Resuscitation 84 (2013) Contents lists available at SciVerse ScienceDirect. Resuscitation

Transcription

1 Resuscitation 84 (2013) Contents lists available at SciVerse ScienceDirect Resuscitation jo u rn al hom epage : Simulation and education The quality of cardiopulmonary resuscitation using supraglottic airways and intraosseous devices: A simulation trial Dena A. Reiter, Christopher G. Strother, Scott D. Weingart Department of Emergency Medicine, Mount Sinai School of Medicine, New York, NY, United States a r t i c l e i n f o Article history: Received 10 November 2011 Received in revised form 28 June 2012 Accepted 5 July 2012 Keywords: Cardiac arrest Critical care Airway management Simulation a b s t r a c t Study objective: To assess whether using interventions such as laryngeal mask airways (LMA) and IO lines lead to improved resuscitation in a simulated cardiac arrest when compared to standard methods of endotracheal intubation (ETI) and central line placement. Methods: Emergency Medicine residents at a single academic center were grouped into teams of four. Each team participated in two simulated ventricular fibrillation cardiac arrests using a high fidelity simulator. Peripheral IV access was unobtainable. Only ETI supplies and a central line kit were available in one case (control) and in the other case those supplies were replaced by an LMA and an EZ-IO drill kit (experimental). Groups were randomized to which set up they were given first. Data examined included time to airway placement, duration and success rate of airway placement, time to vascular access, time to defibrillation, and percent hands off time. Results: 44 residents in 11 teams participated. Mean time to airway was shorter in the experimental group (122.8 seconds (s) vs s, p = 0.001). Mean duration of airway attempt was also shorter (7.6 s vs s, p = 0.002). Time to access was shorter in the experimental group (49.0 s vs s, p = <0.001). Time to defibrillation and percent hands off time did not significantly differ between the two groups. Conclusion: Use of an LMA and an IO device led to significantly faster establishment of an airway and vascular access in a simulated cardiac arrest. The variation in devices did not affect time to defibrillation or percent hands off time Elsevier Ireland Ltd. All rights reserved. 1. Introduction The American Heart Association (AHA) has recently updated recommendations for Advanced Cardiac Life Support (ACLS) in the 2010 guidelines. 1 Based on literature demonstrating improved neurologic outcomes with minimally interrupted chest compressions during CPR, new emphasis has been placed on compressions as the most important component of ACLS, which has lead to a change in the mnemonic A B C to C A B for cardiac arrest patients. 2 5 Additional updates include the use of the intraosseous device as an alternative access device and placement of laryngeal mask airway as a substitute for endotracheal intubation. Endotracheal intubation has a high incidence of complications including aspiration, esophageal intubation, hemodynamic alterations, and failure to oxygenate and ventilate, the consequences of which cannot be underestimated. 6 9 In addition, endotracheal A Spanish translated version of the abstract of this article appears as Appendix in the final online version at Corresponding author at: One Gustave L. Levy Place, Box 1620, New York, NY 10029, United States. Tel.: ; fax: address: dena.asaad@gmail.com (D.A. Reiter). intubation requires technical skill and this procedure can prove difficult to perform in a patient undergoing high quality continuous chest compressions. As a result, health care providers may ask for short pauses in CPR to aide in establishing airway. It has been demonstrated in the literature that pauses in chest compressions can adversely affect coronary perfusion pressure and ultimately lead to worse neurologic outcome. 3,4 Supraglottic airway devices offer the advantage of requiring less technical skill to place, and are more quickly and easily placed in a cardiac arrest scenario They are considered to be effective and safe and ventilation may be superior when compared to standard bag mask ventilation. 14,15 End tidal CO 2 monitoring which has great utility in the intra-arrest period may also be performed using these devices. 16,17 Intravenous access is required during cardiac arrest not only for administration of indicated medications (pressors and antiarrhythmics) but also for the treatment of reversible causes of cardiac arrest (electrolyte imbalance, hypoglycemia, massive pulmonary embolism, hypovolemia). Peripheral intravenous access is often very difficult to establish in pulseless victims of cardiac arrest. High failure rates and significant delays have been reported for traditional peripheral IV access Placement of central lines requires technical skill and may be difficult in a patient actively undergoing chest compressions. Pauses in chest compressions, /$ see front matter 2012 Elsevier Ireland Ltd. All rights reserved.

