Original Contribution

A comparison of 2 types of chest compressions in a porcine model of cardiac arrest^?

Jun-Yuan Wu MM, Chun-Sheng Li MD?, Zhao-Xia Liu MM, Cai-Jun Wu MM, Gui-Chen Zhang MM

Emergency Department, affiliated Chao Yang Hospital of Capital Medical University

Received 16 April 2008; revised 25 June 2008; accepted 2 July 2008

Abstract

Objective: Chest compressions performed by some medical workers are of poor quality, which are too few and shallow with incomplete release. This study was designed to compare the effects of these clinical quality chest compressions with standard manual chest compressions in a porcine model of cardiac arrest.

Methods: Ventricular fibrillation was induced in 18 pigs by programed electrical stimulation. Then, 40 mg Methylene blue was injected into right atrium after 4 minutes of untreated ventricular fibrillation (VF), followed by cardiopulmonary resuscitation for 9 minutes. Defibrillation was attempted at 13 minutes of cardiac arrest. Animals of no restoration of spontaneous circulation after 4 times of defibrillations were announced dead and dissected immediately to observe the cerebral perfusion with methylene blue coloration. Resuscitated animals were executed to remove the tissues of pallium, cardiac muscle, kidney, and liver for histopathology after evaluating a porcine Cerebral performance category score at 24 hours after cardiac arrest. All animals were randomized to the following 2 groups: (1) standard manual chest compressions group (n = 9)-chest compression rates were kept at 100 +- 5 cpm and compression depth at 50 +- 1 mm with complete release by Heartstart MRx Monitor; (2) clinical quality chest compressions group (n = 9)-chest compression rates were kept at 80 +- 5 cpm and compression depth at 37 +- 1 mm with incomplete release.

Results: Compared with clinical quality chest compressions, standard manual chest compressions produced greater restoration of spontaneous circulation, neurologically normal 24-hour survival, and histopathologic findings.

Conclusions: High-quality chest compressions improve outcomes of resuscitation, especially postresuscitation brain damage.

(C) 2009

^? The study was supported by capital medical science development scientific research fund (NO 2005-1006). We also appreciate professor Ai-Hu Wang who provided technical help in inducing VF with programed electrical stimulation instrument.

* Corresponding author. Chao yang district, Beijing 100020, China.

Tel.: +86 10 85231051; fax: +86 10 85231051.

E-mail address: [email protected] (C.-S. Li).

Introduction

That the Quality of cardiopulmonary resuscitation (CPR) impacts on the survival after cardiac arrest has been documented in some Animal experiments [1,2] and clinical quality researches [3-6]. Although we have recognized the importance of the Quality of CPR and made the newest CPR

0735-6757/$ - see front matter (C) 2009 doi:10.1016/j.ajem.2008.07.001

guidelines [7] in 2005, it is still difficult to radically improve the CPR quality. Some studies [2,8-11] have revealed that the quality of CPR even performed by medical staff in hospital is significantly inferior than that recommended in international CPR guidelines. The poor CPR quality is too few, and shallow chest compressions with incomplete release.

Giving audible or video feedback of CPR quality to rescuer can improve performance of chest compressions [12-14]. An automated external defibrillator (HeartStart MRx Monitor/Defibrillator, Philips Medical Systems, Hol- land), which is incorporated with a feedback system, has been invented by Laerdal Company and Philips Medical Systems. The chest pad sensor provided acceleration and force signals enabling accurate measurements of compres- sion depth, rate, and compression/decompression Duty cycle, which were continuously displayed on the screen. We used this instrument to perform standard manual chest compressions recommended by the 2005 guidelines and to make a model of clinical quality chest compressions in a porcine model of cardiac arrest. Then, we compared the effects of these 2 compressions.

