Single Rescuer Exertion Using a Mechanical Resuscitation Device: A Randomized Controlled Simulation Study


  • Presented at the European Resuscitation Council Congress, porto, Portugal, December 2010.

  • The authors give special thanks to Alber Antriebstechnik GmbH (Albstadt, Germany) who provided the mechanical resuscitation device (MRD) “Animax,” to Roche-Austria (Vienna, Austria) for providing a lactate meter, to Gepa-Med (Vienna, Austria) for providing a digital Holter recorder, and to Chemomedica (Vienna, Austria) for providing the blood pressure measurement device. None of these companies reviewed or revised the manuscript at any stage. None of the authors have any potential conflicts of interests to report.

Address for correspondence and reprints: Stephanie Neuhold, MD, MSc; e-mail:



The goal of this experimental study was to investigate rescuer exertion when using “Animax,” a manually operated hand-powered mechanical resuscitation device (MRD) for cardiopulmonary resuscitation (CPR), compared to standard basic life support (BLS).


This was a prospective, open, randomized, crossover simulation study. After being trained, 80 medical students with substantial knowledge in BLS performed one-rescuer CPR using either the MRD or the standard BLS for 12-minute intervals in random order. The main outcome parameter was the heart rate pressure product (RPP) as an index of cardiac work. Secondary outcome parameters were physical exhaustion quantified by the Borg scale (measurement of perceived exertion), Nine Hole Peg Test (NHPT; measurement of fine motor skills), and capillary lactate concentration during testing.


While no significant difference could be found for the RPP, a significantly increased mean heart rate during the final minute of standard BLS compared to the MRD was found (139 ± 22 beats/min vs. 135 ± 26 beats/min, p = 0.027). By contrast, subjective exertion using the MRD was rated significantly higher on the Borg scale (15.1 ± 2.4 vs. 14.6 ± 2.6, p = 0.027). Mean serum lactate concentration was significantly higher when the MRD was used compared to standard BLS (3.4 ± 1.5 mmol/L vs. 2.1 ± 1.3 mmol/L, p ≤ 0.001).


Use of the MRD leads to a RPP of the rescuers comparable to standard BLS. These findings suggest that there is no clinically relevant reduction of exertion if this MRD is used by a single rescuer. If this kind of MRD is used for CPR, frequent changeovers with a second rescuer should be considered as the guidelines suggest for standard CPR.


El Esfuerzo de un Unico Reanimador Mediante el Uso de un Dispositivo de Resucitación Mecánica: Un Estudio de Simulación Controlado con Asignación Aleatorizada


El objetivo de este estudio experimental fue investigar el esfuerzo de un reanimador cuando se utiliza “Animax”, un dispositivo manual de resucitación mecánica (DRM) para la resucitación cardiopulmonar (RCP), en comparación con el soporte vital básico (SVB) estándar.


Estudio prospectivo de simulación cruzado, abierto y con asignación aleatoria. Tras un entrenamiento inicial, 80 estudiantes de medicina con formación en la realización de SVB llevaron a cabo una RCP con un único resucitador, bien mediante el uso del DRM o mediante SVB estándar durante intervalos de 12 minutos de forma aleatorizada. El resultado principal fue el producto de la presión y la frecuencia cardiaca (PTF) como un indicador de trabajo cardiaco. Los resultados secundarios fueron el cansancio físico cuantificado con la escala de Borg (medida del esfuerzo percibido), el test de Nine Hole Peg (medida de las habilidades motoras finas) y la concentración de lactato capilar durante la prueba.


A pesar de no encontrarse una diferencia significativa para la PTF, se encontró un incremento significativo en la media de la frecuencia cardiaca durante el minuto final del SVB comparado con el uso del DRM (139 ± 22 vs. 135 ± 26 lpm, p = 0,027). Por otro lado, el esfuerzo subjetivo fue significativamente mayor en la escala de Borg al usar el DRM (15,1 ± 2,4 vs. 14,6 ± 2,6, p = 0,027). La concentración media de lactato sérico fue significativamente mayor cuando se usó el DRM en comparación con el SVB (3,4 ± 1,5 mmol/l vs. 2,1 ± 1,3 mmol/l, p ≤ 0,001).


El uso del DRM produce una PTF de los reanimadores comparable con el SVB estándar. Estos hallazgos sugieren que no hay una reducción clínicamente relevante del esfuerzo si el DRM se usa por un único reanimador. Si se usa este tipo de DRM para la RCP, debería considerarse realizar cambios frecuentes con un segundo reanimador tal como sugieren las guías clínicas para la RCP clásica.

Even under ideal circumstances, when cardiac arrest is witnessed and the chain of survival is initiated quickly,[1] rates of long-term survival and favorable outcome are unsatisfyingly low.[2, 3] Human and animal studies show that the quality of cardiopulmonary resuscitation (CPR) substantially affects outcome.[4-6] However, CPR is extremely exhausting,[7] and therefore the rescuer's capacity to perform efficient CPR over an extended period of time or during transport is limited.[8-10] It has been shown that rescuer fatigue leads to a decrease in the quality of basic life support (BLS) after just 90 seconds of CPR.[9] In an effort to support high-quality CPR that can be sustained over a long period, mechanical resuscitation devices (MRDs) for closed chest compression have entered the clinical arena of CPR.[11, 12]

Over the past decades various MRDs have been tested in preclinical and clinical settings with the aim of offering high-quality CPR and to reduce rescuer fatigue, but so far no device has become a clinical standard.[13] The MRD “Animax” (AAT Alber Antriebstechnik GmbH, Albstadt, Germany) represents a new generation of such devices. With this device both chest compressions and assisted bag–mask ventilation can be delivered. It is designed for out-of-hospital CPR on adults for health care professionals and laypersons alike to provide high-quality CPR that can be sustained over a long period.

Our group has demonstrated the MRD's superiority over standard BLS concerning compression parameters (except compression rate), as well as a higher minute-volume and less gastric inflations in the MRD group. However, ventilation volumes were below the 2005 European Resuscitation Council (ERC) guidelines for both methods. Hands-off time was significantly reduced in the MRD group, compared to standard BLS.[14]

Little is known about rescuer fatigue when using this new MRD, which is purely hand-powered without an external energy source. Therefore, the use of the MRD could lead to considerable rescuer exertion, similar or possibly even greater than the effects observed during standard manual CPR.[7, 9, 15, 16] The aim of this study was to compare rescuer exertion using the MRD to standard BLS.


Study Design

This study was part of a larger project, an open randomized crossover trial investigating the performance of the MRD “Animax” compared to standard BLS.[14] The study was approved by the ethical committee of the Medical University of Vienna, and all participants provided written informed consent for their participation.

Study Setting and Population

Eighty-one third-year medical students from the Medical University of Vienna, Austria, volunteered for this study (Figure 1). All participants were recruited over a period of 2 weeks and underwent extensive training in BLS according to the ERC guidelines 2005[17] during their mandatory classes.

Figure 1.

Study flow.

Study Protocol

All participants were taught standard BLS and the use of the MRD using the four-stage approach for skill acquisition in CPR.[18] This teaching took about 15 to 20 minutes for each of the two methods.[14] Afterward, the study team evaluated the participants' competence with both methods. To pass competence testing, each participant individually had to demonstrate two correct standard BLS cycles (30:2) within 60 seconds. The MRD had to be assembled and two correct cycles had to be demonstrated within 120 seconds.

Subsequently, each participant performed CPR using both the MRD and standard BLS in accordance with the ERC guidelines for one-rescuer technique.[17] The starting method was randomized, and the randomization list was provided by a Web-based computer program ( and distributed by an independent researcher in sealed envelopes.

Each CPR cycle lasted 12 minutes, because this interval reflects the mean response time for the Vienna Municipal Medical Emergency Service. A 30-minute break served to avoid the influence of rescuer fatigue on the subsequently performed method. Additionally, each participant's baseline pulse rate before the second CPR cycle had to be no more than 10% above his or her initial baseline pulse rate.[14]

Single Rescuer CPR With the MRD

The MRD “Animax” has been described previously.[14] In brief, it is a purely mechanical, hand-powered device, independent of any external energy source.[19, 20] After assembly (according to the manufacturer this takes 15 to 20 seconds) one person pushes the lever up and down to achieve CPR. The device automatically alternates between chest compressions and ventilation in a rhythm of 30:2. The rescuer remains in the same place at the side of the manikin's head during the 12-minute testing period.

Conventional Standard BLS for a Single Rescuer

To provide chest compressions, the rescuer was positioned at the side of the manikin's chest. To ventilate, the rescuer moved to the head of the manikin using a size 4 face mask with a self-inflating bag (Ambu Mark IV Adult, Ambu silicon face-mask size 4, Ambu, Ballerup, Denmark).

Physical Strain

To measure each participant's heart rate, a four-lead electrocardiogram (ECG) Holter digital recorder (Aria, Spacelabs Healthcare, OSI Systems, Hawthorne, CA) was attached via standard ECG electrodes to the chest. Blood pressure was measured by a noninvasive oscillometeric device (Model TM910, Physiogard, Schiller, Germany) attached to the participant's upper arm. Each participant was asked to rate his or her subjective exertion on the 15-point Borg scale (6 = no exertion; 20 = maximal exertion), a method to measure symptoms of fatigue during exercise.[21] This Borg scale has already been used in CPR studies.[15, 16, 22] All participants were familiar with the scale before the investigation began.

Accu-Check Softclix pro pens and Accu-Check lancets (Roche-Austria, Vienna, Austria) were used to draw capillary blood from each participant's earlobe. The blood was taken using 32-μL ring caps (Hirschmann, Germany) and transferred to test strips that were analyzed with the device Accutrend Lactate (Roche-Austria).

A Nine Hole Peg Test[23] (NHPT) measured fine motor skills. This test consists of nine small pegs that have to be put into corresponding holes in a block and removed again as fast as possible. The time to complete the test is measured and can be compared to standard values for each hand and age group. The NHPT has been used in previous studies to measure manual dexterity after CPR.[15, 24]

Outcome Parameters

The main outcome reflecting physical exertion was the rate pressure product (RPP) as an index of cardiac work, akin to previous studies.[15, 16, 25, 26] For the RPP the rescuer's maximum heart rate during CPR was multiplied by the systolic blood pressure (sBP) immediately after CPR; further sBP was measured before and 5 minutes after testing.

Secondary physical exertion outcomes were mean heart rate during the last minute of CPR and the Borg scale[21] compiled at the third, sixth, ninth, and 12th minute of CPR. Each provider's capillary lactate concentration and NHPT for both hands were measured immediately after each BLS cycle.

Data Analysis

After a residual plot was used to determine the normal distribution of data, we used analysis of variance (ANOVA) or Mann-Whitney U-tests to determine differences in the RPP and further heart rate and blood pressure measurements: maximum heart rate during CPR, heart rate during last minute of CPR, sBP pre CPR, diastolic blood pressure (dBP) pre CPR, sBP post CPR, dBP post CPR, sBP 5 minutes post CPR, and dBP 5 minutes post CPR, as well as the Borg, lactate, and NHPT between the MRD and standard BLS group, including possible period and carryover effects. If a statistically significant (p < 0.05) period or carryover effect was detected, the second period was excluded from the analysis of the according characteristic. In these cases, the remaining first period was analyzed using unpaired t-tests or Mann-Whitney U-tests. Otherwise (no period or carryover effect) the primarily calculated ANOVAs or Mann-Whitney U-tests are directly used as our results. Data are presented as mean ± standard deviation (SD) or median and interquartile ranges for normally distributed and skewed continuous variables, respectively. p-values < 0.05 were considered statistically significant. SPSS 11.5 (SPSS, Chicago IL) was used for statistical analysis.

Sample Size Estimation

Based on the sample size calculated for the main study,[14] a power calculation was done on the RPP. The aim of the study was to detect a pairwise difference between standard BLS and MRD in RPP. Given the design-specific properties of a crossover study, we handled data as paired observations within each subject. The null hypothesis was that there would be no difference.

The power estimation for the difference between two dependent means was performed assuming a two-sided alpha level of 0.05 and a sample size of 80. Based on means and SDs of RPP (21,406 ± 4,514 in MRD and 20,016 ± 4,253 in BLS), the achieved power was calculated to be 0.80. No further sample size calculation for specific variables was performed. When the sample size is 80, a two-sided 95% confidence interval (CI) for the difference in paired means will extend 986.088 for RPP from the observed mean, assuming that the SD is known to be 4,500 and the CI is based on the large-sample z-statistic.



Eighty-one medical students volunteered for this study; due to data loss in one case we could analyze 80 participants' data sets (Figure 1). Mean (±SD) age was 23 (±3) years with a mean (±SD) body mass index of 22 (±3) kg/m2, 40 participants (50%) were female, and 77 (96%) were right-handed. All of them had sufficient BLS knowledge due to their medical training and all reached the predefined level of competence with both CPR methods.

Primary Outcome Parameter

The RPP using the MRD was 21,406 (±4,514) compared to 20,016 (±4,253) for standard BLS (p = 0.163; Table 1).

Table 1. Parameters Reflecting Exhaustion
VariableMRDStandard BLSp-valuea
  1. n = 80. Observation period, 12 minutes. Data are reported as mean ± SD or median [interquartile ranges].

  2. BLS = basic life support; CPR = cardiopulmonary resuscitation; dBP = diastolic blood pressure; MRD = mechanical resuscitation device; NHPT = Nine Hole Peg Test; RPP = rate pressure product; sBP = systolic blood pressure.

  3. a

    Differences between the MRD and standard BLS were analyzed using ANOVAs including period and carryover effects.

  4. b

    Maximum heart rate during CPR multiplied by the sBP immediately after CPR.

  5. c

    A period effect was detected and therefore the second period was discarded from analysis. The remaining first period was analyzed using unpaired t-tests or Mann-Whitney U-test.

  6. d

    n = 79 due to data lost.

  7. e

    n = 78 due to data lost.

  8. f

    A period and carryover effect was detected and therefore the second period was discarded from analysis. The remaining first period was analyzed using Mann-Whitney U-test.

Heart RPPbcd 21,406 ± 4,51420,016 ± 4,2530.163
Maximum heart rate during CPRcd (beats/min)151 ± 23145 ± 240.299
Heart rate during last minute of CPR (beats/min)135 ± 26139 ± 220.027
Pre CPRc
sBP131 ± 16127 ± 14≤0.001
dBP80 ± 1178 ± 120.926
Post CPRc
sBP141 ± 15136 ± 130.925
dBP79 [72.2–86.7]79 [68.2–86.7]0.448
5 minutes post CPR
sBPce127 ± 14124 ± 140.472
dBPef74 [68–79]76 [70–82.2]0.041
Borg last value during CPR15.1 ± 2.414.6 ± 2.60.027
Lactate post CPR (mmol/L)3.4 ± 1.52.1 ± 1.3≤0.001
NHPT post CPR (seconds)
Right hand16 ± 216 ± 20.233
Left hand17 ± 217 ± 20.601

Secondary Outcome Parameter

Using the Borg scale, participants rated their subjective physical strain during the last minute of each 12-minute test as more intense with the MRD (mean ± SD = 15.1 ± 2.4) than with standard BLS (mean ± sd = 14.6 ± 2.6; p = 0.027). The subjective exertion increased over the CPR time, and we found significant differences in both groups at each time interval (Figure 2). Serum lactate concentrations were higher using the MRD (3.4 ± 1.5 mmol/l]) than with standard BLS (2.1 ± 1.3 mmol/l; p ≤ 0.001).

Figure 2.

Mean subjective exertion measured by Borg scale over 12 minutes of CPR. Possible scores range from 6 (no exertion) to 20 (maximum exertion). The solid line denotes the MRD group; the dashed line denotes the standard BLS group. The bars represent SDs. BLS = basic life support; MRD = mechanical resuscitation device. Significant difference between MRD and standard BLS at any time (3 minutes, p ≤ 0.001; 6 minutes, p ≤ 0.001; 9 minutes, p = 0.004; 12 minutes, p = 0.027).

The mean (±SD) heart rate during the last minute of CPR with the MRD was 135 (±26) beats/min compared to standard BLS at 139 (±22) beats/min (p = 0.027). No difference was found between the two methods in regard to the NHPT for the right or for the left hand (Table 1).


The results of our study confirm the physical challenges of CPR. We found no statistically significant difference between the two methods for our primary outcome parameter the RPP of the rescuer. That is somewhat surprising, since the MRD was designed to alleviate rescuer effort during BLS. Participants rated their subjective physical exertion on the Borg scale as slightly more intense when using the MRD compared to standard BLS, and capillary lactate concentrations (a parameter measuring physical exertion) were significantly higher after the MRD was used compared to standard BLS. That lactate concentration was similar to compression-only CPR of 8 minutes, as reported by others.[15, 16] A reason why CPR with the MRD was more exhausting than standard BLS might be the differences in CPR performance between the two groups. In the primary study, we found a significantly lower absolute hands-off time of 79 (±40) seconds for the MRD group, compared to 264 (±57) seconds with standard BLS, translating into a significantly higher number of “effective compressions.”[14] The main reason for this is that during standard BLS, the single rescuer must discontinue chest compressions to apply ventilations and therefore move between the victim's head and chest. This is not necessary if CPR is applied with the MRD, rendering it a continuous strenuous activity and leading to more exhaustion. Such a decrease in hands-off time could therefore not only increase the CPR performance, but also lead to more exhaustion. That means effective CPR attempts using either a MRD or standard BLS seems to be a considerable challenge to unfit rescuers. Our findings underscore the importance of physical fitness for CPR of any kind, as previous investigations have already suggested.[7] Physical exhaustion during standard CPR might decrease the number and depth of the chest compressions, increase incorrect decompressions, and lead to an incorrect compression point. The design of the MRD provides increased control over these parameters; thus it might limit the margin for such errors. Yet improving physical fitness might have a positive effect on CPR efforts,[7] independent of the CPR method.

The mean heart rate during the last minute of the 12-minute CPR period was higher during standard BLS. The measured heart rates were comparable to the findings of other investigations after a CPR cycle of 18 minutes.[7] Despite its statistical significance, the clinical relevance of the observed difference of four beats/min is uncertain.

The NHPT was not only included into the study as a surrogate parameter for measuring exertion, but as a test of manual dexterity after physical exhaustion.[24] Fine motor skills during CPR are certainly relevant for advanced life support measures such as endotracheal intubation or administration of medication. According to our data, gripping and pushing the lever of the device did not result in a decrease in fine motor skills. Both CPR methods resulted in comparable NHPT measurements. This raises the question of whether the NHPT, usually used for neurologic evaluation in ataxia[27] and multiple sclerosis,[28] is sensitive enough to detect marginal difference in the effects of exertion on fine motor skills during CPR.

The extensive training and competence testing before data collection is one of the study's strong points, since this homogenized the level of skill among the participants. Also, the testing period of 12 minutes is very realistic,[14] and the sample size of 80 subjects in a crossover trial is considerable, compared to similar studies.


This study was experimental and performed on manikins. The rescuers' stress level is considerably higher during real CPR, affecting physiologic parameters. In addition, the study cohort had medical background knowledge and a mean age of about 23 years, which may not be the case for all potential users of MRD. Due to the relatively young age of the participants, we did not evaluate their physical fitness before the study.

This study was performed before the 2010 ERC guidelines were published. The now higher compression rate and the deeper compression depth of 5 to 6 cm could have intensified the effect of our findings.


Our findings confirm that cardiopulmonary resuscitation is physically demanding even with support of a mechanical resuscitation device. While cardiopulmonary resuscitation performance in the mechanical resuscitation device group was superior,[14] there was no clinically relevant difference in the level of exhaustion during cardiopulmonary resuscitation with the mechanical resuscitation device compared to standard basic life support in this single rescuer scenario. If this kind of mechanical resuscitation device is used for prolonged life support, frequent changeovers with a second rescuer, as suggested for standard cardiopulmonary resuscitation,[17] should be strongly considered.

The authors thank all medical students who participated in the study and Eva Hochbrugger, Claudia Maurer, and Peter Karner, who helped with data processing. Thanks also to Julie Tremetsberger for the English proofreading.