2 94 D.A. Reiter et al. / Resuscitation 84 (2013) which are detrimental, may be required in order to safely place these lines. Complication rates of central venous cannulation have been estimated to be between 10 and 15%. 21,22 Risks involved include pnemothorax, vessel laceration, arterial cannulation and health care provider needlestick. Intraosseous devices are becoming widely accepted as an alternative to both peripheral and central IV access. They are safely, easily and rapidly placed and do not require any pauses in chest compressions They minimize risk to both the patient and the healthcare provider. Medication and IV fluids are safely administered via intraosseous lines with similar efficacy We hypothesized that in a simulated adult cardiac arrest, the use of a laryngeal mask airway and intraosseous device in the place of an endotracheal tube and central line would lead to improved quality of cardiopulmonary resuscitation secondary to faster airway establishment, vascular access, earlier defibrillation, and an overall reduction in percent hands-off time. To our knowledge, the simultaneous use of both supraglottic airways and intraossous devices in place of more traditional endotracheal intubation and central line placement for cardiac arrest has not been studied in previous literature. 2. Methods We conducted a randomized crossover study between the months of September 2010 and November This study was approved by the Institutional Review Board of the Mount Sinai School of Medicine with waiver of written consent. This study took place at an urban academic medical center with a 4 year Emergency Medicine residency program. Participants consisted of Emergency Medicine residents at the Mount Sinai School of Medicine, levels PGY1-4. All participants had been previously certified in ACLS within the past two years. These volunteer participants were verbally consented to participate in two simulation trials and consented to have all data recorded via the high fidelity simulator as well as audio/video recording. Eleven groups were formed by having residents sign up for preferred time slots, Each group consisted of a PGY1, PGY2, PGY 3 and PGY 4. Groups were then assigned to either A or B in alternating order, starting with A. The A groups underwent a simulated case of cardiac arrest with the control scenario, then underwent a second case of cardiac arrest with the intervention scenario. The B groups, in a crossover fashion, underwent a simulated case of cardiac arrest using the intervention scenario first, and the control scenario second. They were given an approximately 10 min break between cases and were not debriefed until both cases were completed. Prior to the simulation trials, groups were not informed of the purpose of the study or the content of the scenarios. Team members were not assigned roles by the investigators. Participants were allowed to establish roles at their own discretion during each scenario, as they would in real emergency department resuscitation. The control scenario consisted of a 55 year old male who presented to an emergency department by EMS with crushing chest pain. EKG revealed an anterior wall ST elevation MI. Peripheral IV access was unable to be obtained in this patient. This patient subsequently underwent a ventricular fibrillation arrest at approximately 5 min. Participants were required to perform ACLS and were required to establish an airway and vascular access. In the control scenario, participants were only given the following equipment: nasal cannula, non-rebreather mask, bag valve mask, mac 4 laryngoscope, 8.0 endotracheal tube, triple lumen catheter central line kit, and a monophasic defibrillator. Participants were able to intubate the manikin using the same materials and technique one would use in a real clinical scenario. Central line insertion was simulated using a penrose drain filled with water. The penrose drain was placed over the distribution of the femoral and sublcavian vein and participants were able to perform a central line insertion using the Seldinger technique as one would use on a human subject. The experimental scenario consisted of a 40 year old male who presented to an emergency department by EMS with palpitations and dizziness. EKG revealed frequent PVCs. Peripheral IV access was unable to be obtained in this patient. The patient subsequently underwent a ventricular fibrillation arrest at approximately 5 min. Participants were required to perform ACLS and were required to establish an airway and vascular access. In this experimental scenario, participants were again given a nasal cannula, non-rebreather mask, bag valve mask, and a monophasic defibrillator. However, the laryngoscope, endotracheal tube, and central line kit were removed and replaced with a laryngeal mask airway and an powered-drill intraosseus access device (EZ-IO, Vidacare). Participants were able