Automatic compression improves adherence to advanced life support protocol in two-paramedic team. A randomized simulation study

1. Introduction

One-year survival rates for sudden cardiac arrest (SCA) victims remain poor, despite some improvements in trends over time from 1985 to 2018 [1]. Only immediate delivery of high-quality chest compressions (CC) and defibrillation in cases of ventricular fibrillation (VF) followed by appropriate implementation of post-resuscitation care have been linked to to improved survival and neurologic outcomes [2].

The Advanced Life Support (ALS) protocol, in addition to high-quality CC and ventilation, also includes critical elements such as heart rhythm analysis, defibrillation, intravascular access and drug administration. Although basic procedures are essential, recent studies have revealed benefits when basic life support (BLS) is followed by administration of ALS protocols within 11 minutes of CPR [3]. Adrenaline improved survival until hospital discharge and resulted in a meaningful clinical outcome according to a 2019 review and meta-analysis. [4]. However, cumulative doses of adrenaline may provoke vasoconstriction leading to stroke or myocardial ischaemia, prompting the recommendation for administration at proper intervals. Continuous assessment of the quality of resuscitation should be performed during the whole action. There is also a need to ensure proper diagnosis for potentially reversible causes of cardiac arrest. Performing these various tasks in a short period of time requires excellent organization and is the basis for the effectiveness of the resuscitation team. Furthermore, exposure of paramedics to actual resuscitation cases is low [5]. In a central European agglomeration, paramedics’ response to SCA was estimated at 1% of all emergency ambulance responses [6]. Even the presence of experienced staff does not always guarantee high-quality care. These aforementioned factors oblige practitioners to find methods that will optimize the quality of resuscitation [7].

Protocols and guidelines ensure high-level care and reduce the risk of human errors when various tasks must be done effectively and urgently. A guideline recommendation is defined as any statement that promotes or advocates a particular course of action in clinical care [8]. The protocol is a pathway of treatment developed based on guidelines. It indicates step by step what activities should be performed. It has been shown that the use of protocols increases healthcare professionals’ adherence to guidelines [9]. Even when key personnel are present, adherence to the specific content and timing of guidelines is often unsatisfactory [10]. Ebben et al. indicate that in life-threatening situations, adherence to international emergency guidelines shows a wide variation [11]. McEvoy et al. found that the number of wrong actions undertaken during ALS correlated with survival rate [12]. Consequently, Cheskes et al. strongly recommended strategies to improve overall guideline compliance that might significantly impact outcomes after out-hospital cardiac arrest (OHCA) [13].

Automated Chest Compressions Devices (ACCD) seem to be a promising method that can improve quality of CC compared to manual compressions provided by rescuers, both in timing and compression depth. ACCD may also be useful for continued resuscitation during prolonged procedures, transport of patients, and advanced diagnostics or treatment procedures like computed tomography or percutaneous coronary intervention. Currently, there is no consensus on whether the use of these devices actually improves the outcome of SCA cases [14, 15, 16].

The aim of this study was to evaluate if the use of ACCD in resuscitation provided in two-paramedic teams helps in adherence to ALS protocols.

2. Materials and methods

2.3 Interventions

Each team completed the same scenario twice, once providing manual resuscitation (control group - MAN), and once using the automated CC device (experimental group - ACC). The LUCAS 2 Chest Compression System (Physio-Control, Redmond, Washington, USA) was used. Between the scenarios, the teams had at least 20 minutes for complete physical recovery. The flowchart of this study according to Consolidated Standards of Reporting Trials (CONSORT) statement is presented on Fig. 1.

Fig. 1.Study flowchart according to CONSORT statement.

2.6 Variables

Our primary outcomes were: duration of resuscitation loop; time and duration of heart rhythm analyses and its deviation from the pattern; time of administering the first dose of epinephrine and interval between its subsequent doses. The clock was stopped after 10 minutes even if a scenario with mandatory actions was not completed. All the findings were included in further statistical analysis.

For the purpose of the study, the timeline of model resuscitation was created according to ERC Guidelines 2015. This pattern is presented in Fig. 2. It was assumed that the first dose of adrenaline should be administered as soon as possible-subsequent doses optimally in 3-5 minutes intervals.

Fig. 2.Timeline model for the research protocol.

Our secondary outcomes were insertion times of intravenous line and SAD, as well as CCF.

To validate the simulation model and to check for any bias related to the scenarios sequence (MAN followed by ACCD vs. ACCD followed by MAN), at the beginning of our analysis, we compared the findings (both primary and secondary outcomes) in two subsets of MAN group, one in which manual chest compression scenario (MAN 1) was the first one versus the second one (MAN 2) (seen Table 1).

Table 1.A Baseline demographics of participants randomized to study groups of (52 paramedics, 26 teams) each.

*	Manual compressions first	Automated compressions first
Age (years, mean $\pm$ standard deviation)	31.11 $\pm$ 4.65	30.73 $\pm$ 3.29
Profesional experience (years, mean $\pm$ standard deviation)	9.25 $\pm$ 4.22	9.00 $\pm$ 3.46
Sex (Males, females)	39, 13	31, 21

3. Results

3.3 Primary outcomes

All events were completed within 600 seconds. The duration of the initial assessment of vital signs was comparable in both groups. In the group where ACCD was used, all times of analyzed parameters included in the study primary outcomes were markedly shorter (Table 3). Additional comparisons of the difference ($\mathrm{\Delta }$) between the time specified in the ERC protocol and the one reached by paramedics showed higher deviation from optimal (target) value in MAN group (Table 4).

Table 3.A comparison and the statistical summary of rhythm analyzes and loops time.

*	Control (MANUAL) [s]	Experimental (ACCD) [s]	P-value
ABC	16 $\pm$ 2	15 $\pm$ 2	0.0683
1${}^{st}$# analysis	29 $\pm$ 6	24 $\pm$ 7	0.0002
2${}^{nd}$ analysis	177 $\pm$ 38	158 $\pm$ 18	0.0024
3${}^{rd}$ analysis	311 $\pm$ 50	291 $\pm$ 21	0.0107
4${}^{th}$ analysis	449 $\pm$ 54	416 $\pm$ 28	0.0002
5${}^{th}$ analysis	551 $\pm$ 36	538 $\pm$ 25	0.0362
Loop time	143 $\pm$ 17	133 $\pm$ 11	$$ 0.0001
DA VF	5 $\pm$ 1	4 $\pm$ 1	0.0263
DA PEA	7 $\pm$ 1	8 $\pm$ 2	0.0035
*All variables are presented as the means $\pm$ standard deviation. #analysis - number of following heart rhythm analyzes. Abbreviations: ABC - , ACCD - automatic chest compression device, DA PEA - duration of analysis during pulseless electrical activity, DA VF - duration of analysis during ventricular fibrillation, ERC - European Resuscitation Council.

Table 4.Results of deviation time between recommended in ERC protocol and reached by study participants.

*	ERC	MAN [n = 52]	ACC [n = 52]	P value#
A1 vs ERC ($\mathrm{\Delta }$) [s]	20	10 (5 -13)	3 (0 - 8)	0.0002
A2 vs ERC ($\mathrm{\Delta }$) [s]	150	15 (9 - 31)	5 (0 - 10)	$<$ 0.0001
A3 vs ERC $\mathrm{\Delta }$ [s]	280	21 (4 - 60)	8 (0 - 17)	0.0035
A4 vs ERC ($\mathrm{\Delta }$) [s]	410	23 (9 - 73)	2 ((-11) - 21)&	0.0003
A5 vs ERC ($\mathrm{\Delta }$) [s]	540	14 (1 - 39)	-1 ((-13) - 15)	0.0070
Loop time vs ERC ($\mathrm{\Delta }$) [s]	120	18 (14-23)	13 (7 - 17)	0.0003
*all variables are presented in seconds as the medians with interquartile ranges (1${}^{st}$ quartile - 3${}^{rd}$ quartile); #in the non-parametric Mann-Whitney U test. &negative values indicate that in some scenarios rhythm analyses were earlier than recommended by ERC. Abbreviations: A - analysis, ACC - automatic chest compression group, ERC - European Resuscitation Council, MAN - manual chest compression group.

The detailed dispersion for both groups is presented in Fig. 3 and Fig. 4.

Fig. 3.Dispersion chart for the control group.

Fig. 4.Dispersion chart for the experimental group.

Intervals between consecutive rhythm analyses in both studygroups were calculated. Interestingly, there were no differences in the intervals between first and fifth analysis (A5 - A1), probably because only 77% (the fastest 40 out of all 52) in the MAN group were able to carry out the fifth analysis. Therefore, we have also included analysis of A4-A1 interval as seen in Table 5.

Table 5.Intervals between consecutive heart rhythm analyses in seconds.

*	MAN	ACC	P value
Interval 1 (A2 - A1)	16 $\pm$ 2	15 $\pm$ 2	0.0167
	[n = 52]${}^{\mathrm{\#}}$	[n = 52]
Interval 2 (A3 - A2)	29 $\pm$ 6	24 $\pm$ 7	0.8871
	[n = 52]	[n = 52]
Interval 3 (A3 - A2)	177 $\pm$ 38	158 $\pm$ 18	0.0045
	[n = 52]	[n = 52]
Interval 4 (A3 - A2)	311 $\pm$ 50	291 $\pm$ 21	0.2378
	[n = 40]	[n = 49]
Interval A5 - A1	551 $\pm$ 36	538 $\pm$ 25	0.1683
	[n = 40]	[n = 49]
Interval A4 - A1	143 $\pm$ 17	133 $\pm$ 11	0.0012
	[n = 52]	[n = 52]
*All variables are expressed in seconds as the means $\pm$ standard deviation. ${}^{\mathrm{\#}}$ values within the square brackets indicate how many teams reached a given point of heart rhythm analysis. Abbreviations: A - heart rhythm analysis, ACC - automatic chest compression group, MAN - manual chest compression group.

During either manual or automated compressions, the time that paramedics spent for a reevaluation of the manikin was within the normal range. The differences between groups were only one second, but this result was found to be significant when PEA was presented on the monitor. In both groups, the rhythm analyses were longer for PEA than for VF (7 $\pm$ 1 s and 5 $\pm$ 1 s in MAN, 8 $\pm$ 2 s and 4 $\pm$ 1 s in ACC, respectively). The comparison of the described above findings is presented in Table 3.

The mean duration of a single resuscitation loop for MAN vs. ACC was 143 $\pm$ 17 s vs 133 $\pm$ 11 s (P value if available?). The difference between MAN and the recommended ERC protocol time was 23 s, significantly longer than that of ACC’s difference of 13 s (P = 0.0003).

The first and second doses of adrenaline (epinephrine) were administered earlier in ACC vs. MAN (272 $\pm$ 58 vs 232 $\pm$ 57 for the first dose and 486 $\pm$ 96 vs 424 $\pm$ 69 for the second dose). Adrenaline was administered three times in one MAN simulation but four ACC scenarios. Therefore the time of the third dose was excluded from further comparisons. All of these results were significantly different as presented in Table 6.

Table 6.Time [s] of epinephrine administration, 1${}^{st}$, 2${}^{nd}$ and 3${}^{rd}$ dose, time interval and IV line implementation.

*	MAN [s]	ACCD [s]	P-value
1${}^{st}$ dose of epinephrine	272 $\pm$ 58	232 $\pm$ 57	0.0014
2${}^{nd}$ dose of epinephrine	486 $\pm$ 96	424 $\pm$ 69	0.0007
3${}^{rd}$ dose of epinephrine	In 1 scenario	In 4 scenarios	n/a
Time interval${}^{\mathrm{\#}}$	230 $\pm$ 54	203 $\pm$ 40	0.0128
IV. line placed	183 $\pm$ 45	162 $\pm$ 35	0.0111
*All variables are presented as the means $\pm$ standard deviation. ${}^{\mathrm{\#}}$time between consecutive doses of epinephrine. Abbreviations: ACCD - automatic chest compression device, i.v. - intravenous, ACC - automatic chest compression group, MAN - manual chest compression group.

4. Discussion

To our knowledge, this is the first study measuring providers’ adherence to resuscitation protocols when providing automated vs. manual CCs.

The risk of a mistakes for providers working under stress and time pressures are high. Making key decisions with limited data is difficult. The 2015 International Liaison Committee on Resuscitation (ILCOR) Consensus on Science reported that there were no studies presenting optimal intervals between rhythm checks.

There is no strong scientific evidence to support the opinion that two-minute interval improves the number of return of spontaneous circulation (ROSC), survival to discharge, increase coronary perfusion pressure (CPP) or cardiac output. On the other hand, it has been estimated that after every two minutes, the quality of CC decreases. ERC recommends that CC should be paused every two minutes to assess the hearth rhythm [2]. Patients with the CCF of 60% and above had the highest chance to survive, with an adjusted OR for survival to the hospital discharge of 1.11 (1.01 to 1.21 5-95${}^{th}$% confidence intervals) per 10% increase in CCF. It has also been found that interruptions in CC for 10 s decreased CPP [18]. In addition, a pre-shock interruption longer than 15 s significantly compromises the outcomes of CPR and increases the risk of severity of post-resuscitation myocardial dysfunction [19].

Interruptions in CC are only allowed during activities that cannot be performed while moving the patient’s body [20]. These are: an assessment of heart rhythm, defibrillation and sliding of the endotracheal tube through vocal cords. Bjørshol found that there was not significant extension of noflow time in the first minutes of resuscitation. This problem occurred from the 7${}^{th}$ minute onwards [21]. In this study, however, in contrast to ours, CPR was conducted by a three-person team. Additionally, we did not assess the quality by at regular intervals of time.

The results presented in this study are in agreement with the ERC. The rhythm assessment was performed in less than 10 seconds, regardless of whether the ACCD was used or not. Previous studies reveal the most common causes of interruptions in CC were due to switching compressors (25%), pulse check (24%), and rhythm analysis (15%) [22]. The possibility of shortening the evaluation time translates directly into shortening of hands-off time. Currently, automatic methods available to assess the rhythm during compression have both high sensitivity (97%) and specificity 99% [23]. This technology, combined with the use of ACCD gives the chance to recognize shockable rhythms and perform defibrillation without any interruption in CC. While it is safe for a rescuer during automated CC, carrying it out during manual CC may harm a rescuer. It can be noticed, that the mean duration of a single resuscitation loop, was longer than it was supposed to be in both groups. However, in ACC, this time was more similar to the one defined in ERC Guidelines. In the latest American publications, some authors suggest that the 2-minute rhythm check is not essential for CC in patients with a non-shockable rhythm [24]. We did not assess the correlation between the duration of the loop and the quality of CPR. Such results have been presented, for example, by Sugerman et al. indicating that in 90. second CPR compressions were significantly shallower. While in the case of using ACCD, the quality of each compression is the same, in the manual compressions group, any extension of the loop will result in a deterioration of quality [25].

The difference in SAD insertion time was almost 1.5 minutes. Tomte et al. found no significant differences in time of intubation between manual and automated compressions groups [26]. In this research however, teams included emergency medical technicians, paramedics and ambulance physicians. Moreover, scenarios were two-tiered with a second team arriving 5 minutes after the first team. Airway patency is not a priority during resuscitation efforts. Manual maneuvers may be used in the initial phase of the action. However, other advantages of SAD should be considered. It allows providing CC and ventilation asynchronously even when performed by inexperienced medics [27]. This allows for minimizing interruptions associated with ventilation. It should be assumed that faster SAD insertion provides higher CCF value. That suggestion is strongly supported by Sanson et al. who found that during asynchronous CPR, higher ventilation rate, CCF, and lower CC rate per minute are delivered [28].

The role of adrenaline in SCA is still being discussed. The ERC 2015 Guidelines recommend adrenaline administration every 3-5 minutes. In non-shockable rhythms, the first dose should be administered as soon as possible. The mean interval between doses in our study was within recommended limits. Adrenaline administered in a 2-minute or shorter interval was associated with a better outcome when compared to longer time to the first dose [29]. Most teams in our study simulations administered adrenaline within the first 5 minutes. Nolan et al. found that drug administration is possible 10-20 minutes from incident onset [30]. In this study however, medical records of real events were analyzed and ambulance arrival time was important. It must be stressed that classic ampules and intravenous cannulas were used in our trial. A proper solution may be a routine application of intraosseous access and ready-to-use prefilled adrenaline syringes. However, this is a more expensive strategy.

Advanced procedures and pharmacotherapy are of importance, but they must not overtake high-quality CC. One can speculate that ALS procedures should not be initiated by a two-person team not equipped with ACCD or it should be postponed until a second team support is provided.

Our results suggest that in both groups, some protocol deviations occurred. Other authors imply that this is seen in nearly a half of CPR. McEvoy et al. proved that suboptimal timing of the actions, drug administration, omission indicated by ACLS protocol were associated with a lower chance to survive SCA and reach ROSC [12]. Johansson et al. found that in out-of-hospital cardiac arrest adherence to guidelines was lower than in-hospital cardiac arrest [31]. In this case, the number of team members in the resuscitation team may play a large role. In many countries, as in Poland, the ambulance is manned by two paramedics. If such a team were to lead the ALS standard resuscitation, a medical compromise should probably need to be used.

There are good quality studies indicating that ACCD can provoke a number of complications. For example, treatment with LUCAS was associated with higher rates of sternal and ribs fracture, severe soft tissue injury, and other serious intrathoracic injuries [32]. All of them may be unfavorable, especially if the patient achieves ROSC. Nevertheless, it should be emphasized that the use of ACCD significantly improved the compliance of the actions with the algorithm. It can be seen in better timing, faster analyzes and overall better adherence to the protocol, which was presented by the ERC experts. In addition, faster implementation of SAD led to increase in CCF, which has been shown to directly increase chances of survival. Taking into account the importance of procedures in current medicine, we think the benefits may outweigh the risks.

5. Limitations

Sudden cardiac arrest is a complex problem. There are many factors influencing survival. CPR should be performed according to an accepted pattern. Although medical simulation is a useful training tool, even high fidelity modeling will never be a complete reflection of real resuscitation scenarios. Performing procedures such as obtaining intravenous lines or SAD may be more time-consuming in real life, because of environmental pressures and working with patients whose anatomy varies. Additionally, participants were aware of taking part in an experiment that could have led to a Hawthorne effect, meaning they were trying to perform CC at their very best. Furthermore, our study included paramedics with a median experience of more than 9 years. We do not know if CC quality in general is different for less experienced, but usually younger paramedics. These aforementioned doubts should warrant further research in this area.

6. Conclusions

The use of ACCD may improve the quality of resuscitation performed in two-paramedic team. This can be achieved by improving adherence to ALS protocol, faster adrenaline administration and increasing CCF.

Ethics approval and consent to participate

The study protocol was approved by the Institutional Review Board of Poznan University of Medical Sciences (no: KB764/19). Written consent to participate in the study was obtained from each participant.

Author contributions

TK, MP designed the research study. TK, MD performed the research. TK, MP, ŁS, analyzed the data. TK, MP, MD, ŁS wrote the manuscript. MP, BP critically reviewed the manuscript and provided supervision. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript.

Acknowledgements

We would like to thank all the peer reviewers and editors for their valuable opinions and suggestions.

Conflict of interest

The authors declare no competing interests.