Increase in pre-shock pause caused by drug administration before defibrillation: An observational, full-scale simulation study☆

1. Background 

The critical importance of sufficient circulation during cardiopulmonary resuscitation has led to increased efforts to decrease no-flow time.1, 2, 3, 4, 5 No-flow time refers to the period during cardiac arrest without sufficient circulation, and is equivalent to the time without chest compressions. Several initiatives have been taken to reduce no-flow time (also referred to as “hands-off time”) in the treatment recommendations published by the International Liaison Committee on Resuscitation (ILCOR). Based on ILCOR recommendations, the European Resuscitation Council (ERC) and the American Heart Association (AHA) now recommend a compression–ventilation ratio of 30:2 (rather than previously 15:1), and only one defibrillation between each 2-min series of basic life support (BLS) (rather than three).6, 7, 8, 9, 10, 11, 12 Furthermore, the AHA recently added a “hands-only CPR” recommendation to their guidelines.13

During advanced life support (ALS), BLS is interrupted every 2min to assess the cardiac rhythm and to guide therapy. For non-perfusing tachy-arrhythmias (e.g., ventricular fibrillation (VF) and pulseless ventricular tachycardia), therapy includes defibrillation alone or drug administration and defibrillation together.8, 9, 12 Defibrillation and drug administration are potentially life saving actions, but they may also increase hands-off time if interrupting chest compressions. Thus, it is necessary to weigh the disadvantages of interrupting chest compressions against the advantages of other actions.

The no-flow time that starts from the interruption of chest compressions until the delivery of DC-shock is referred to as the pre-shock pause. A shorter pre-shock pause increases the chance for successful defibrillation and, accordingly, the return of spontaneous circulation (ROSC).14, 15, 16, 17 Current recommendations state that pre-shock pauses should not exceed 10s.18

The recommended timing for drug administration during ALS differs between authorities, including the ERC, AHA, and the Norwegian Resuscitation Council (NRC) (Fig. 1). The ERC and AHA agree that drugs should be administered just before delivery of DC-shock9, 12; however, this apparent agreement encompasses two different interpretations. As a safety precaution, the ERC discourages touching the patient while charging the defibrillator7, 9; consequently, drug administration is to occur prior to charging the defibrillator. In contrast, the AHA recommends continued chest compressions – and therefore drug administration – during the charge phase.12 Yet another recommendation adds to the discordance as the NRC recommend drug administration happens 60s after defibrillation, i.e., during the following cycle of BLS.19

View Large Image Download to PowerPoint

Fig. 1. Comparison of advanced life support (ALS) guidelines. The figure shows the interpretation of ALS guidelines from the American Heart Association (AHA), the European Resuscitation Council (ERC), and the Norwegian Resuscitation Council (NRC).

The rationale for delaying drug administration to 60s after defibrillation is dual. One part is pharmacokinetic and pathophysiological aspects of resuscitation; the other is cognitive considerations about complexity of guidelines.

The pharmacokinetic and pathophysiological arguments against administering adrenaline just prior to defibrillation are, according to the NRC, that myocardial perfusion ceases almost immediately when chest compressions are interrupted,20, 21 and that it takes up to 90s to restore adequate perfusion of the heart after resuming chest compressions.20, 21 Thus, adrenaline injected into a peripheral vein only reaches its peak concentration after 90–150s.22 As such, there may be no immediate benefit from adrenaline injected just prior to defibrillation.9

The cognitive considerations about guideline complexity can be summarised to questioning if the focus on drug administration may shift focus away from defibrillation and thereby increase the pre-shock pause.19

In theory, drug administration in itself should not significantly influence the pre-shock pause since teams should use the BLS cycles to prepare drugs for subsequent injection.9 Further, intravenous injection of 1ml (as in the case for 1mg adrenaline) should be very quick, as should the mandatory saline flush9; however, it remains unclear if drug administration affects the pre-shock pause in practice. Therefore, the aim of our study was to compare the pre-shock pause with and without drug administration in a full-scale simulation study.

2. Methods 

The data used in this paper originated from our previous observational study of junior physicians’ skills and behaviour during simulated resuscitation.23 The study included 72 participants who had graduated within 5 years and were working in internal medicine departments with acute admissions. Participation was voluntary, informed consent was obtained, and data were kept confidential.

The simulations took place in a working ambulance with genuine equipment and personnel (a paramedic and an emergency medical technician). A computer-controlled manikin with simulated cardiac rhythm, respiration, peripheral blood saturation, and blood pressure, was placed on the stretcher (ResusciAnne Simulator & Laerdal PC SkillReporting System, Laerdal Medical, Norway). Supplemental oxygen, intravenous accesses, electrocardiogram, and self-adhesive defibrillation pads were established in advance.

The scenario involved a patient case with acute coronary syndrome in need of percutaneous coronary intervention at a specialised cardiac department. During transfer, the patient experienced a ventricular fibrillation (VF) cardiac arrest that was refractory to treatment for 5min. In the following 3min, defibrillation would invoke ROSC. If not defibrillated within these 3min, ROSC would appear no later than 8min after the onset of VF. In all simulations, the used monitor/defibrillator, LIFEPAK-12 (Physio-Control, USA) was of the same type and set to manual mode before initiating the simulation (in contrast to advisory mode).

2.1. Data and statistics 

Video recordings from a digital surveillance camera mounted in the ambulance documented all simulations, and recordings were continuously time-stamped by the camera with a built-in on-screen digital clock. Calculation of intra- and inter-observer variability was done by random selection of three simulations that were reviewed (in their full length) twice by two independent persons (a physician and a medical student) and by the first author of this paper. The inter- and intra-observer coefficients were calculated using Stata/IC 10.1 (StataCorp, USA).

Events identified in the video recordings included time to onset of VF, all ventilations, start/stop of all series of chest compressions (including the number of chest compressions in each series), and time for defibrillation and drug administration (Fig. 2).

View Large Image Download to PowerPoint

Fig. 2. Example of registrations. The figure illustrates a segment of a simulation exercise. The diagonal numbers represent net time in the simulation, starting with the debut of the ventricular fibrillation at the time 00:00:00 (hours:minutes:seconds).

Sequences that included defibrillation and/or drug administration were identified in the dataset and assigned to one out of four categories: DC-only, Drug+DC, Drug-only, and Drug-during (Table 1). If a pause in chest compressions included several defibrillations and/or several drug administrations or if in correct numbers but in the wrong order, they were excluded.

Table 1. Defibrillations and/or drug administrations were assigned to one out of the four categories listed in the table. The table shows the actions included in each of the four categories.

	Category	Action
		Chest compressions	Drug administration	Defibrillation
	DC-only	Interrupted	No	Yes
	Drug+DC	Interrupted	Yes	Yes
	Drug-only	Interrupted	Yes	No
	Drug-during	Ongoing	Yes	No

For the two categories including defibrillation (DC-only and Drug+DC), the pre-shock pause was calculated (Fig. 3). In order to compare sequences including defibrillation and sequences including drug administration, but not defibrillation, the term “pre-drug pause” was introduced (Fig. 3): The pre-drug pause describes the hands-off time from chest compressions are interrupted until drugs are administered and thus represent an equivalent to the pre-shock pause calculated in events including defibrillation. Time used for drug administration during chest compressions (Drug-during) was by definition zero.

View Large Image Download to PowerPoint

Fig. 3. Illustration of pauses and categories used in analysis. The upper part of the figure illustrates two different pre-shock pauses: upper left illustrates a sequence with only defibrillation (DC-only) and upper right a sequence with drug administration and defibrillation (Drug+DC). The lower left part of the figure shows a sequence of a pre-shock pause, when only drugs are administered instead of a DC-shock, that is, a pre-drug pause (Drug-only). The lower right of the figure illustrates drug injection during chest compressions (Drug-during).

In order to prevent results from being skewed by paired observations, the median value in each category (DC-only, Drug+DC, and Drug-only) was calculated for each simulation: If a simulation included more than one case in a single category, e.g. three cases of defibrillation alone (DC-only), the median duration of the three pre-shock pauses was calculated and used for further analysis.