to insert the laryngeal mask airway into the manikin in the same manner as one would insert the LMA into a human subject. Additionally the manikin had an EZ-IO tibia bone simulator attached to its lower leg, and participants were able to drill intraosseous access devices into the simulated tibia. Time (s) to airway establishment was recorded from time zero (defined as onset of v-fib arrest) to when ventilations were detectable on the manikin by an observer in the room (DAR), and confirmed by video recording. Duration (s) of airway attempt was defined by the time the observer noted the airway device entering the mouth to the time ventilations were detected by the manikin. Number of airway attempts included all endotracheal attempts resulting in esophageal intubation, unsuccessful attempts to visualize the airway during endotracheal intubation, successful endotracheal intubation confirmed by manikin detected ventilations, and both improperly and properly inserted supraglottic devices (as confirmed by manikin detected ventilations). Time (s) to definitive vascular access was also recorded by the observer starting at time zero to when the triple lumen catheter was threaded into a simulated vein, or the IO needle was drilled completely into an adult sized tibia practice model. The time (s) of defibrillation from time zero, and percent hands off time was recorded by the manikin. All recorded data was cross referenced and confirmed by the same observer using the audio and video recording. If the results between real time recording and video observation did not agree, the times were averaged. The observer was not blinded to the hypothesis of the study. Each group s performance was compared between case 1 (control scenario) and case 2 (experimental scenario). Separate analyses for each experimental variable time to airway establishment, duration of airway attempts, number of airway attempts, time to vascular access, time of defibrillation, and percent hands off time were performed using IBM SPSS Statistics 19. Data were analyzed using a within-group paired sample comparison of means for all variables except number of airway attempts, which was not normally distributed and thus analyzed with a Wilcoxson signed rank test. In light of the multiple separate analyses, a Bonferroni corrected alpha cutoff of was chosen for all tests. A further set of analyses were performed using SPSS to examine the effect of order of presentation, in which different scores for each variable were calculated and compared between the group that received the control scenario first versus the group that received the experimental scenario first. 3. Results Forty-four out of 48 total residents participated in the study. Five groups participated in the experimental scenario first and the control scenario second. Six groups participated in the control scenario

3 D.A. Reiter et al. / Resuscitation 84 (2013) Table 1 Mean and median outcomes in the control versus experimental scenario, with standard deviations (SD), p values, and confidence intervals of the difference. Outcomes Control (SD), n = 11 Experimental (SD), n = 11 p value Confidence interval of the difference Mean time to airway s (83.8) s (67.7) s Mean duration of airway attempt 22.7 s (11.0) 7.6 s (3.0) s Median number of airway attempts Mean time to vascular access s (49.7) 49.0 s (22.2) < s Mean time to defibrillation 96.6 s (85.1) 73.3 s (39.4) s Mean percent hands off time 18.1% (9.8) 17.0% (8.6) % first and the experimental scenario second. PGY levels were evenly distributed among groups. Each group consisted of a PGY1, PGY2, PGY3 and PGY4 resident. Roles were determined at the discretion of the appointed team leaders and did not remain consistent between cases or across groups. In the within-group analysis, the mean time to airway was significantly shorter during the experimental scenario as compared to the control scenario (122.8 s (s) vs s, p = 0.001). Mean duration of airway attempt was also shorter (7.6 s vs s, p = 0.002). The median number of airway attempts did not differ between groups, each of which succeeded in establishing the airway after either one or two attempts; a trend toward fewer attempts in the experimental group did not reach statistical significance (p = 0.059). Time to access was shorter in the experimental scenario (49.0 s vs s, p = <0.001). Time to defibrillation and percent hands off time did not significantly differ between the two scenarios (Table 1). The order in which the scenarios were performed did not significantly change outcomes between groups. (p > in all comparisons) (Table 2). 