Aim Chest compression devices are useful during mountain rescue but may cause a delay in transport if not immediately available. The aims of this prospective observational study were to compare manual and mechanical cardiopulmonary resuscitation (CPR) during transport on a sledge connected to a snowmobile with a non-moving setting and to compare CPR quality between manual and two mechanical chest compression devices.
Methods Sixteen healthcare providers simulated four different combined CPR scenarios on a sledge in a non-moving setting and during transport and two mechanical chest compression devices during transport on the sledge. The study was conducted in May 2015 in a mountain in Norway. The primary outcome measures were compression rate (compressions per minute), compression depth in millimetres, leaning (incomplete chest wall release after compression in millimetres) and chest compression fraction (fraction of total time were compression were performed). The results were analysed by descriptive and graphical methods and paired t-tests were used to compare the differences between techniques.
Results We did not observe a significant difference between moving and non-moving conditions with respect to manual compression rate (p=0.34), compression depth (p=0.50) or leaning (p=0.92). However, both the manual compression depth (p<0.001) and the leaning (p=0.04) showed a significantly larger variance during the moving runs.
Conclusion Manual chest compression is possible on a snowmobile during transport even in challenging terrain. This experimental study shows that high-quality chest compressions and manual ventilation can be performed in an intubated patient during a short-term transportation on a sledge.
Statistics from Altmetric.com
What is already known on this subject?
Chest compression devices are useful during long and demanding resuscitations but may cause a delay in transport if not immediately available.
What this study adds?
Comparing manual cardiopulmonary resuscitation with two mechanical devices in actual mountain conditions, we found that high-quality chest compressions and ventilation can be delivered on a moving sledge during short-term transportation on a sledge.
Patients experiencing injuries and medical emergencies in remote areas are susceptible to hypothermia. Out-of-hospital cardiac arrest (OHCA) due to severe hypothermia is associated with increased survival with good neurological outcome, even after prolonged resuscitation.1–5 The European Resuscitation Council (ERC) recommends that some patients with OHCA should be transported with continuous cardiopulmonary resuscitation (CPR) to a hospital that can provide extracorporeal life support.6 In a terrestrial mountain rescue and retrieval situation, the quality of chest compression may be influenced by rough terrain, an unstable platform on which to perform the procedure (eg, stretchers and snowmobiles), fatigue, lack of resources/rescuers, darkness, altitude, cold and wet conditions. Recognising these difficult situations, it was recently suggested that a patient with a core temperature <28°C or unknown with unequivocal hypothermic CA should receive 5 min CPR alternating with ≤5 min without CPR. With core temperature <20°C, evidence supports alternating 5 min CPR and ≤10 min without CPR.7
An alternative to ensure continuous CPR in these conditions could be to use a mechanical chest compression device. These devices have been useful during prolonged transports in situations with limited resources or for personnel safety reasons.8 9 Chest compression devices are increasingly available to terrestrial rescue teams and, in many circumstances, may be transported to an accident site within reasonable time. If this is a possibility, the dilemma would be either to wait on site providing continuous CPR until device arrival or start transport with manual (and anticipated suboptimal) CPR. The aim of this study was to compare manual CPR quality during transport on a snowmobile sledge with a non-moving setting. In addition, we compared the CPR quality of manual compression versus two mechanical chest compression devices.
The field study and the controls were conducted in May 2015 at Hemsedal Ski Resort, Southern Norway. The temperature was 4.0° C (range 0.8–8.7), wind speed was 0 m/s and it was overcast with periods of light snow and rain.
The participants were 16 healthcare personnel (nurses, paramedics and doctors). Their median age was 39 years (range 25–54) and median body mass index (BMI) was 24.6 kg/m2 (range 21.4–32.1). All had experience in prehospital CPR. The evening before the study, the participants received 30 min of practical training on the chest compression devices, based on the manufacturers’ manuals, to learn how to deploy and operate the device. The duration of the training was less than the manufacturer’s recommendation.
Chest compression devices
We used two different battery-powered mechanical chest compression devices: Lund University Cardiac Assist System (LUCAS2, Physio-Control, Redmond, Washington, USA) and the Load Distributing Band (LDB) device AutoPulse (Zoll Medical, Chelmsford, Massachusetts, USA). The LUCAS2 is a chest compression device that provides compressions based on a piston inside a suction cup with an anterior posterior displacement of the sternum of 5.3 cm at a rate of 100 compressions/min and a duty cycle (compression:release) of 50%. The LDB device provides compressions based on a band that is placed semicircumflex on the chest and fastened to a board under the manikin. The anterior–posterior displacement of the chest is 20% at a rate of 80 compressions/min and a duty cycle of 50%.
The participants used the Resusci Anne QCPR Full Body (Laerdal Medical, Stavanger, Norway) with a SimPad SkillReporter V.5.7.0-79 (Laerdal Medical, Stavanger, Norway) to store and read data. This manikin system measured compression rate, depth, correct release for each compression, appropriate ventilation volume, the amount of compressions and ventilations, and the frequency and length of interruptions. It is important to note that compressions were detected by sensors built into the manikin to sense the correct compression spot, depth and leaning (incomplete chest wall release). These sensors are not designed for detecting LDB compressions other than rate; therefore, they will misinterpret LDB compressions as soft, with increased leaning and incorrect placement on the sternum.
Three professional ski patrollers operated two snowmobiles with an attached sledge. Each scenario started at an altitude of 923 m and descended along a 1.1 km track, down to 719 m. The average speed was 20 km/h (±3 km/h) and the descent time was about 3 min (±10 s). The track was part of a groomed and salted ski slope with two minor melting cracks (30 cm wide). An intubated Resusci Anne QCPR Full Body manikin was placed on each snowmobile sledge with feet in the heading direction in a transport bag and secured according to local protocol for patient transport. Before start, the rescuer performing manual chest compressions was positioned facing backwards on the sledge, straddling the manikin’s hips. The second rescuer was positioned by the manikin’s head and performed ventilations (see online supplementary figure B). Each participant simulated four combined CPR scenarios with chest compressions and manual ventilations of an intubated manikin (one non-moving and three moving) (see online supplementary figure C). The scenarios were: (1) manual chest compressions on the sledge in an indoor, non-moving setting; (2) manual chest compressions on a sledge during transport; (3 and 4) mechanical chest compressions on a sledge during transport with two types of chest compression devices (LDB AutoPulse CPR and LUCAS2 CPR). During the manual chest compression scenarios, the first rescuers were instructed to perform high-quality continuous chest compressions at a rate of 100 per minute. For all scenarios, the second rescuer performed one-hand bag endotracheal tube ventilation after every 10 compressions. During the scenarios, the participants were instructed to adjust the positioning of the compression machines if they deemed it necessary due to displacement. The primary outcome measures were compression rate (compressions per minute), compression depth (millimetres) and leaning (millimetres) defined as incomplete chest wall release after compression. Chest compression fractions were also measured (non-compression times were defined as more than 2 s without electronically registered compressions).
Data recording and preparation
To measure the outcome variables, we read the raw data for each compression and ventilation during the measurement process from the Sim Pad SkillReporter. This enabled us to remove some noise that generated artefacts during the preprocessing of the data. The recording of data started when the snowmobile started to move and ended when the snowmobile stopped after the ride. However, the first and last 10 compressions were excluded to extract the time window where the compressions and ventilations could be considered the most representative during transport. During each ventilation, we recorded an increased leaning and compression depth. This led to an initial overestimation of irregular compressions in relation to leaning, and an underestimation of irregular compressions in relation to compression depth. To eliminate these false measurements generated from the manikin’s lungs, we removed the first registered compression after each ventilation from the analysis of both leaning and compression depth. For frequency analysis and the computation of the chest compression fractions, we used all available data (see online supplementary figure A).
To describe compression depth and leaning for each run, we computed the median and the SD of the compressions for each test person. They were analysed by descriptive and graphical methods and compared between the different techniques by paired t-tests. For the graphical illustration of the depth and the leaning, we smoothed the measurement series by a moving average of five neighbouring compressions. Assuming a detectable mean difference in compression depth of 3 per cent points (SD=2.9), a significance level of 0.05 and a power of 0.08, we needed 16 test persons for a pairwise comparison. All computations were done using Matlab R2010a.
A total of 64 series (48 runs on the sledge and 16 non-moving) were performed. Three runs had to be repeated due to data recording failure. Table 1 shows the descriptive data for the runs. We did not observe a significant difference between manual compressions in moving and non-moving conditions with respect to compression rate (p=0.34), compression depth (p=0.50) or leaning (p=0.92) (figure 1A,B,D). However, both the manual compression depth (p<0.001) and the leaning (p=0.04) showed a significantly larger variation during the moving runs (figure 1C,E). The mean compression rate, depth, leaning and chest compression fraction with corresponding 95% CI are reported in table 1 and figure 1.
Both mechanical devices delivered chest compressions according to their technical specification. There was no significant difference in compression rate (p=0.12) or compression depth (p=0.84) between manual-moving conditions and the LUCAS2 (figure 1A,B), while the leaning was significantly higher during the manual moving conditions (figure 1D).
The smoothed trajectories for compression depth and leaning are shown in figure 2, where each line represents one test person. We observed a larger variation in compression depth both between and within the series for manual compressions compared with the LUCAS2 device. A similar variation was observed for leaning, but it was not as clear as for depth.
In this study, we investigated the quality of CPR during a typical situation in mountain rescue, that is, transport on a sledge. The main finding shows that high-quality manual CPR can be performed during transport. There were no significant differences in median compression rate, depth, leaning or chest compression fraction when comparing manual CPR on a moving snowmobile sledge with a non-moving setting. This finding may have clinical implications in an OHCA where transport CPR has to be performed and a mechanical compression device is not immediately available. The slope used in this study was part of a ski resort and had only minor bumps and melting cracks. Despite this, we observed a larger variation in compression depth and leaning for manual compressions compared with the LUCAS2 during transport. It is likely to believe that the differences between manual compressions and LUCAS2 will increase with rougher terrain.
The quality of chest compressions during ambulance transport has been documented to be suboptimal in both manikin models and patients. In a randomised cross-over design study, the percentage of correct compressions decreased from 54% on the floor to 21% (p<0.001) on a moving stretcher.10 Another study simulating maternal cardiac arrest showed that correctly performed compressions during transport over a level floor decreased from 93% to 32% (p=0.002).11 In a study simulating OHCA in an urban setting, the proportion of correct compressions decreased from 68% at the scene to 32% during transport down a stairway.12 In these three studies, nearly all the subjects walked adjacent to the stretcher while performing compressions. We assume the high quality of compressions in our study may be due to the relatively stable position where the rescuer straddled the manikin during transport. Reduction of movements due to legs locked under the sledge sidebars, a kneeling front position and padding around the patient may have enhanced the quality.
During resuscitation, the non-compression time (hands-off time) should be minimised.13 In a setting where CPR is performed during transport, there will be several phases where continuous compressions are challenging (removal from ground to stretcher, carrying the stretcher to the transport vehicle and start and stop of transport). In a retrospective observational study of 36 OHCA cases receiving manual CPR on site and during transport in an ambulance, the hands-off ratio increased from 0.19 to 0.27 (p<0.002).14 Our study did not include any of these phases and the findings will not be representative for a clinical setting combining relocation of the patient and the start and stop of transport.
The low temperature, microvibrations from the snowmobile, acceleration and deceleration during transport seemingly did not influence the performance of the machines. The differences between the mechanical devices regarding compression rate, depth and leaning are explained by their technical specifications. In addition, the increased leaning in the AutoPulse group is explained by the LDB over the chest that triggers the manikin sensors to detect leaning.
Despite the ability to deliver consistent high-quality CPR, meta-analyses have failed to show that compression machines improve patient outcomes.15 Two multicentre randomised clinical trials (LUCAS In Cardiac Arrest (LINC) Trial and Pre-hospital randomised assessment of a mechanical compression device in cardiac arrest (PARAMEDIC) study) concluded no difference in 4-hour survival and 30-day survival, respectively, between manual and mechanical compressions.16 17 The Circulation improving resuscitation care (CIRC) trial showed a significant benefit for survival in the mechanical compression group compared with manual compressions if the compression duration was longer than 16.5 min.18 Future research may look into the effect and feasibility of compression machines in other subgroups of OHCA.
The participants were highly motivated to perform high-quality CPR, but they were not specially trained for this type of transport CPR. We believe that other prehospital healthcare personnel with experience with OHCA would perform the same quality of CPR in a similar setting. The ERC Guidelines 2015 suggest that automated mechanical chest compression devices should not be used routinely to replace manual chest compressions, but are a reasonable alternative to high-quality manual chest compressions in situations where sustained high-quality, manual chest compressions are impractical or compromise provider safety. The most important is still to ensure high-quality chest compressions with adequate depth, rate and minimal interruptions, regardless of whether they are delivered by a machine or by humans.
The duration of each scenario was about 3 min and was short in distance (1.1 km); therefore, fatigue was unlikely to develop. In longer transports, fatigue inevitably will result in decreased quality of manual chest compressions and the benefits of a compression machine may be evident. More than 50 runs with snowmobiles were conducted in the same standardised track, but only minor bumps with minimal impact on stability and lurching occurred at the end of the day. The clinical relevance of this study can be questioned since snowmobile transport of OHCA victims is quite rare, but there is a wide range of OHCA transportations with similar conditions, where our findings may be applicable.
This experimental study shows that high-quality chest compressions and manual ventilation can be performed in an intubated patient during transportation on a snowmobile sledge for 3 min.
The study group wishes to thank the participants and the ski patrol at Hemsedal Ski Resort for exceptional attitude and endurance, despite a long day with demanding conditions.
Contributors All authors have contributed to the design of the study, writing the manuscript and the revision process. OT, SCC, JA, OO, GB and JKH performed the field study.
Competing interests LW is a NAKOS representative in the Medical Advisory Board for PhysioControl and a principle investigator for studies with the Load Distributing Band (AutoPulse) and Lund University Cardiac Assist System (LUCAS2) mechanical chest compression devices. The other authors have no conflict of interest to declare. The AutoPulse and LUCAS2 were borrowed from the manufacturers, but the study did not receive financial or other sorts of support from any manufacturer.
Ethics approval The study was approved by the Data Protection Officials for Research (2015/5342). The Regional Committee for Medical Research Ethics assessed the study and decided that no approval was needed.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.