Two comparisons were done: First, the duration of the median pauses in the three categories (DC-only, Drug+DC, and Drug-only) were compared using one-way analysis of variances. Second, intra-individual comparison between the pre-shock pause in DC-only and Drug+DC sequences was calculated for all simulations including both DC-only and Drug+DC sequences using paired t-test. Both comparisons were made using GraphPad Prism 5.02 (GraphPad Software, USA). All time values are given in seconds as median (lower; upper quartiles) [minimum–maximum].

3. Results 

Almost all of the 72 simulations (68 (94%)) included DC-only sequences, while only 24 (33%) included Drug+DC sequences. Drug-only sequences were found in 33 (46%) of the simulations (Fig. 4). Furthermore, 24 (33%) simulations included drug administration during ongoing chest compressions (Drug-during). The three sequences (DC-only, Drug+DC, and Drug-only) did not necessarily happen in each simulation; consequently, the numbers in each group do not add up to the 72 simulations described

View Large Image Download to PowerPoint

Fig. 4. Time in seconds since interruption of chest compressions to DC-shock (DC-only (n=68) and Drug–DC (n=24)) or drug injection (Drug-only (n=33)), respectively. The three sequences depicted (DC-only, Drug+DC, and Drug-only) did not necessarily happen in each simulation; consequently the numbers in each group does not add up to the 72 simulations described. The differences were statistically significant (one-way analysis of variances, p≪0.001). Values depicted are medians, upper/lower quartiles, 5–95 percentiles (whiskers), and outliers (dots).

In the simulations including DC-only sequences (n=68), the median pre-shock pause for DC-only sequences was 18s (14; 23) [1–38]. In the 24 simulations including Drug+DC sequences the median pre-shock pause for Drug+DC sequences was 32s (21; 36) [12–85]. Drug administration alone happened in 33 simulations and had a median pre-drug pause of 6s (3; 12) [1–41]. The difference between the three groups was statistically significant (one-way analysis of variances, p≪0.001) (Fig. 4). The relative increase in pre-shock pause between DC-only and Drug+DC was 78% (14/18).

A total of 22 (31%) of the simulations included both DC-only and Drug+DC sequences, which made it possible to compare intra-individual differences between the pre-shock pause in DC-only and Drug+DC sequences. In those 22 simulations, the median pre-shock pause when only defibrillation was done (DC-only) was found to be 17s (15; 22) [6; 38] compared to 25s (21; 36) [12–51] when drugs were administered prior to defibrillation (Drug+DC) (Fig. 5). The difference in means was statistically significant (p≪0.001, paired t-test).

View Large Image Download to PowerPoint

Fig. 5. Pre-shock pause. The figure illustrates the pre-shock pause in the DC-only and the Drug–DC-groups for the 22 simulations that included both type of sequences. The difference in pre-shock pause was statistically significant (paired t-test, p≪0.001). Values depicted are medians, upper/lower quartiles, 5–95 percentiles (whiskers), and outliers (dots).

Only two physicians performed the recommended flushing with saline after drug administration. Raising the limb after drug injection9 happened only once.

The intra-observer variability coefficients were 0.9966, 0.9981, and 0.9971, respectively, and the inter-observer variability coefficients were 0.9897, 0.9913, and 0.9906, respectively.

4. Discussion 

In this observational, full-scale cardiac arrest simulation study, we found an increase in the median pre-shock pause when drugs were administered prior to defibrillation (25s) compared to defibrillation alone (17s) in 22 simulations that included both sequences. This 8-s difference in medians is equivalent to a 47% relative increase (8/17) in the pre-shock pause. Further, comparison of all 72 simulations revealed a relative increase in pre-shock pause between DC-only and Drug+DC at 78% (14/18).

This is an important finding given that a prospective, multi-centre, observational study of cardiac arrest showed survival to be 71% if the pre-shock pause was within 10.2–20.0s, but only 60% if the it was within 21.1–30s.14

Pre-shock pauses shorter than 10s might not be attainable in all settings. We found pre-shock pauses of 17 and 25s in the Drug-only and Drug+DC groups, respectively, and our findings are supported by those of other studies that found median pre-shock pauses of 15s,1 15.3s,14 11s,24 and 17s.25

The findings of pre-shock pauses longer than 10s need attention. One explanation is that the charging time of the defibrillator prolongs the interruption of chest compressions. This will be the case if guidelines that discourage touching the patient during charging the defibrillator are followed (ERC and NRC guidelines, Fig. 1). Another explanation for longer pre-shock pauses may be inappropriately designed equipment.26

A third and perhaps more important explanation may be that the less experienced physician needs more time than the more experienced to assess the cardiac rhythm before deciding whether defibrillation is appropriate.

In this study, we observed that performing two resuscitation actions instead of one correlates with a significant increase in pre-shock pause. Information overload, i.e., a situation with more incoming stimuli than the physician has cognitive abilities to process, may be one explanation.27, 28, 29, 30

Drug administration is not only a matter of emptying the syringe, but also entails other actions, such as flushing with saline (10–20ml) and raising the limb in the air for 10–20s.9, 12 Thus, after deciding on appropriate drug administration, the physician almost instantaneously has to estimate if this (injection, flushing, raising the limb, etc.) will cause the patient to be left without chest compressions for more than 10s, and, in that case omit drug administration.9, 12, 18

To the experienced physician, such actions may be implicit, but to the less experienced physician substantial efforts may be necessary in order to recall these details. Further increasing the number of information units to be processed within the very short time frame from chest compressions are interrupted to defibrillation is supposed to be done, constitutes the core concept of “safe defibrillation”. Safe defibrillation is advocated by the ERC, which discourages touching the patient while charging the defibrillator (18, pp. 82–83); however, the ERC also recommends administering drugs while charging the defibrillator as a method of shortening the pre-shock pause.31

In our study we observed, that when drugs were administered apart from defibrillation, but during a pause in chest compressions, the median pre-drug pause was 6s. Thus, performing one instead of multiple procedures can be associated with shorter pauses in chest compressions. This supports our concerns about information overload under the current guidelines, a question also raised by Meertens et al.32 who suggested that drug administration just before defibrillation may divert focus from defibrillation to drug administration, thereby prolonging the pre-shock pause.

If the onset of the effects of adrenaline does not occur until minutes after administration,9, 20, 21, 22 it should be carefully considered if this equals that the timing of administration could be changed without negative consequences. It could be argued that drug administration during BLS is merely relocating the time without chest compressions. A counter-argument is the immense impact that the length of the pre-shock pause has on the chance for successful defibrillation.

In our study, we saw 24 simulations that included drug administration during ongoing chest compressions (Drug-during). One interpretation is that the physicians intuitively prioritised chest compressions higher than drug administration and therefore chose to administer drugs during ongoing chest compressions instead. Another, of course, would be that they did not know guidelines.

A limitation in this study is that performance during simulation is not equal to performance in real life. A manikin is not able to simulate all vital signs, skin pallor and temperature are just some of the signs missing as well as it is clear that human life is not at stake.33 The link between real life and the results presented in this paper is that the simulations were held within the actual context. The setup was a genuine ambulance, and the fellow players in the simulation were real ambulance crews. The drugs, syringes, needles, and fluids were real. Naturally, the manikin was not real. However, the chest moved, the pulse was palpable — and vivid communications took place between the physicians and the simulated patient through the first minutes of the simulations.

Life-like surroundings and a patient case representing a frequent challenge to the junior physician constitute a major strength to this study. In the light of this, and the fact that simplified guidelines have been shown to increase guideline adherence,34, 35, 36 it seems reasonable to carefully consider if it is possible to reduce the number of actions to be performed during the pre-shock pause, thereby reducing information overload. In order to further elucidate this point, it will also be necessary to perform studies that focus on only single actions in order to decide whether the action by itself has latent weaknesses.

5. Conclusions 

Observations from our simulation study show that the pre-shock pause is considerably longer in cases that include intravenous drug administration prior to defibrillation (25s) compared to cases that include only defibrillation (17s). Considering the recommended maximum 10-s delay, both pauses were too long; however, the delay related to drug administration is noteworthy as the pre-shock pause represents time without sufficient circulation.

Our results suggest that altering the resuscitation sequence from administering drugs before defibrillation to after defibrillation may improve the outcome of resuscitation. As Kern et al. concluded in 2002, “Any technique that minimizes lengthy pauses in chest compressions […] should be given serious consideration”.17

Conflict of interest statement 

None to declare.