4. Discussion In this study we demonstrated significant advantages using simplified interventions such as laryngeal mask airway and intraosseous device in place of endotracheal tube and central line during a simulated cardiac arrest. The use of these simplified interventions was associated with significantly faster establishment of airway and vascular access. Several studies support the use of supraglottic and intraosseous devices for resuscitation. The use of supraglottic devices as compared to endotracheal intubation and bag valve mask ventilation was studied by Wiese et al. 29 One hundred fifty paramedics working in teams of 2 were randomized to use laryngeal tube suction-device (LTS-D) versus endotracheal tube versus standard BVM in a simulated cardiac arrest in a manikin. The LTS-D was inserted significantly faster than the ET tube (15 s versus 44 s, p < 0.01) and ventilations were established significantly faster (57 s with LTS-D versus 116 s with ET versus 111 s with BVM). The noflow time was significantly reduced in the LTS-D group (125 s with LTS-D versus 207 s with ET versus 160 s with BVM, p < 0.01) Laryngeal mask airway as compared to endotracheal tube was also studied by Chen et al. in a simulated prehospital pediatric cardiac arrest. 11 Fifty-two emergency medical technicians were included in a randomized crossover study using laryngeal mask airway or endotracheal tube in a simulated pediatric arrest using a manikin. On average, the LMA was placed 50% faster than the endotracheal tube (23 s vs. 46 s). The mean difference was 23 s (95% CI s). LMA was placed fewer attempts compared to ETI (1.1 vs. 1.27, p = 0.03). The LMA had a lower incidence of misplacement as well (9.5% vs. 34%, p < 0.001) Ruetzler et al. conducted a manikin study which compared six different airway devices and their effect on hands-off time during cardiopulmonary resuscitation. 30 Emergency medical technicians were asked to perform airway management on a simulator using an endotracheal tube, Combitube, EasyTube, laryngeal tube, Laryngeal Mask Airway (LMA), and I-Gel in a randomized sequence during ongoing chest compressions. The hands-off time associated with endotracheal intubation was significantly longer (48 s, 95% CI s) compared with any of the other airway systems (Combitube 10 s, 95% CI s, EasyTube 11.4 s, 95% CI s, laryngeal tube 8.4 s, 95% CI s, LMA 13.3 s, 95% CI s, I-Gel 15.9 s, 95% CI s). There were no significant differences among the airway systems outside of the endotracheal tube. The use of IO device versus central venous catheter for vascular access in a population of ED patients undergoing medical or trauma resuscitation was examined in a prospective observational study by Leidel et al. 31 Forty consecutive adult patients undergoing resuscitation received both intraosseous and central venous access. Procedural time and success rate were recorded. Mean procedure time was lower for IO cannulation (2.0 min vs. 8.0 min, p < 0.001). First time success rate was higher for IO cannulation (85% versus 60%, p = 0.024) Reades et al. conducted a prehospital randomized control trial comparing tibial intraosseous access, humeral interosseous access, and peripheral intravenous access. 32 The primary outcome was first attempt success rate. Individuals randomized to tibial intraosseous access were more likely to experience a successful first attempt at vascular access (91%, 95% CI 83 98%) compared with either humeral intraosseous access (51%, 95% CI 37 65%) or peripheral intravenous access (43%, 95% CI 31 55%) groups. Time to initial success was shortest in the tibial interosseous group (4.6 min; interquartile range min) compared to the humeral interosseous access group (7.0 min, interquartile range min) and peripheral intravenous access group (5.8 min, interquartile range min). However the difference between Table 2 Order analysis comparing outcomes between groups performing case #1 first versus case #2 first. Mean change in value from C1 to C2 Order 1 2 (SD), n = 6 Order 2 1 (SD), n = 5 p value Confidence interval of the difference Time to airway s (37.79) s (133.58) s Duration of airway attempt s (5.68) 14.6 s (18.10) Number of airway attempts 0.50 (0.55) 0.40 (0.89) Time to vascular access s (63.50) s (25.24) Time to defibrillation s (61.19) s (46.31) Percent hands off time 0.55% (6.09) 1.78% (8.67) %

4 96 D.A. Reiter et al. / Resuscitation 84 (2013) interosseous and peripheral intravenous access groups was not significant. Because we found time to airway establishment and vascular access to be significantly faster using supraglottic and intraosseous devices, our study supports the findings in previous resuscitation literature. We did not find any differences in time to defibrillation and percent hands off time between the two groups. This may have occurred secondary to having four providers participating in the resuscitation. Had fewer people been available to assist with chest compressions and defibrillation, these outcomes may have been more affected by the airway and vascular access interventions. Additionally, early defibrillation and minimized hands off time is reinforced during both ACLS training and real life resuscitation in both of our emergency departments. Participants in this trial may have been more mindful of defibrillating early and minimizing hands off time, and our results may not reflect the performance of other healthcare providers. A study by Jiang examining whether real time video feedback led to improved resuscitation reported hands off times ranging between 24% and 34%. 