Methods

Animal preparation

All trials abided by the administration regulation of Beijing on animal experimentation. Eighteen domestic pigs (2-3 months, 30 +- 5 kg) of either sex were used in this study. Animals were fasted overnight except for free access to water. Anesthesia was induced by Intramuscular injection of ketamine (20 mg/kg) and then by ear vein injection of propofol (1.0 mg/kg) and maintained in a surgical plane of anesthesia with intravenous infusion of pentobarbital (8 mg/ kg per hour) [15,16]. A cuffed 6.5-mm endotracheal tube was advanced into the trachea, and animals were mechanically ventilated with a volume-controlled ventilator (Servo 900c, Siemens, Germany) using a tidal volume of 15 mL/kg, a respiratory frequency of 12 per minute, and room air. End- tidal PCO2 (EtPCO2) was measured by an in-line infrared capnograph. Respiratory frequency was adjusted to maintain EtPCO2 between 35 and 40 mm Hg before inducing cardiac arrest and after restoration of spontaneous circulation . Room temperature was adjusted at 26 centigrade thermometer. A Swan-Ganz catheter (7F, Edwards Life Sciences, USA) was advanced from the right femoral vein and flow-directed into the pulmonary artery for the measurement of right atrial pressure (RAP) through a pressure transducer (Biosensors International Corp, Singa- pore), Continuous cardiac output (CCO), and mixed venous Blood oxygen saturation (SvO2) with CCO monitor (Vigilance II, Edwards Life Sciences). A catheter was retrograde inserted into the Right internal jugular vein for collection of vein blood for blood gas analyses. An

angiographic catheter was inserted from the femoral artery into the aortic arch for reference blood sample and for measuring aortic pressure (AOP) through a pressure transducer. Electrocardiogram (ECG), AOP, RAP, and EtPCO2 were monitored with HP monitor (M1165, Hewlett- Packard, USA). Another catheter was inserted into the left external jugular vein to place an electrode catheter to induce ventricular fibrillation by programed electrical stimulation instrument (GY-600A, Kai Feng Huanan Instrument Limited Company, China).

Experimental equipment

HeartStart MRx Monitor/Defibrillator with Q-CPR was used to monitor the Quality of chest compressions. It provides objective measurement with compression sensor, which measures the acceleration of chest during chest compressions, and an algorithm in the HeartStart MRx converts this to compression depth. Compression depth is presented as a waveform in screen that represents about 10 seconds of compressions, and compression rate (cmp) is also displayed in the upper left corner above the wave. As the chest is compressed, the compression is represented as a downward stroke of the wave, rebounding up to a baseline as compression pressure is completely released. Two horizontal lines in the wave sector drawn at -38 mm and -51 mm indicate the target zone to help achieve good compression depth in accordance with American Heart Association (AHA) guidelines.

Experimental protocol

After surgery, the animals were allowed to equilibrate for 30 minutes to achieve stable resting level. All animals were randomized to the following 2 groups before inducing cardiac arrest: (1) standard chest compressions group (S-CC) (n = 9)-the animals received standard manual chest compressions (the 2005 guidelines recommended) by Heart- Start MRx Monitor/Defibrillator, which helped to keep chest compression rates at 100 +- 5 cpm and compression depth at 50 +- 1 mm with complete release; (2) clinical quality chest compressions group (C-CC) (n = 9)-manual chest compression rates at 80 +- 5 cpm and compression depth at 37 +- 1 mm with incomplete release using the same monitor. VF was induced by programed electrical stimulation [17], mode S1S2 (300/200 milliseconds), 40v, 8:1 proportion, and

-10 milliseconds step length, and then mechanical ventila-

tion was discontinued. VF was verified by ECG and blood pressure. Electrode catheter was extracted after VF. After

4 minutes of untreated VF, 40 mg methylene blue was injected into left external jugular vein followed by CPR for

9 minutes. compression-to-ventilation ratios of 2 groups were both 30:2, and ventilations were performed by bag respirator with room air. Defibrillation (Heartstart MRx Monitor/Defibrillator, Philips Medical Systems) was attempted at 13 minutes of cardiac arrest. Defibrillation

Outcomes	S-CC (n = 9)	C-CC (n = 9)	P
ROSC	8/9	2/9	.015
Shocks before ROSC	1.4 +- 0.5	1.5 +- 0.7	.846
Duration of CPR before	586 +- 62	600 +- 84	.846
ROSC (s)
4-h survival	8/9	2/9	.015
24-h survival	8/9	2/9	.015
24-h and “normal” neurology	7/9	0/9	.002

shocks were administered at 120 J (Smart Biphasic) for the first attempt. All subsequent attempts used the 150-J dose. If the first defibrillation was unsuccessful, epinephrine (0.02 mg/kg) was given intravenously, followed by 2 minutes of CPR. After each 2 minutes of CPR, a 10-second pause was interjected to analyze rhythm and prepare for another defibrillation attempt. Additional epinephrine, if needed, was given at 3-minute intervals. Mechanical ventilation began at the beginning of the first defibrillation attempt with 100% oxygen and was continued until successful resuscita- tion, after which room air was used. Restoration of spontaneous circulation was defined as maintenance of a systolic blood pressure of at least 50 mm Hg for at least 10 consecutive minutes. Animals of no ROSC after 4 times of defibrillation were announced dead and were dissected immediately to observe cerebral perfusion with methylene blue coloration. After successful resuscitation, the animals underwent a 4-hour intensive care period when Ringer’s solution (20 mL/kg) was administered, and analgesics (Fentanyl) and antiarrhythmic (Amiodarone), if needed, were administered. With the exception of 1 jugular vein sheath that was used for fluid administration, all other vascular sheaths were removed. The animals were allowed to recover from anesthesia, placed in observation cages, and

Table 1 Outcomes

Table 2 Hemodynamics during CPR

monitored over the ensuing hours until 24 hours after resuscitation. Water was given during the observation period. Animals were evaluated a porcine Cerebral Performance Category score at 24 hours after cardiac arrest and then were killed with intravenous potassium chloride, the tissues of the pallium, cardiac muscle, kidney, and liver (the same place every time) were removed to deposit in 10% formaldehyde for pathology. The person who performed the histopatholo- gic examination belonged to the pathology department of capital medical university and was blinded to the groups from which the animals belonged.

Measurements

Hemodynamic data (cardiac output [CO], RAP, AOP, SVO2) and ECG were continuously measured and recorded. In the normal beating heart, coronary perfusion pressure was measured as the difference between the aortic diastolic pressure and right atrial diastolic pressure during diastole. During CPR, CPP was measured during the relaxation time of CPR. Arterial and internal jugular vein blood samples for blood gas and lactate levels analyses (GEM Premier 3000 Blood gas analyzer, Instrumentation Laboratory, USA) were drawn at baseline, VF 4 minutes, and CPR 3, 6, and 9 minutes. Mixed venous blood samples for hemoglobin (KX-21, Sysmex, Japan) were drawn at the same time. Blood oxygen content (mL/dL) was calculated by standard formula. Arterio- internal jugular venous oxygen content difference (Ca-ijvDO2) was calculated as the difference between arterial oxygen content (CaO2) (CaO2 = [Hb x 1.34 x (SaO2 / 100)] + (PaO2 x 0.0031)) and internal jugular venous oxygen content (CijvO2) (CijvO2 = [Hb x

1.34 x (SijvO2 / 100)] + (PijvO2 x 0.0031)). Cerebral Oxygen extraction ratios (O2ER) was calculated by the formula: O2ER = (CaO2 - CijvO2)/ CaO2 x 100.

Baseline		VF 4 min	CPR 3 min	CPR 6 min	CPR 9 min
CO (L/min)
S-CC	3.53 +- 0.26	0.25 +- 0.06	1.60 +- 0.16	1.73 +- 0.23	1.72 +- 0.18
	(3.35-3.71)	(.21-.29)	(1.49-1.71)	(1.57-1.89)	(1.60-1.84)
C-CC	3.56 +- 0.25	0.26 +- 0.06	1.25 +- 0.15 ??	1.34 +- 0.12 ??	1.35 +- 0.11 ??
	(3.40-3.72)	(0.22-0.30)	(1.15-1.35)	(1.26-1.42)	(1.27-1.43)
SvO2 (%)
S-CC	59 +- 2 (57-61)	20 +- 2 (19-21)	31 +- 3 (29-33)	32 +- 2 (30-34)	32 +- 2 (31-33)
C-CC	60 +- 3 (58-62)	20 +- 2 (19-21)	27 +- 2 ? (26-28)	27 +- 2 ?? (26-28)	28 +- 2 ?? (27-29)
MAP (mm Hg)
S-CC	105 +- 4 (102-108)	23 +- 2 (22-25)	53 +- 3 (51-55)	53 +- 3 (51-55)	52 +- 2 (51-53)
C-CC	102 +- 5 (99-105)	23 +- 2 (21-25)	35 +- 3 ?? (33-37)	37 +- 2 ?? (36-38)	38 +- 1 ?? (37-39)
CPP (mm Hg)
S-CC	92 +- 5 (89-95)	-	25 +- 3 (23-27)	25 +- 3 (23-27)	23 +- 3 (21-25)
C-CC	89 +- 6 (85-93)	-	11 +- 3 ?? (9-13)	12 +- 3 ?? (10-14)	13 +- 2 ?? (12-14)
Values are given as the mean +- SD (95% confidence intervals). - indicates negative value. * P b .05 vs S-CC. ?? P b .01 vs S-CC.

Neurological evaluation

The Cerebral Performance Category neurological func- tion evaluation uses a 5-grade scale [1]. 1 (normal): no difficulty with standing, walking, eating, or drinking, being alert and fully responsive to environmental stimuli; 2 (mild disability): able to stand but exhibiting an Unsteady gait, drinking but not eating normally, with a slower response to environmental stimuli; 3 (severe disability): unable to stand or walk without assistance, incapable of drinking or eating, awake but unable to respond to noxious stimuli; 4 (coma); and 5 (death).

Statistical analysis

Data are reported as mean +- SD. Discrete variables, such as ROSC, 4-hour survival, 24-hour survival, and normal neurological function survival at 24 hours, were compared

Table 3 cerebral metabolism (blood gas, lactate value, Ca-ijvDO2, and O2ER)

with Fisher exact test testing. Continuous variables, includ- ing all hemodynamic data, blood gases, lactates, and oxygen content parameters measured during the CPR period, were compared with unpaired 2-tailed t test. A value of P b .05 was regarded as being statistically significant.

Results

Eighteen pigs (8 female and 10 male) were randomized to the 2 groups. Female/Male was 3:6 in the S-CC group and 5:4 in the C-CC group (P = .637).

The ROSC, 4-hour survival, 24-hour survival, neurologi- cally normal 24-hour survival of the S-CC group were significantly better than those in the experimental group receiving C-CC (P b.05). However, there were no significant differences in the duration of CPR or Shocks before ROSC (Table 1).

Baseline VF 4 min CPR 3 min CPR 6 min CPR 9 min
PaH
S-CC	7.46 +-	.05	7.53 +- .06	7.58 +-	.08	7.51 +-	.02	7.48	+- .02
C-CC	7.45 +-	.04	7.52 +- .06	7.54 +-	.07	7.54 +-	.06	7.51	+- .06
PaCO2 (mm Hg)
S-CC	39 +-	4	33 +- 4	22 +-	6	24 +-	5	25	+- 5
C-CC	41 +-	3	32 +- 3	22 +-	5	23 +-	4	25	+- 4
PaO2 (mm Hg)
S-CC	86 +-	6	58 +- 3	78 +-	2	80 +-	2	81	+- 2
C-CC	87 +-	5	55 +- 5	63 +-	6 ??	66 +-	6 ??	66	+- 6 ??
aLac (mmol/L)
S-CC	1.78 +-	.60	5.62 +- .90	6.42 +-	.79	7.07 +-	.86	7.25	+- .57
C-CC	1.68 +-	.25	6.10 +- 1.34	7.42 +-	.91 ?	8.78 +-	.39 ??	9.49	+- .53 ??
PijvH
S-CC	7.43 +-	.05	7.38 +- .06	7.32 +-	.06	7.28 +-	.08	7.26	+- .09
C-CC	7.45 +-	.03	7.38 +- .05	7.33 +-	.06	7.27 +-	.07	7.24	+- .07
PijvCO2 (mm Hg)
S-CC	46 +-	4	54 +- 5	57 +-	4	59 +-	5	59	+- 5
C-CC	45 +-	5	55 +- 3	59 +-	6	67 +-	6 ?	74	+- 7 ??
PijvO2 (mm Hg)
S-CC	31 +-	4	14 +- 2	18 +-	2	16 +-	2	17	+- 2
C-CC	31 +-	3	12 +- 2	18 +-	2	13 +-	4	12	+- 5 ??
ijvLac (mmol/L)
S-CC	2.75 +-	.43	5.97 +- .66	6.91 +-	.73	7.95 +-	.76	8.68	+- .84
C-CC	2.57 +-	.29	6.35 +- 1.33	7.77 +-	.89 ?	9.28 +-	1.02 ??	9.91	+- 1.01 ??
Hgb (g/L)
S-CC	97 +-	6	97 +- 5	99 +-	4	99 +-	4	100	+- 2
C-CC	98 +-	4	98 +- 4	100 +-	3	99 +-	2	100	+- 2
Ca-ijvDO2 (ml/L)
S-CC	60.34 +-	3.44	101.48 +- 5.29	98.56 +-	4.51	101.63 +-	4.23	104.04	+- 4.35
C-CC	63.27 +-	3.94	101.81 +- 4.90	107.56 +-	4.73 ??	106.66 +-	3.10 ?	110.60	+- 3.05 ??
O2ER(%)
S-CC	47 +-	2	85 +- 2	76 +-	2	77 +-	2	79	+- 2
C-CC	48 +-	2	86 +- 2	84 +-	2 ??	84 +-	1 ??	86	+- 1 ??
Values are given as the mean +- SD. * P b .05 vs S-CC. ?? P b .01 vs S-CC.

Fig. 1 Methylene blue staining extent of brain tissue after CPR.

There were no significant differences in the baseline and VF values of CO, mean aortic pressure (MAP), CPP, SvO2, blood gas or lactate measurements, and Ca-ijvDO2 (Tables 2 and 3) (P N .05). However, CO, MAP, CPP, and SvO2 during the CPR were significantly greater in the group of S-CC compared with C-CC (P b .05) (Table 2).

From Fig. 1, methylene blue coloration extent of cerebral tissue during the CPR was greater with S-CC compared with that of C-CC. The mean CO and MAP during CPR of the animal with S-CC were, respectively,

1.77 L/min and 53 mm Hg, but only 1.33 L/min and

38 mm Hg of the animal with C-CC. In addition, the animal with S-CC also had a better mean CPP during CPR than the animal with C-CC (24 vs 12 mm Hg).

From cerebral metabolism, the S-CC group had lower internal jugular vein lactate measurements values (P b .05), Ca-ijvDO2 (P b .05), and O2ER (P b .01) (Table 3).

The histopathology alterations of pallium, cardiac muscle, kidney, and liver were compared, as shown in Fig. 2. Postresuscitation tissue damage of the C-CC group,

especially brain tissue damage, was more serious than the S-CC group.

Discussions

Because of the limitation of animal provision, we did not choose pigs of the same sex. Some studies documented that hormonal differences due to sex dimorphism may be important in outcome from adverse circulatory conditions [18]. However, the sexual immaturity of the typical swine model may preclude this effect [19]. Although sexual immature pigs (2-3 months) were used in our study and there was no significant difference in sex between 2 groups, sex still may be a potentially confounding variable.

The quality of chest compressions often did not meet published guidelines recommendations, even when per- formed by hospital staff. Abella et al [9] documented that the in-hospital chest compression rates were less than 80 per minute in 37% of the total segments. Wik et al [10] found that only 28% of the out-of-hospital compressions adhered to international guidelines for CPR (between 38 and 51 mm with complete release). This study compared clinical quality chest compressions with standard manual chest compres- sions (the 2005 guidelines recommended) in a porcine VF model, especially with the addition of the histopathologic findings to hemodynamic and survival data. We also firstly infuse methylene blue into the heart to evaluate the cerebral perfusion through observing brain tissue coloration.

It is proven in this study that C-CC induced worse outcomes as compared with S-CC. The end point of neurologically normal 24-hour survival in the S-CC group

Fig. 2 Photographs of histopathologic findings. A, A few neuron cells lost the processus and swelled mildly. B, Most neuron cells lost the processus, swelled heavily, nuclear pycnosis and Nissl bodies lost. A part of the cell membrane is dissolved and neuron cellular necrosis (thin arrow) with the reaction of the neuronophage (thick arrow). C, Few cytoplasms of myocardial cells are deeply stained with nuclear pycnosis (thin arrow). D, Most cytoplasms of myocardial cells are deeply stained with nuclear pycnosis (thin arrow); myocardial interstitial substance are widen and interstitial infiltration of few inflammatory cells (thick arrow). E, Glomeruluses are normal. Renal tubular epithelial cells swelled mildly. F, Glomeruluses are basic normal. Renal tubular epithelial cells swelled, and interstitial infiltration of lots of inflammatory cells, most of which are leukomonocytes (arrow). G, Hepatocytes swelled, sinus hepaticus is broaden with blood congested. H, Hepatocytes swelled, sinus hepaticus is broaden obviously with blood congested(arrow). S-CC: A, C, E, and G; C-CC: B, D, F, and H (H&E x400).

was significantly better than the C-CC group, which also had serious postresuscitation tissue damage, especially brain damage.

Coronary perfusion pressure values during CPR have been identified as the leading predictor of the success of CPR [20,21]. The values of MAP and CPP during CPR in the S- CC group were, respectively, 50% and 26% of baseline levels, whereas only 35% and 13% in the C-CC group. Ristagno et al [16] yielded analogue result by comparing 2 types of compression depth (6.0 vs 4.2 cm). The CCO and SvO2 have been applied extensively in clinic but not always used in the study of CPR. SvO2 is related to oxygen-delivery and oxygen-consumption. In the same oxygen-consumption condition, the lower SvO2 values in the C-CC group demonstrated that systemic oxygen supply was more deficient compared with the S-CC group. Cardiac output during CPR is related to stroke volume and compression rates. It has been documented in animal experiments that increased chest compression depth can improve cardiac output [22]. and incomplete chest wall recoil during the decompression phase of CPR decreases mean arterial pressure and Coronary and cerebral perfusion pressures [23]. Our study supports these earlier findings.

That methylene blue coloration extent of brain tissue was greater in the S-CC group directly reflected the more cerebral perfusion with S-CC than with C-CC. However, we were only able to compare methylene blue data from 1 animal in the S-CC group with 7 animals in the C-CC group because of the different ROSC, and Salaris et al

[24] had documented that methylene blue could decrease the damage of ischemia/reperfusion injury because of inhibition of superoxide generation. Although we used the same dose (40 mg) in each group so that we could ignore the affection, we still hoped that another stain can be used to eliminate the possible interferences.

The increase of Ca-ijvDO2 or O2ER represents that cerebral blood flow is insufficient relative to cerebral oxygen consumption. The study S-CC group with lower Ca-ijvDO2 values (P b .05) and O2ER (P b .01) documented that there might be improved cerebral perfusion compared with C-CC. The differences of lactate levels of internal jugular vein also has proven this.

Compared with S-CC in histopathologic findings, C-CC induced more serious postresuscitation tissue damage, especially brain tissue. As it is show in Fig. 2, although postresuscitation brain damage with S-CC also occurred, it was significantly better than that in the C-CC group in which there was much neuron cell degeneration or necrosis. Therefore, neurologically normal 24-hour survival of the S- CC group (7/9) was significantly better than in the experimental group receiving C-CC (0/9).

It is a systemic ischemia/reperfusion process that begins with cardiac arrest next follow by CPR and then returns to ROSC. Although CPR can produce systemic perfusion after cardiac arrest, perfusion volume was limited. As shown above, the values of MAP and CPP during CPR were just,

respectively, 50% and 26% of baseline levels with S-CC, and they were worse with C-CC. Negovsky et al [25] suggested that the degree of postresuscitation tissue damage was decided to the extent of earlier period systemic ischemia, and earlier period systemic ischemia is related to time of untreated VF, cheat compressions quality, and duration of CPR. In the present study, time of untreated VF was fixed, and there was no significant difference in the duration of CPR between 2 groups. Therefore, postresuscitation tissue damage was finally decided to the chest compressions quality of the 2 groups.

Conclusions

High-quality chest compressions improved outcomes of resuscitation after cardiac arrest. Compared with C-CC, S-CC produced greater ROSC, 4-hour survival, 24-hour survival, and neurologically normal 24-hour survival. The quality of chest compressions also had direct impact on postresuscita- tion tissue damage, especially brain tissue damage.

References

1. Kern KB, Hilwig RW, Berg RA, et al. Importance of Continuous chest compressions during cardiopulmonary resuscitation: improved out- come during a simulated single lay-rescuer scenario. Circulation 2002:645-9.
2. Aufderheide TP, Sigurdsson G, Pirrallo RG, et al. Hyperventilation- induced hypotension during cardiopulmonary resuscitation. Circula- tion 2004;109:1960-5.
3. Van Hoeyweghen RJ, Bossaert LL, Mullie A, et al. Quality and efficiency of bystander CPR. Resuscitation 1993;26:47-52.
4. Wik L, Steen PA, Bircher NG. Quality of bystander cardiopulmonary resuscitation influences outcome after prehospital cardiac arrest. Resuscitation 1994;28:195-203.
5. Gallagher EJ, Lombardi G, Gennis P. Effectiveness of bystander cardiopulmonary resuscitation and survival following out-of-hospital cardiac arrest. Jama 1995;274:1922-5.
6. Ko PC, Chen WJ, Lin CH, et al. Evaluating the quality of prehospital cardiopulmonary resuscitation by reviewing automated external defibrillator records and survival for out-of-hospital witnessed arrests. Resuscitation 2005;64:163-9.
7. 2005 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2005; 112:IV1-IV203.
8. Abella BS, Alvarado JP, Myklebust H, et al. Quality of cardiopulmonary resuscitation during in-hospital cardiac arrest. Jama 2005;293:305-10.
9. Abella BS, Sandbo N, Vassilatos P, et al. Chest compression rates during cardiopulmonary resuscitation are suboptimal: a prospec- tive study during in-hospital cardiac arrest. Circulation 2005;111: 428-34.
10. Wik L, Kramer-Johansen J, Myklebust H, et al. Quality of cardiopulmonary resuscitation during out-of-hospital cardiac arrest. Jama 2005;293:299-304.
11. Aufderheide TP, Pirrallo RG, Yannopoulos D, et al. Incomplete chest wall decompression: a clinical evaluation of CPR performance by EMS personnel and assessment of alternative manual chest compres- sion-decompression techniques. Resuscitation 2005;64:353-62.
12. Handley AJ, Handley SA. Improving CPR performance using an audible feedback system suitable for incorporation into an automated external defibrillator. Resuscitation 2003;57:57-62.
13. Williamson LJ, Larsen PD, Tzeng YC, et al. Effect of automatic external defibrillator audio prompts on cardiopulmonary resuscitation performance. Emerg Med J 2005;22:140-3.
14. Chiang WC, Chen WJ, Chen SY, et al. Better adherence to the guidelines during cardiopulmonary resuscitation through the provision of audio-prompts. Resuscitation 2005;64:297-301.
15. Grund F, Tjomsland O, Sjaastad I, et al. Pentobarbital versus medetomidine-ketamine-fentanyl anaesthesia: effects on haemody- namics and the incidence of ischaemia-induced ventricular fibrillation in swine. Lab Anim 2004;38:70-8.
16. Ristagno G, Tang W, Chang YT, et al. The quality of chest compressions during cardiopulmonary resuscitation overrides impor- tance of timing of defibrillation. Chest 2007;132:70-5.
17. Hamer AW, Karagueuzian HS, Sugi K, et al. Factors related to the induction of ventricular fibrillation in the normal canine heart by programmed electrical stimulation. J Am Coll Cardiol 1984;3: 751-9.
18. Angele MK, Schwacha MG, Ayala A, et al. Effect of gender and sex hormones on immune responses following shock. Shock 2000;14:81-90.
19. Niemann JT, Rosborough JP, Youngquist S. Is the tumour necrosis factor-alpha response following resuscitation gender dependent in the swine model? Resuscitation 2008;77:258-63.
20. Sanders AB, Kern KB, Atlas M, et al. Importance of the duration of inadequate coronary perfusion pressure on resuscitation from cardiac arrest. J Am Coll Cardiol 1985;6:113-8.
21. Paradis NA, Martin GB, Rivers EP, et al. Coronary perfusion pressure and the return of spontaneous circulation in human cardiopulmonary resuscitation. Jama 1990;263:1106-13.
22. Bellamy RF, DeGuzman LR, Pedersen DC. Coronary blood flow during cardiopulmonary resuscitation in swine. Circulation 1984;69: 174-80.
23. Yannopoulos D, McKnite S, Aufderheide TP, et al. Effects of incomplete chest wall decompression during cardiopulmonary resus- citation on coronary and cerebral perfusion pressures in a porcine model of cardiac arrest. Resuscitation 2005;64:363-72.
24. Salaris SC, Babbs CF, Voorhees III WD. Methylene blue as an inhibitor of superoxide generation by xanthine oxidase. A potential new drug for the attenuation of ischemia/reperfusion injury. Biochem Pharmacol 1991;42:499-506.
25. Negovsky VA. The second step in resuscitation-the treatment of the

‘post-resuscitation disease’. Resuscitation 1972;1:1-7.