideline recommendations. In-target rate = Proportion of total chest compressions delivered at a rate of 100–120 compressions per minute, in accordance with guideline recommendations. In-target compression = Proportion of total chest compressions simultaneously achieving both the recommended depth (5–6 cm) and rate (100–120 compressions per minute). Total interrupt time = Cumulative duration of pauses in chest compressions from the initiation to the termination of the resuscitation event. CCF (Chest compression fraction) = Proportion of total resuscitation time (from event initiation to termination) during which chest compressions were actively delivered. Mean ± SD (standard deviation); t-test (parametric). Median (interquartile range, IQR); Mann-Whitney U test (non-parametric). In parametric tests, the presented P-values were corrected with Bonferroni test and P < 0.05 is considered statistical significant. In non-parametric tests, P-value < 0.008 (after Bonferroni correction) is considered statistical significant. Total interruption time and CCF were used to assess team smoothness of chest compression across the 10 CPR sessions in each study group (n = 10). When total interruption time and chest compression fraction (CCF) were used as measures of team performance, the 1-min groups demonstrated significantly longer median interruption times and lower CCF compared with their 2-min counterparts. Despite this difference, both 1-min groups—regardless of AVF device use—maintained CCF values exceeding 80%, indicating preservation of acceptable compression continuity (CCF 86.9% [IQR 84.5, 87.7] and 86.9% [IQR 86.1, 87.5] and in 1-min no-AVF and with AVF groups, respectively vs 88.8% [IQR 87.9, 89.6] and 89.8% [IQR 88.7, 90.3] in 2-min no-AVF and with AVF, respectively; all p-values < 0.001 among 1-min and 2-min groups). Importantly, the 2-min AVF-guided group achieved the shortest interruption time and the highest CCF, reflecting superior overall team performance. Detailed comparisons of interruption time and CCF across study groups are presented in Table 2. With respect to individual physical fatigue, compressors in the 2-min AVF group reported significantly higher median fatigue scores than those in the 1-min groups, irrespective of AVF device use. In contrast, compressors in the 2-min no-AVF group demonstrated fatigue scores comparable to those observed in both 1-min groups. Comparisons of individual rescuer physical fatigue in the study groups are presented in Table 2. Discussion In this randomized simulation study comparing CPR performance with compressor rotation intervals of either 1 or 2 min, with or without AVF, we demonstrated that the 2-min rotation interval combined with AVF yielded the best chest compression fraction, despite higher perceived fatigue levels among the compressors. AVF significantly improved both the in-target depth and rate of chest compression in both the 1-min and 2-min rotation groups. The recent CPR guidelines recommend maintaining the compression depth and rate as well as emphasizing on reducing the interruption time and therefore increasing the CCF during all BLS or ACLS attempts. Although 2-min compression cycle and rescuer switch remains in the standard for a long time, this recommendation is considered weak as the optimal duration is still unknown.19 1-min chest compression rotation may offer a possible way to maintain recommended depth and rate by reducing the rescuer fatigue. A single manikin simulation study has demonstrated that there was no significant difference in the effectiveness of chest compressions when switching two rescuers every 1 min or 2 min within 8 min of CPR.7 Most of the 1-min CPR rotation data was derived from prior studies conducted with the primary focus on continuous chest compressions in simulated BLS settings.11, 12, 13, 14, 15, 20 In ACLS resuscitations, prolonged code duration, concurrent life-saving interventions (e.g., intubation, defibrillation, pulse checks, and invasive procedures), and increased cognitive demands may compromise chest compression quality and continuity, potentially resulting in suboptimal overall CPR performance. To our knowledge, this is the first study to simultaneously evaluate the effects of the strategies to enhance the CPR quality (1-min rotation to mitigate compressor fatigue and the use of AVF devices) in the simulated setting of extended ACLS scenarios, using a comprehensive set of chest compression quality metrics. Comparing within the 1-min and 2 min without AVF groups, both groups had similar median compression depth and rate but at a very low proportion of compressions within the target range. This finding suggests that even when maximal physical capacity is preserved during shorter compression efforts (e.g., within the first minute), the rescuers did not constantly perform the compressions with recommended depth. This is further reflected by a median compression depth below guideline recommendations and a very low proportion of compressions within the target range in the 1-min no-AVF group. Notably, median self-reported fatigue scores were comparable between the 1-min and 2-min no-AVF groups. These findings imply that the preservation of physical strength—or the absence of perceived fatigue of the rescuers—does not necessarily ensure the delivery of high-quality CPR. One possible explanation is that rescuers may subconsciously regulate the force they apply during chest compressions according to an internal perception of what feels tolerable or sustainable, rather than exerting their maximal effort. This self-regulatory behavior may occur independently of the duration of the compression cycle and could lead to consistently suboptimal compression depth despite preserved physical capacity. Comparing within the 1-min and 2 min AVF groups, no significant differences were observed in compression depth or rate between rotation intervals. Notably, when AVF was implemented, device guidance markedly improved compression depth in both rotation interval groups but had little effect on compression rate. Overall, AVF guidance enhanced chest compression quality toward guideline-recommended targets, as reflected by the substantially higher proportion of in-target compressions. This finding is supported by a number of studies.21, 22, 23, 24, 25 In our study, improvements of the performance appeared to be primarily attributable to optimization of compression depth rather than rate, and this effect was consistent across rotation intervals. As expected, the median fatigue scores were highest in the 2-min AVF group compared with the other study groups. The results suggest that the guidance from AVF device prompts the rescuers to exert greater physical force to counteract the progressive decline in compression performance that typically occurs during prolonged manual CPR rotations. Although the 1-min rotation groups demonstrated a slightly lower CCF than the 2-min rotation groups with statistically significant differences, the absolute values remained above 80% and are unlikely to be clinically meaningful. However, the findings suggest that more frequent compressor rotation may increase the likelihood of brief interruptions in chest compressions, even under standardized simulation conditions. This effect may be more pronounced during resuscitation of real patients, where the clinical environment is less controlled and more complex. Accordingly, when physical fatigue or other clinical circumstances necessitate compressor rotation at intervals shorter than 2 min, such a strategy may be acceptable, with caution regarding the potential for increased interruptions. Although prior studies suggest that small differences in compression pauses or CCF may not substantially affect clinical outcomes in cardiac arrest.26, 27, 28 any premature rescuer switch should be performed as seamlessly as possible to minimize cumulative interruptions and avoid meaningful reductions in CCF. Similar considerations may apply to other CPR performance parameters that can deteriorate during prolonged rotations, environmental challenges or extended resuscitation efforts in real patients, such as mean chest compression depth during prolonged code events. This study has noteworthy limitations. First, as a single-center manikin study, the findings may not be generalizable and may not reflect clinical practice in real patients. Additionally, all participants were of Asian ethnicity, which may further limit the applicability of the findings to other populations due to potential differences in body structure and physical strength. Second, participants were repeatedly sampled from a common pool; however, participant identifiers were not retained across study rounds. Consequently, repeated observations from the same individual could not be identified or modeled. Statistical analyses therefore treated observations as independent, which may have affected the results if participants contributed multiple observations. Third, although the CPR scenario closely simulated prolonged resuscitation, it did not fully replicate real-world complexities, including environmental challenges, concurrent interventions, and scene safety concerns that may affect CPR quality. Additionally, the standardized protocol did not capture the diagnostic uncertainty and dynamic decision-making required during actual cardiac arrest. Furthermore, our standardized three-rescuer rotation protocol was designed to minimize cumulative fatigue but may not fully reflect the widely-used practice of two-rescuer rotation with a third rescuer replacing a fatigued compressor. Therefore, the generalizability of our findings to this approach may be limited. Fourth, although each scenario was conducted once per day to minimize rescuer fatigue and allow a 2-day washout period, prior exposure to the study protocol may still have influenced rescuer performance. Participants may have assimilated experience from previous sessions, potentially improving or altering their behavior during subsequent tests. Finally, the Hawthorne effect and the unblinded study design may also have influenced personal and team performance. Conclusion In the absence of AVF guidance, 1-min rescuer rotation achieved compression depth and rate comparable to 2-min rotation, with a slight decrease in CCF; however, overall compression quality remained suboptimal in both groups. When AVF guidance was applied, performance metrics improved markedly and were more likely to meet guideline-recommended targets regardless of the rotation duration. In our study, 2-min rotation under AVF guidance achieved the highest compression quality metrics and CCF, accompanied by greater rescuer fatigue. Funding disclosure The research was supported by Faculty of Medicine, Chulalongkorn University. Declaration of generative AI During the preparation of this work, the authors used the generative AI (ChatGPT 3.5) solely for assistance in checking and refining the English language. The authors entirely generated the content and findings presented in this work without AI assistance. After using this tool, the author reviewed and edited the content as needed and takes full responsibility for the content of the published article. Data availability statement Access to the data can be granted by the investigators upon reasonable request. CRediT authorship contribution statement Khrongwong Musikatavorn: Writing – review & editing, Writing – original draft, Supervision, Methodology, Formal analysis, Conceptualization. Wittaya Namlimhemnatee: Writing – review & editing, Formal analysis, Data curation, Conceptualization. Jutamas Saoraya: Writing – review & editing, Visualization, Methodology. Declaration of competing interest The authors declare no competing interests. Footnotes Supplementary material to this article can be found online at https://doi.org/10.1016/j.resplu.2026.101423. Contributor Information Khrongwong Musikatavorn, Email: khrongwong.m@chulahospital.org. Wittaya Namlimhemnatee, Email: aong1_1@docchula.com. Jutamas Saoraya, Email: jutamas.sa@chula.ac.th. Supplementary material The following are the Supplementary material to this article: Supplementary Data 1 References Associated Data Supplementary Materials Supplementary Data 1 Data Availability Statement Access to the data can be granted by the investigators upon reasonable request.