33 Wik et al. examined the quality of CPR by ambulance personnel and found hands off time to be 48% (38% when adjusted for defibrillation and pulse check). 34 Olasveengen et al. analyzed CPR quality over a 3 year period and observed a hands off time of 18%, improving to 11%. 35 This was attributed to developing evidence and guidelines focusing on minimizing chest compression interruptions. Our hands off time of 17% and 18.1% fall within the lower end of the range of previously reported hands off times. 5. Limitations Our study had several limitations. Sample size was chosen based on the size of the residency program, and not on a power calculation. Because of the nature of the study, subjects were unblinded to the interventions which may have led to biased performance in resuscitations. However, subjects were not aware of the study hypothesis. Additionally groups may have performed at a higher level based on the fact that they were participating in this study (Hawthorne effect). Each group underwent both the control and experimental scenario on the same day. Groups were randomized as to which scenario they experienced first, and the order in which the scenarios took place did not lead to any significant differences in outcomes. However, we could not assess whether groups performed better in the second scenario because they had recently performed ACLS in the first scenario. There are several characteristics of our high fidelity Laerdal simulator which may not make our data generalizeable to all emergency department patients. Our simulator only provided a very specific airway anatomy which may not reflect the potentially varying anatomies of human subjects. In addition our simulator did not include potential complicating factors such as secretions, vomiting or difficult anatomy. We are unable to assess whether endotracheal intubation and insertion of laryngeal mask airway is more or less difficult when compared to true clinical scenarios. Additionally, the simulated insertion of intraosseous and central line may not have reflected the level of difficulty one would find in a true cardiac arrest. We felt that the models that were in place may have been easier in the simulated model when compared to a true ED patient. This is particularly true for the central line model, in which the central line was inserted into a visible penrose drain. This would be considerably harder if done blindly or with ultrasound guidance as it is done in true cardiac arrest scenario. This limitation would only favor the placement of intraosseous device and support our hypothesis. Lastly, in the management of cardiac arrest, the most important patient oriented outcome is neurologically meaningful survival. Given that this was a simulation trial, this outcome was not available for analysis. In conclusion, our study data lends support to the use of simultaneous supraglottic and intraosseous devices in place of traditional endotracheal intubation and central line placement in a simulated cardiac arrest scenario. In a simulated cardiac arrest, the use of these devices was associated with significantly faster airway establishment and vascular access. The use of these devices did not significantly improve time to defibrillation or hands off time. Additional studies are needed to further assess the use of such interventions in human subjects. Conflict of interest statement The authors have no conflicts of interest or financial disclosures to report. Acknowledgement The authors thank Tavis Allison for conducting the statistical analysis of this study. References 1. Field JM, Hazinski MF, Sayre MR, et al. Part 1: Executive summary: 2010 American heart association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2010;122(Suppl. 3):S Hallstrom A, Cobb L, Johnson E, et al. Cardiopulmonary resuscitation by chest compression alone or with mouth-to-mouth ventilation. N Engl J Med 2000;342: Berg RA, Sanders AB, Kern KB, et al. Adverse hemodynamic effects of interrupting chest compressions for rescue breathing during cardiopulmonary resuscitation for ventricular fibrillation cardiac arrest. Circulation 2001;104: Christenson J, Andrusiek D, Everson-Stewart S, et al. Chest compression fraction determines survival in patients with out-of-hospital ventricular fibrillation. Circulation 2009;120: Bobrow BJ, Clark LL, Ewy GA, et al. Minimally interrupted cardiac resuscitation by emergency medical services for out-of-hospital cardiac arrest. JAMA 2008;299: Jaber S, Amraoui J, Lefrant J, et al. Clinical practice and risk factors for immediate complications of endotracheal intubation in the intensive care unit: a prospective, multiple-center study. Crit Care Med 2006;34: Mort TC. Complications of emergency tracheal intubation: hemodynamic alterations: part I. J Intensive Care Med 2007;22: Mort TC. Complications of emergency tracheal intubation: immediate airwayrelated consequences: part II. J Intensive Care Med 2007;22: Schwartz DE, Mathay MA, Cohen NH. Death and other complications of emergency airway management in critically ill adults. A prospective investigation of 297 tracheal intubations. Anesthesiology 1995;82: Wiese CH, Semmel T, Muller JU, et al. The use of the laryngeal tube disposable (LT- D) by paramedics during out-of-hospital resuscitation-an observational study concerning ERC guidelines Resuscitation 2009;80: Chen L, Hsiao AL. Randomized trial of endotracheal tube versus laryngeal mask airway in simulated prehospital pediatric arrest. Pediatrics 2008;122:e Gatward JJ, Thomas MJ, Nolan JP, et al. Effect of chest compressions on the time taken to insert airway devices in a manikin. Br J Anaesth 2008;100: Kokkinis K. The use of the laryngeal mask airway in CPR. Resuscitation 1994;27: Heuer JF, Barwing J, Eich C, et al. Initial ventilation through laryngeal tube instead of face mask in out-of-hospital cardiopulmonary arrest is effective and safe. Eur J Emerg Med 2010;17: Kurola J, Harve H, Kettunen T, et al. Airway management in cardiac arrest comparison of the laryngeal tube, tracheal intubation and bag-valve mask ventilation in emergency medical training. Resuscitation 2004;61: Eckstein M, Hatch L, Malleck J, et al. End-tidal CO 2 as a predictor of survival in out-of-hospital cardiac arrest. Prehosp Disaster Med 2011;26: Levine RL, Wayne MA, Miller CC. End-tidal carbon dioxide and outcome of outof-hospital cardiac arrest. N Engl J Med 1998;337: Lapostolle F, Catineau J, Carrigue B, et al. Prospective evaluation of peripheral venous access difficulty in emergency care. Intensive Care Med 2007;33: Minville V, Pianezza A, Asehnoune K, et al. Prehospital intravenous line placement assessment in the French emergency system: a prospective study. Eur J Anaesthesiol 2006;23: Costantino TG, Parikh AK, Satz WA, et al. Ultrasonography-guided peripheral intravenous access versus traditional approaches in patients with difficult intravenous access. Ann Emerg Med 2005;46:

5 D.A. Reiter et al. / Resuscitation 84 (2013) Steele R, Irvin CB. Central line mechanical complication rate in emergency medicine patients. Acad Emerg Med 2001;8: Taylor RW, Palagiri AV. Central venous catheterization. Crit Care Med 2007;35: Levitan RM, Bortle CD, Snyder TA, et al. Use of a battery-operated needle driver for intraosseous access by novice users: skill acquisition with cadavers. Ann Emerg Med 2009;54: Ngo AS, Oh JJ, Chen Y, et al. Intraosseous vascular access in adults using the EZ-IO in an emergency department. Int J Emerg Med 2009;2: Wampler D, Schwartz D, Shumaker J, et al. Paramedics successfully perform humeral EZ-IO intraosseous access in adult out-of-hospital cardiac arrest patients. Am J Emerg Med 2011 [Epub ahead of print]. 26. Spivey WH, Lathers CM, Malone DR, et al. Comparison of intraosseous, central, and peripheral routes of sodium bicarbonate administration during CPR in pigs. Ann Emerg Med 1985;14: Von Hoff DD, Kuhn JG, Burris 3rd HA, et al. Does intraosseous equal intravenous? A pharmacokinetic study. Am J Emerg Med 2008;26: Buck ML, Wiggins BS, Sesler JM. Intraosseous drug administration in children and adults during cardiopulmonary resuscitation. Ann Pharmacother 2007;41: Wiese CH, Bahr J, Bergmann A, et al. Reduction in no flow time using a laryngeal tube: comparison to bag-mask ventilation. Anaesthesist 2008;57: Ruetzler K, Gruber C, Nabecker S, et al. Hands-off time during insertion of six airway devices during cardiopulmonary resuscitation: a randomised manikin trial. Resuscitation 2011;82: [Epub 2011, April 6]. 31. Leidel BA, Kirchhoff C, Bogner V, et al. Comparison of intraosseous versus central venous vascular access in adults under resuscitation in the emergency department with inaccessible peripheral veins. Resuscitation 2012;83:40 5 [Epub 2011, September 3]. 32. Reades R, Studnek JR, Vandeventer S, et al. Intraosseous versus intravenous vascular acces during out-of-hospital cardiac arrest: a randomized controlled trial. Ann Emerg Med 2011;58: Jiang C, Zhao Y, Chen Z, et al. Improving cardiopulmonary resuscitation in the emergency department by real-time video recording and regular feedback learning. Resuscitation 2010;81: Wik L, Kramer-Johansen J, Myklebust H, et al. Quality of cardiopulmonary resuscitation during out-of-hospital cardiac arrest. JAMA 2005;293: Olasveengen TM, Wik L, Kramer-Johansen J, et al. Is CPR quality improving? A retrospective study of out-of-hospital cardiac arrest. Resuscitation 2007;75: