UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Emergency medical services design and response : the role of chain-of-survival clinical interventions… Andrusiek, Douglas Lorne 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_november_andrusiek_douglas.pdf [ 2.6MB ]
Metadata
JSON: 24-1.0372327.json
JSON-LD: 24-1.0372327-ld.json
RDF/XML (Pretty): 24-1.0372327-rdf.xml
RDF/JSON: 24-1.0372327-rdf.json
Turtle: 24-1.0372327-turtle.txt
N-Triples: 24-1.0372327-rdf-ntriples.txt
Original Record: 24-1.0372327-source.json
Full Text
24-1.0372327-fulltext.txt
Citation
24-1.0372327.ris

Full Text

Emergency Medical Services Design and Response: The Role of Chain-of-Survival Clinical Interventions on Survival Following Out-of-Hospital Cardiac Arrest by  Douglas Lorne Andrusiek  B.A., The University of Western Ontario, 1987 M.Sc., The University of British Columbia, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Population and Public Health)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  SEPTEMBER 2018  © Douglas Lorne Andrusiek, 2018 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Emergency Medical Services Design and Response: The Role of Chain-of-Survival Clinical Interventions on Survival Following Out-of-Hospital Cardiac Arrest  submitted by Douglas Lorne Andrusiek  in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Population and Public Health Examining Committee: KS Joseph, Population and Public Health Supervisor  Sam Sheps, Population and Public Health Supervisory Committee Member  John M Tallon, Emergency Medicine Supervisory Committee Member   Timothy C.Y. Chan, Industrial Engineering External Examiner Boris Sobolev, Population and Public Health University Examiner Rollin Brant, Department of Statistics University Examiner Additional Supervisory Committee Members: Riyad B Abu-Laban, Emergency Medicine Supervisory Committee Member iii  Abstract Out-of-hospital cardiac arrest (OHCA) resuscitation is guided by a set of published guidelines that includes several interventions based on the organizing framework known as the chain-of-survival. The aim of the studies in this dissertation was to examine the effects of these interventions in improving survival following non-traumatic adult OHCA. Data for this purpose was obtained from two multi-centre studies carried out in the United States and Canada. Analysis methods included multivariable logistic regression and matched propensity score analysis.  The first study was an analysis of a cohort of 12,821 OHCA cases that showed that younger age, public location, shorter response time, initial rhythm of ventricular fibrillation/tachycardia, and advanced airway management were associated with improved survival, while epinephrine administration was associated with a reduction in survival. Specialized post-arrest care appeared to be associated with improved survival but the magnitude of this was attenuated in analyses of cases transported to hospital. There was a non-significant association between advanced airway management and reduced survival (adjusted odds ratio (AOR) 0.82, 95% confidence interval (CI) 0.59–1.14) and a substantial reduction in survival among those who received epinephrine (AOR 0.13, 95% CI 0.10–0.17).  The second study was based on a cohort of 14,673 non-traumatic OHCA cases and showed that initial rhythm modified the effect of prehospital epinephrine administration: asystole was associated with a 14-fold decrease in the adjusted odds of survival if epinephrine was administered, while ventricular fibrillation/tachycardia was associated with a 5.8-fold decrease in the adjusted odds of survival if epinephrine was administered (p-value for interaction < 0.01). iv   The third study used propensity score matching and showed that epinephrine was negatively associated with return of spontaneous circulation (ROSC; AOR 0.50, 95% CI 0.42–0.60) and survival (AOR 0.15, 95% CI 0.11–0.19), while tracheal intubation was positively associated with ROSC (AOR 2.31, 95% CI 1.92–2.77) and survival (AOR 1.97, 95% CI 1.62–2.40).   These findings raise concerns regarding the efficacy of some commonly used interventions in OHCA. Some interventions, such as epinephrine, may potentially compromise survival at least in some subgroups, and should to be more rigorously studied through randomized controlled trials.   v  Lay Summary The chain-of-survival is the central metaphor for the organization and evaluation of EMS efforts to resuscitate victims of non-traumatic cardiac arrest. In this dissertation, I sought to measure the effect of epinephrine and advanced airway placement including intubation on return of spontaneous circulation and survival for adults who suffered a non-traumatic out-of-hospital cardiac arrest. We first confirmed the effects of many of the factors known to improve survival, such as younger age, public location, faster EMS response time, and type of first known heart rhythm. However, we found that some elements of the chain-of-survival could not be assessed due to survival bias or a mixing of effects due to the conditions under which the care was provided. We attempted to address the mixing of effects using different datasets, and statistical methods that allow for very close matching of cases based on recorded characteristics. In these analyses as well, we found that epinephrine was associated with reduced survival, while advanced airway placement, including tracheal intubation, improved survival. Our findings point to the need for rigorous studies, using a randomized controlled trial design to definitively assess the risks and benefits of epinephrine use in out-of-hospital cardiac arrest.       vi  Preface This statement certifies that the work presented in this dissertation was conceived, conducted, and written by Douglas Andrusiek. All studies were conducted after receipt of ethics approval from the University of British Columbia Behavioural Research Ethics Board (Certificate No.: H08-02293).   Chapters 2, 3 and 4 of this dissertation were conceived and written as stand-alone manuscripts for submission to peer-reviewed academic journals for publication. The Resuscitation Outcomes Consortium (ROC) Publications Committee reviewed these three chapters and approval was granted for publication in peer-reviewed journals.     Data used in this dissertation were collected by the ROC under two study protocols: the Epistry-Cardiac protocol, and the ROC PRIMED randomized clinical trial. Epistry-Cardiac data, including de-identified cases and EMS agency information, were provided directly by the ROC.  ROC PRIMED de-identified case information was acquired through a Research Materials Distribution Agreement (RMDA) from the National Heart, Lung, and Blood Institute Biological Specimen and Data Repository BioLINCC data access program (https://biolincc.nhlbi.nih.gov /about/). The ROC PRIMED case information was augmented with additional data that allowed for the assembly of the combined ROC PRIMED cohort with ancillary EMS agency information.  As a condition of use of data provided by the ROC for off-site data analysis, a ROC faculty member/statistician was assigned to supervise and support all research activities.    vii  As Manager of Research at British Columbia Ambulance Service, I oversaw and coordinated collection of clinical prehospital care records for cases enrolled by the British Columbia site of the Resuscitation Outcomes Consortium for Epistry-Cardiac. As Director of Research and Evaluation at British Columbia Emergency Health Service (BCEHS), I oversaw and coordinated collection of clinical prehospital care records for cases enrolled in PRIMED through the British Columbia site of the Resuscitation Outcomes Consortium PRIMED clinical trials.  I conducted all analyses and wrote all of the statistical analysis code. I wrote the first draft of all manuscripts. My supervisor (K.S. Joseph) and thesis committee members (Riyad Abu-Laban, Sam Sheps and John Tallon) made contributions to the study design, analysis and interpretation of data and reviewed each of the 3 manuscripts for intellectual content, provided editorial input and signed off on the final versions for journal submission. My contribution was greater than 90% for each manuscript of this dissertation.  Chapter 2: Effects of the chain-of-survival interventions and other factors on survival following non-traumatic out-of-hospital cardiac arrest. A version of Chapter 2 is under review at a peer-reviewed journal. I conceptualized and designed the study, managed all data acquisition, performed the data manipulation and analysis, and wrote the first draft of the manuscript. The author list includes my committee members (Andrusiek DL, Abu-Laban RB, Sheps S, Tallon JM, Joseph KS).   Chapter 3: Survival following non-traumatic out-of-hospital cardiac arrest: Do prehospital clinical interventions modify the prognostic effect of initial rhythm? viii  A version of Chapter 3 is under review at a peer-reviewed journal. I conceptualized and designed the study, managed all data acquisition, performed the data manipulation and analysis, and wrote the first draft manuscript. The author list includes my committee members (Andrusiek DL, Abu-Laban RB, Sheps S, Tallon JM, Joseph KS).   Chapter 4: Effect of epinephrine and tracheal intubation in out-of-hospital cardiac arrest: A propensity score analysis. A version of Chapter 4 is under review at a peer-reviewed journal. I conceptualized and designed the study, managed all data acquisition, performed the data manipulation and analysis, and created the first draft manuscript. The author list includes my committee members (Andrusiek DL, Abu-Laban RB, Sheps S, Tallon JM, Joseph KS).   ix  Table of Contents ABSTRACT .............................................................................................................................................................iii LAY SUMMARY .......................................................................................................................................................v PREFACE ................................................................................................................................................................ vi TABLE OF CONTENTS ............................................................................................................................................. ix LIST OF TABLES ..................................................................................................................................................... xii LIST OF FIGURES .................................................................................................................................................. xvi LIST OF ABBREVIATIONS ..................................................................................................................................... xvii ACKNOWLEDGEMENTS........................................................................................................................................ xix DEDICATION ....................................................................................................................................................... xxii CHAPTER 1: INTRODUCTION .......................................................................................................................................... 1 1.1 Tracheal intubation and airway management in out-of-hospital cardiac arrest ................................ 2 1.2 Epinephrine administration in out-of-hospital cardiac arrest resuscitation ....................................... 6 1.3 Summary of background literature .................................................................................................11 1.4 Study objectives .............................................................................................................................13 1.5 Methods summary .........................................................................................................................13 CHAPTER 2: EFFECTS OF THE CHAIN-OF-SURVIVAL INTERVENTIONS AND OTHER FACTORS ON SURVIVAL FOLLOWING NON-TRAUMATIC OUT OF HOSPITAL CARDIAC ARREST .......................................................................................................................................16 2.1 Summary ...............................................................................................................................................16 2.2 Introduction ...........................................................................................................................................17 2.3 Methods ................................................................................................................................................18 2.3.1 Population and setting .....................................................................................................................................19 2.3.2 Determinants, outcomes, and covariates .........................................................................................................19 x  2.3.3 Outcomes........................................................................................................................................................20 2.3.4 Analytic methods.............................................................................................................................................20 2.3.5 Ethics approval ................................................................................................................................................21 2.3.6 Results ................................................................................................................................................21 2.4 Discussion ..............................................................................................................................................26 2.5 Conclusion..............................................................................................................................................30 CHAPTER 3: SURVIVAL FOLLOWING NON-TRAUMATIC OUT-OF-HOSPITAL CARDIAC ARREST: DO PREHOSPITAL CLINICAL INTERVENTIONS MODIFY THE PROGNOSTIC EFFECT OF INITIAL RHYTHM? ........................................................................................................62 3.1 Summary ...............................................................................................................................................62 3.2 Introduction ...........................................................................................................................................63 3.3 Methods ................................................................................................................................................65 3.3.1 Population.......................................................................................................................................................65 3.3.2 Determinants and outcomes ............................................................................................................................65 3.3.3 Analytic methods.............................................................................................................................................66 3.4 Results ...................................................................................................................................................67 3.5 Discussion ..............................................................................................................................................71 3.6 Conclusion..............................................................................................................................................75 CHAPTER 4: EFFECT OF EPINEPHRINE AND TRACHEAL INTUBATION IN OUT-OF-HOSPITAL CARDIAC ARREST: A PROPENSITY SCORE ANALYSIS ........................................................................................................................................................................... 100 4.1 Summary ............................................................................................................................................. 100 4.2 Background .......................................................................................................................................... 101 4.3 Methods .............................................................................................................................................. 102 4.3.1 Population and Setting .................................................................................................................................. 103 4.3.2 Determinants, outcomes, and covariates ....................................................................................................... 103 4.3.3 Outcomes...................................................................................................................................................... 104 4.3.4 Analytic methods........................................................................................................................................... 104 4.3.5 Ethics approval .............................................................................................................................................. 105 xi  4.4 Results ................................................................................................................................................. 105 4.5 Discussion ............................................................................................................................................ 108 4.6 Conclusion............................................................................................................................................ 113 CHAPTER 5: DISCUSSION AND IMPLICATIONS .................................................................................................................. 140 5.1 Key findings .......................................................................................................................................... 140 5.2 Analytic approach ................................................................................................................................ 141 5.3 Role of advanced airway management and tracheal intubation ............................................................ 142 5.4 Role of epinephrine .............................................................................................................................. 144 5.6 Strengths and limitations...................................................................................................................... 146 5.7 Implications ......................................................................................................................................... 147 5.7 Postscript ............................................................................................................................................. 149 REFERENCES ....................................................................................................................................................... 151 APPENDIX ........................................................................................................................................................... 158  xii  List of Tables   Table 1.1 Study Summary ......................................................................................................... 14 Table 2.1 Study subject demographic and clinical characteristics (n=12,821). ........................... 33 Table 2.2. Survival by demographic and clinical characteristics (n=12,821). ............................. 35 Table 2.3. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and survival (n=12,821). ........................................................... 38 Table 2.4. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and return of spontaneous circulation (ROSC) (n=12,281)......... 41 Table 2.5. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients transported to hospital capable of providing specialized post-arrest care* (n=4681)......................................... 44 Supplemental Table 2.1. Return of spontaneous circulation (ROSC) rates by demographic and clinical characteristics (n=12,821). ............................................................................................ 47 Supplemental Table 2.2. Logistic regression model (restricted cohort) showing unadjusted and adjusted association between interventions and other factors and survival among patients transported to hospital with ongoing resuscitation or ROSC (n=7612). ...................................... 50 Supplemental Table 2.3. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients who were NOT transported to a hospital capable of providing specialized post-arrest care* (n=8140). ................................................................................................................................................. 53 Supplemental Table 2.4. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients with a xiii  response time of less than 6 minutes who were transported to hospital capable of providing specialized post arrest care (n=3094). ........................................................................................ 56 Supplemental Table 2.5. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients with a response time of less than 6 minutes who were NOT transported to a hospital capable of providing specialized post-arrest care (n=4710). ....................................................................... 59 Table 3.1.  Study subjects’ demographic and clinical characteristics (n=14,673). ...................... 77 Table 3.2. Survival by demographic and clinical characteristics (n=14673). .............................. 79 Table 3.3. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and survival (n=14,673). ........................................................... 82 Table 3.4. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and return of spontaneous circulation (ROSC) (n=14,673)......... 85 Table 3.5. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients who received specialized post-arrest care (hypothermia, fibrinolysis or diagnostic cardiac catheterization) (n = 2,006). ...................................................................................................................................... 88 Supplemental Table 3.1. Return of spontaneous circulation (ROSC) at emergency department by demographic and clinical characteristics (n=14,673). ................................................................ 91 Supplemental Table 3.2. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between of interventions and other factors and survival among patients with ROSC at arrival in the emergency department (n=3,713). .................................................. 94 xiv  Supplemental Table 3.3. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients who did NOT receive specialized post-arrest care (n=12,667). .......................................................... 97 Table 4.1 Distribution of patient characteristics and other factors among those who did not receive epinephrine and endotracheal intubation (ETI). ........................................................... 115 Table 4.2 Survival rates by patient characteristics, interventions, and other factors. ................. 118 Table 4.3. Logistic regression model showing unadjusted and propensity score matched effect of epinephrine and tracheal intubation (ETI) on survival at discharge from hospital and return of spontaneous circulation present at emergency department among different populations........... 121 Supplemental Table 4.1. Unadjusted and propensity score matched odds ratios of receiving epinephrine, and standardized difference and variance ratio of propensity score matched cases*. ............................................................................................................................................... 122 Supplemental Table 4.2. Unadjusted and propensity score matched odds ratios of receiving tracheal intubation, and standardized difference and variance ratio of propensity score matched cases*. .................................................................................................................................... 125 Supplemental Table 4.3. Unadjusted and propensity score matched odds ratios of receiving epinephrine for cases transported to hospital capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*. ................ 128 Supplemental Table 4.4. Unadjusted and propensity score matched odds ratios for epinephrine for cases that were not transported to hospitals capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*. ................ 131 xv  Supplemental Table 4.5. Unadjusted and propensity score matched odds ratios of receiving tracheal intubation for cases transported to hospital capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*. ........ 134 Supplemental Table 4.6. Unadjusted and propensity score matched odds ratios for tracheal intubation for cases that were not transported to hospitals capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*. ............................................................................................................................................... 137 Appendix Table 1.2 Systematic reviews and meta-analyses of airway interventions research in OHCA .................................................................................................................................... 158 Appendix Table 1.3 Systematic reviews and meta-analyses of epinephrine vs. no epinephrine 161 Appendix Table 1.4 Studies published subsequent to the 2015 systematic reviews and meta-analyses .................................................................................................................................. 163  xvi  List of Figures Figure 2.1 Resuscitation Outcomes Consortium pre-hospital time record web entry page. ......... 31 Figure 2.2 Subject enrolment details. ........................................................................................ 32 Figure 3.1 Subject enrolment details. ........................................................................................ 76 Figure 4.1. Box plots of propensity scores (raw vs. matched) for epinephrine and tracheal intubation treated cases and controls. ...................................................................................... 114 Appendix Figure 1. Truncated conditional probability patient flowchart — proportions of patients alive at each node when patient achieved ROSC or had ongoing resuscitation at transport from scene. ............................................................................................................................. 166 Appendix Figure 2. Truncated conditional probability patient flowchart — proportions of patients alive at each node when resuscitation was terminated in the field. .............................. 167    xvii  List of Abbreviations 95% CI  95% Confidence Interval AED  Automatic External Defibrillator AHA  American Heart Association ALS  Advanced Life Support AOR  Adjusted Odds Ratio BLS  Basic Life Support CPR  Cardiopulmonary Resuscitation EMS  Emergency Medical Services ETCO2  End tidal carbon dioxide ETI  Tracheal Intubation ILCOR International Liaison Committee on Resuscitation IQR  Interquartile Range N  Number NNT  Number Needed to Treat OHCA  Out-of-Hospital Cardiac Arrest OR  Odds Ratio  PEA  Pulseless Electrical Activity RCT  Randomized Controlled Trial ROC  Resuscitation Outcomes Consortium ROSC   Return of Spontaneous Circulation RR  Relative Risk RSI  Rapid Sequence Intubation xviii  SD   Standard Deviation SGA  Supraglottic airway SPSS  Statistical Package for the Social Sciences VF/VT  Ventricular Fibrillation/Ventricular Tachycardia xix  Acknowledgements As with all PhD research, this was not done alone. I am indebted to many people who wittingly and unwittingly provided guidance and encouragement over the past 12 years.  I would like to thank those people who were part of my thesis committee in previous years: Charlyn Black, Scott Emerson, Jim Christenson, and Jason Sutherland. Thank you so much for sparking thought and helping me to narrow down the focus of my dissertation.    My current committee has been incredibly supportive and responsive as I ran for the finish line.  Thank you to John Tallon, who has challenged my thinking and encouraged me to apply my epidemiological training to EMS. Thank you to Sam Sheps, who has been part of my training and reflection for over 15 years. Much of my foundational thinking is a result of Sam’s influence. Riyad Abu-Laban seems to have started this whole effort to get a PhD, and despite my often-unfocused efforts to get this done, has remained a strong supporter and advocate. And I want to thank KS Joseph, whose laser sharp intellect hides behind a very caring and supportive man. All of you, Sam, John, Riyad, and KS have been so incredibly cohesive and supportive over the past couple of years, and I don’t believe I would be writing this if it weren’t for all of you.  I have had the tremendous benefit of working with an extra-ordinary team at the Resuscitation Outcomes Consortium, who provided me data, analytic support and guidance. Thank you Rob Schmicker, Scott Emerson, Susan May, Siobhan Brown, and Tom Rea—the members of the publication committee. Thanks also to all of the ROC investigators, coordinators and paramedics xx  at the collaborating sites who helped to stoke the fire of curiosity in me to learn more about out-of-hospital cardiac arrest resuscitation.  While there has been formal support from academics and mentors, there are many informal supports I have received along the way. George Papadopoulos has encouraged me to continue probing, by acknowledging that it helped him better understand his own problems. The other paramedics turned academics have also helped encourage me. Thank you Ian Blanchard, Judah Goldstein, and Jan Jensen for continuing to push the agenda for an academically rigorous EMS system. Thanks to Imelda Wong for taking an interest in paramedics as a subject of study for her own PhD. Thanks to David Hostler and Daniel Paterson for being examples of where I can end up when I am done.   My profound gratitude for important personal support goes to Barb Murphy, whom I love, and thanks for putting up with this project. Thanks for teaching me to be a better person, regardless of the circumstances.    I would like to acknowledge the financial support that I have received throughout my PhD, including the Western Regional Training Centre for Health Services and Policy Research for the WRTC fellowship training I received. Additionally, I would like to acknowledge the PhD tuition fee support initially provided by the Department of Healthcare and Epidemiology and subsequently by the School of Population and Public Health at the University of British Columbia.  I received a Michael Smith Foundation for Health Research Senior Trainee scholarship for which I am incredibly grateful. I was also one of the inaugural recipients of the xxi  Jump Start Resuscitation Doctoral Research Awards, provided by the Heart and Stroke Foundation of Canada and the Institute of Circulatory and Respiratory Health at the Canadian Institutes of Health Research. Finally, I acknowledge the very generous support of the Canadian Institutes of Health Research and the BC Children’s Hospital Research Institute.    Chapter 3 was based on ROC PRIMED Research Materials obtained from the NHLBI Biologic Specimen and Data Repository Information Coordinating Center and does not necessarily reflect the opinions or views of the Resuscitation Outcomes Consortium or the NHLBI. xxii  Dedication I would like to dedicate this to two people.   First to my father, William, who watched me and quietly encouraged me for much of this effort. Unfortunately, I took too long to finish, and he won’t be there to celebrate this accomplishment.   And most importantly, to Barb, who showed me that this isn’t really worth it if you don’t do it with someone you love.  1 Chapter 1: Introduction Cardiac arrest affected approximately 45,000 individuals in Canada in 2012 [1], while approximately 326,200 individuals experienced an Emergency Medical Services (EMS) assessed out-of-hospital cardiac arrest (OHCA) in 2011 in the United States [2]. In the United States, only approximately 10% [3] of those who suffer out of hospital cardiac arrest survive to hospital discharge, and about 25% of those who were treated by EMS had no previous symptoms. The combination of the sudden onset of cardiac arrest and low survival make it an important public health issue that warrants ongoing efforts to improve survival with neurological function intact.  Central to that effort to improve survival after cardiac arrest has been the chain-of-survival [4] originally proposed in 1989 [5] and adopted by the American Heart Association [6, 7]. The chain-of-survival serves as the central framework for the organization and evaluation of the treatment for cardiac arrest. It brings together the efforts of the lay public, EMS system providers, and hospital staff to better define potential treatments and coordinate roles. The chain-of-survival has undergone numerous revisions [4, 5, 8] since its first introduction, but its fundamental structure has remained.   EMS play a pivotal role bridging the activities of the public who may initiate an organized response and perform cardiopulmonary resuscitation (CPR) or deploy a public access defibrillator following OHCA, and emergency department and cardiology staff who take over resuscitation and, in many instances, initiate post-arrest care. Much of the care provided by EMS is prescribed by medical directives, protocols, or treatment guidelines developed by EMS medical directors and supported by research evidence [9].  In the case of cardiac arrest   2 resuscitation, guidelines published by national resuscitation councils in collaboration with the International Liaison Committee on Resuscitation (ILCOR) [7, 10-15] form the foundation of EMS resuscitation. These guidelines contain specific treatment recommendations for EMS clinical care related to the type of CPR [16, 17], airway management, and drug therapy [18, 19]. While published evidence [20] has confirmed the survival benefit of the guidelines used, this evidence is often based on observational studies which have provided conflicting signals about the most efficacious treatment options. For example, the case for advanced life support and its application in a prehospital setting exists, and the role of epinephrine [21] and tracheal intubation [22, 23] remains controversial. These interventions are embedded in the resuscitation guidelines despite substantial uncertainty about their effectiveness and safety.  The following sections describe the existing evidence and proposed theories on the effects of tracheal intubation and epinephrine in out-of-hospital cardiac arrest resuscitation.  1.1 Tracheal intubation and airway management in out-of-hospital cardiac arrest The goal of airway management in cardiac arrest is to ensure adequate oxygenation and ventilation, and to protect the patient’s airway from aspiration due to regurgitation of abdominal contents [24]. This can be accomplished by basic airway management via manual manipulation of the mandible and an oral pharyngeal airway, or advanced airway management techniques via supraglottic airways (SGA) or tracheal intubation (ETI).   The American Heart Association (AHA) and International Liaison Committee on Resuscitation (ILCOR) resuscitation guidelines have relied on a systematic review process to evaluate and   3 grade available evidence [25]. Although the guideline review process is ideally informed by well-conducted large randomized controlled trials (RCT), the absence of such methodologically rigorous studies resulted in resuscitation guidelines being based on less rigorous evidence. Lack of randomized trial evidence on the relative efficacy of Basic Life Support (BLS) vs. Advanced Life Support (ALS) airway management techniques meant that ALS resuscitation guidelines from the AHA have appeared in the 2005 [10], 2010 [15] and 2015 [19] guidelines without definitive evidence on the relative efficacy of these interventions.      A number of systematic reviews have been conducted since the 2010 resuscitation guidelines were released (Appendix Table 1.1). In 2014, Fouche and colleagues [26] conducted a systematic review and meta-analysis comparing advanced airway interventions (AAI) with basic airway interventions (BAI). The studies covered the period of 1988 to 2013, and AAI included laryngeal masks, ETI, SGA, double lumen airways, and trans-tracheal and trans-cricothyroid devices. (ETI and SGAs are the only devices that continue to be used). The meta-analysis included data from 17 studies of patients suffering non-traumatic and traumatic OHCA who were successfully managed with an AAI or BAI respectively. In this mixed population, no short-term survival benefit was found for AAI or BAI (odds ratio (OR) 0.84, 95% CI 0.62–1.13) and reduced odds of survival to hospital discharge or later for those treated with AAI (OR 0.49 95% CI 0.37–0.65). For cases of non-traumatic cardiac arrest who were treated with ETI, no short-term benefit was found (OR 0.86 95% CI 0.53–1.38), while the odds of surviving to hospital discharge or later were reduced following treatment with ETI (OR 0.49 95% CI 0.32–0.74).      4 In 2014, Tiah et al. [27] published a systematic review comparing ETI to SGA. They included 5 studies published between 2009 and 2014 with a total pooled study size of 303,348 adult cases of cardiac arrest. Unfortunately, the investigators did not conduct a meta-analysis but noted the lack of a consistent positive association between ETI and any outcome, leading to their conclusion that the evidence did not support the superiority of ETI over SGA. There was substantial variation in the size of the studies included in this systematic review, with studies ranging from 172 subjects (finding of no difference, with large 95% confidence intervals) to 281,522 subjects (showing a beneficial association for return of spontaneous circulation (ROSC) but a null effect for neurologically intact survival).   A systematic review and meta-analysis in 2015 by Benoit et al. [28], compared outcomes for ETI vs. SGA in 10 observational studies of non-traumatic OHCA with a pooled study size of 75,649 subjects. They found that ETI resulted in higher relative odds of ROSC (OR 1.28, 95% CI 1.05–1.55), and neurologically intact survival (OR 1.33 95% CI 1.09–1.61) compared with SGA.  However, odds of survival to hospital discharge were not significantly different between ETI and SGA (OR 1.15, 95% CI 0.97–1.37).  In 2016, Jeong et al. [29] returned to the question of the superiority of AAI compared with BAI in a systematic review and meta-analysis of 10 studies covering the years 1993 to 2015, with a pooled study size of 84,905 pediatric and adult cases. In a random effects analysis, AAI was associated with reduced survival (OR 0.51, 0.29–0.90) compared with BAI, and ETI resulted in a non-significant reduction in survival (OR 0.44, 95% CI 0.16–1.23) compared with BAI.      5 A recently published RCT by Jabre et al. [30] designed to determine if bag-mask ventilation (BMV) was non-inferior to ETI in a multicenter trial set in Belgium and France (where physicians and nurses, rather than paramedics, are responsible for delivering ALS care following OHCA). These investigators enrolled 2,043 patients and followed them for 28 days post-discharge from hospital to collect information on neurological status. The investigators set a non-inferiority margin of 1% (absolute) and found 4.3% of cases treated with BVM and 4.2% of cases treated with ETI had favourable status at 28 days post discharge (P value for non-inferiority 0.11). Their results were inconclusive as they had failed to demonstrate non-inferiority or inferiority. In the discussion, the investigators noted that their a priori establishment of a 1% difference for demonstrating non-inferiority between the treatment groups may have resulted in an underpowered study and led to their call for the design and execution of a larger, better powered study to definitively establish equivalence or superiority.  Variation in study populations included in these reviews makes drawing conclusions difficult.  Fouche et al. [26] included both pediatric and traumatic arrest cases in their systematic review, and both of these populations have very different rates of survival compared with adult non-traumatic cardiac arrest. Thus, the result of a meta-analysis that combines such populations is not clearly generalizable adult non-traumatic cardiac arrest cases requiring resuscitation. The lack of a meta-analysis in the review performed by Tiah et al. [27] limits the utility of this study. The narrowly focused review conducted by Benoit [28] provides some information to guide decisions regarding use of ETI vs. SGA in the non-traumatic adult population. However, the lack of inclusion of BAI means that results are potentially of little help in guiding EMS systems that are staffed by BLS providers only (who would not normally be trained or capable of managing an   6 airway using advanced techniques like ETI). The inclusion of older studies, which included procedures that relied on airway management techniques/devices that are no longer available, and a mixed pediatric and adult population, make interpreting and applying the results from Jeong et al. [29] difficult with regard to understanding how best to guide airway management in adults using currently available techniques and devices. Finally, the RCT by Jabre et al. [30] is important for a number of reasons, including providing information that will aid in the design of future definitive equivalence or superiority trials, as well as demonstrating that it is possible to conduct a study of ETI versus BAI, which by necessity cannot be blinded.  1.2 Epinephrine administration in out-of-hospital cardiac arrest resuscitation Epinephrine (adrenaline) has been used in cardiac arrest resuscitation for decades. It is thought that the effects of epinephrine are two-fold: first, interaction with alpha receptors results in vasoconstriction, which improves perfusion pressure to the brain and the heart and, second, interaction with beta-1 receptors in the heart increases heart rate and contractility, and improves electrical conduction through the atrio-ventricular node [31]. However, despite epinephrine’s widespread use in resuscitation, evidence is mixed regarding its effectiveness in improving survival to discharge from hospital or achieving good neurological outcomes [32-35].    As with airway management, a number of systematic reviews have been published since 2010 examining the role of epinephrine in non-traumatic OHCA (Appendix Table 1.2). In a systematic review and meta-analysis Atiksawedparit et al. [36] reported on published studies examining the benefit of epinephrine vs. no epinephrine in OHCA. They examined effects of epinephrine on any ROSC, admission to hospital, and survival to discharge. The odds ratio for ROSC at any   7 point in the resuscitation from a total of four studies with a pooled study size of 2,381 subjects was 0.93, 95% CI 0.50–1.75. The odds ratio for survival to hospital admission from a total of 8 studies with a pooled study size of 8,470 cases was 1.05, 95% CI 0.80–1.38 and the pooled odds for hospital discharge from 7 cohort studies with a size of 4,743 cases was 0.69, 95% CI 0.48–1.00.  Lin et al. [37] conducted a systematic review of randomized trials that examined standard dose epinephrine vs. placebo and several other treatment comparisons including standard dose epinephrine vs. high dose epinephrine, standard dose epinephrine vs. epinephrine and vasopressin, and standard dose epinephrine vs. vasopressin. There was no survival-to-discharge benefit or neurological outcome differences found in any groups. In the standard dose epinephrine review, they were only able to identify one study [34]), with 534 subjects, which found that those receiving epinephrine had higher rates of ROSC (relative risk (RR) 2.80, 95% CI 1.78–4.41) but not in terms of survival to discharge from hospital (RR 2.12, 95% CI 0.75–6.02) or improved neurological outcomes (RR 1.73, 95% CI 0.59–5.11).    Loomba et al.[38] conducted a systematic review and meta-analysis of 14 observational studies of patients who received prehospital epinephrine versus those who did not. In the analysis of ROSC, which included 9 studies with a pooled size of 640,258 subjects, they found an increase in the odds of ROSC for those receiving epinephrine (OR 2.84, 95% CI 2.28–3.54). Eight studies, with a pooled size of 6,527 cases, provided data for their examination of discharge from hospital, which showed no survival benefit for epinephrine (OR 0.82, 95% CI 0.46–1.48). There was no survival benefit at 1 month, based on a pooled study size of 647,770 subjects from 5   8 studies (OR 1.03, 0.70–1.34). In their examination of a favourable neurological outcome, based on 9 studies with a pooled size of 641,723 subjects, epinephrine resulted in reduced relative odds of a favourable neurological outcome (OR, 0.51, 0.31–0.84).   A number of studies (Appendix Table 1.3; [39-43]) have been published subsequent to the Loomba review [38] that evaluated the effect of epinephrine vs. no epinephrine. Fukuda et al.[39] used a population-based collection of cases of cardiac arrest enrolled in a Japanese cardiac arrest registry to perform a matched propensity score analysis of the effect of epinephrine in shockable and non-shockable cardiac rhythms. In addition to including standard Utstein variables of age, sex, witnessed arrest, bystander resuscitation, first rhythm, cause of arrest, and response time, they also included duration from time of call to administration of epinephrine and advanced airway management. For shockable rhythms, they found decreased odds of ROSC (OR 0.91 95% CI 0.82–1.00), 1-month survival (OR 0.53 95% CI 0.47–0.59), and favourable neurological outcome (OR 0.38 95% CI 0.33–0.43), while for non-shockable rhythms they found increased odds of ROSC (OR 4.08 95% CI 3.85–4.32), no-benefit or harm for 1-month survival (OR 1.00 95% CI 0.92–1.09), and a reduction in the odds of a favourable neurological outcome (OR 0.47 95% CI 0.39–0.56). These results suggest that epinephrine results in increased ROSC rates among cases with non-shockable rhythms, but that the benefit is not carried over to survival at discharge from hospital, and in fact leads to reduced relative odds of a favourable neurological outcome. Further, cases with a shockable rhythm do not benefit at all from receipt of epinephrine.     9 Tanaka et al. [41] conducted an analysis of the timing of epinephrine administration. Cases were divided into quartiles (early = 5 – 18 minutes, intermediate = 19 – 23 minutes, late = 24 – 29 minutes, and very late = 30 – 62 minutes), with late administration as the reference category.  The analysis adjusted for common Utstein variables plus defibrillation (number) and type of airway. They excluded cases who achieved ROSC within two minutes of receiving bystander CPR, as those cases would not have been eligible for epinephrine administration. Their results showed dose response increase in the relative odds of ROSC when epinephrine was administered in an earlier period compared to late period (administered early vs. late (OR 1.66 95% CI 1.49–1.85) or in the intermediate vs. the late period (OR 1.25 95% CI 1.12–1.40).  Rates of favourable neurological outcomes were also improved following early and intermediate epinephrine administration compared with late administration (OR 2.49, 95% CI 1.90–3.27 and OR 1.53, 95% CI 1.14–2.05, respectively). Odds of survival with a favourable neurological outcome were improved among those not receiving epinephrine compared with those receiving epinephrine (OR among those not receiving epinephrine vs. late epinephrine 3.33, 95% CI 2.60–4.26). However, those not receiving epinephrine had poorer rates of ROSC (OR among those not receiving epinephrine vs. late epinephrine 0.62, 0.57–0.68).   Sagisaka et al.[40] conducted a retrospective cohort study of the effect of repeated doses of epinephrine while attempting to control for timing of administration based on the duration from the time of the emergency call to the first administration of epinephrine. In their analysis of bystander witnessed cases that received epinephrine, they found an inverse dose-response relationship between the number of epinephrine doses administrated and survival and favourable neurological outcome. In a combined analysis where they examined the effect of repeated doses   10 within different time intervals, they found that the inverse relationship held, with increasingly poor survival and neurological outcomes as epinephrine doses and time intervals increased. Given that timing of administration of epinephrine was related to the EMS response time, it was difficult to determine if the effect resulted from an extended period of arrest and was unrelated to the dosing itself. Furthermore, by excluding cases that did not receive epinephrine, the investigators were not able to determine if no epinephrine was worse or better than the dosing under study.  Utea et al. [42] performed an examination of the Japanese national cardiac registry to look at the effect of early vs. late administration of epinephrine (within 10 minutes or less of initial patient contact, and later than 10 minutes after initial patient contact), and within short vs. long response times (defined as 8 minutes or less, and 8 to 16 minutes) in cases that were witnessed by bystanders and received epinephrine. The investigators found that early administration of epinephrine was associated with improved neurological outcomes regardless of the response time (short response time group OR for early vs. late administration 2.12, 95% CI 1.54–2.92 and delayed response time group OR for early vs. late administration 2.66 95% CI 1.97–3.59). As with the two earlier analyses, the investigators did not explore the reasons for the delay in epinephrine administration, including imbalances with regard to confounders such as the need to control a difficult airway or delayed performance of CPR, both of which have been shown to be associated with the timing of epinephrine administration and poor outcomes.   Hagihara et al. [43] assessed the effect of advanced airway placement and/or epinephrine vs. no ALS or epinephrine using the Japanese national cardiac registry. In their matched propensity   11 score analysis of bystander witnessed cardiac arrest they found that ALS care and/or epinephrine was associated with reduced odds of 1-month survival (OR 0.88, 95% CI 0.80–0.97) and reduced neurological status at 1-month (OR 0.56, 95% CI 0.48–0.66) compared with no ALS care and no epinephrine. They found no evidence of an interaction between advanced airway placement and epinephrine administration. However, they did find a stronger negative association between advanced airway management and neurological status than between epinephrine administration and neurological status.   1.3 Summary of background literature While substantial effort has been made to understand the role of epinephrine and advanced airway management, including ETI in OHCA resuscitation, the effect of these treatments compared with no treatment remains unknown. Despite being a central component of the resuscitation guidelines, and resuscitation practice for decades [44], only one randomized trial (by Jacobs et al. (34)) has been conducted comparing epinephrine with placebo. This trial was stopped early, resulting in a suboptimal study size and an inability to adequately assess the effects on the primary study outcome, namely survival. Jacobs et al. (34) found that ROSC rates were substantially improved following epinephrine administration (OR 3.4, 95% CI 2.0–5.6) but survival to hospital discharge was not significantly improved (OR 2.2, 95% CI 0.7–6.3). Stratified analyses of this trial by initial rhythm also showed interesting results. Although larger effects on ROSC rates were found following epinephrine administration among those with non-shockable rhythms compared with those who had shockable rhythms (OR 6.9, 95% CI 2.6–18.4 vs. OR 2.4, 95% 1.2–4.5), overall survival to hospital discharge was low among those with non-  12 shockable rhythms (0.0% among placebo and 1.3% among epinephrine group) compared with those who has a shockable rhythm (4.0% among placebo and 7.6% among epinephrine group).  Two randomized trials comparing ETI to SGA are currently underway [45, 46], and until the results are published there remains a lack of evidence of the superiority of ETI over basic airway management techniques.   The most salient of the findings from the non-experimental studies to date assessing the effect epinephrine are the findings that 1. Effects of interventions vary depending on the outcome assessed. Specifically, effects of intervention on ROSC differ in magnitude and/or direction from effects on survival to hospital discharge and favourable neurological status at hospital discharge;  2. The effect of epinephrine varies among those with a shockable vs. non-shockable initial rhythm; 3. Early administration of epinephrine is associated with significant improvements in outcomes (including neurological outcomes) compared with late administration of epinephrine and the effect appears to be independent of response time. This finding suggests that confounding by indication may not be a significant issue in non-experimental studies of epinephrine efficacy. Administration of epinephrine appears to be protocol-based and not necessarily prioritized to the most critical OHCA cases. If administration of epinephrine was prioritized to the most critical OHCA cases, that would have resulted in early administration being associated with worse outcomes (i.e. confounding by indication);    13 4. Observational studies assessing the effect of epinephrine vs no epinephrine on survival have yielded contradictory results, with effects varying in both magnitude and direction; 5. Efforts to understand these links in the chain-of-survival have focused on isolated effects without adequate consideration of the effects of associated interventions.  Failing to account for clinical interventions associated with the chain-of-survival may result in confounded results, especially if the clinical interventions are performed in a sequenced manner as suggested by the chain-of-survival; 6. Despite nearly 40 years of practice and advocacy of the chain-of-survival, RCT evidence for efficacy of specific interventions is lacking. In the interim, a careful examination of the components of the chain-of-survival is warranted, including an in-depth exploration of potentially confounding and bias. 1.4 Study objectives  The purpose of this dissertation research was to investigate the effect of clinical interventions associated with the chain-of-survival, specifically to examine the effects of two key advanced life support interventions, namely, epinephrine administration and advanced airway placement. The effect of each intervention was assessed within a model that contained all the clinical interventions that form part of the resuscitation chain-of-survival. 1.5 Methods summary This dissertation contains three observational cohort studies of interventions associated with the chain of survival, conducted using two different datasets, and analyzed using several statistical approaches. The following table is intended to provide a brief summary of the data sources, analytic techniques, main findings, and subgroup analyses performed.    14 Table 1.1 Study Summary Chapter Data source Analytic Method Interventions Survival to hospital discharge: adjusted OR (95% CI) ROSC Subgroup analyses 2 ROC Epistry Logistic regression (univariate and multivariable) Bystander attempted resuscitation 1.30 (1.10  - 1.55) 0.98 (0.88 - 1.10) Cases transported to hospitals with specialized post arrest care, NOT transported to hospitals with specialized post-arrest care, cases with an EMS response time of less than 6 mins transported to hospital with specialized post arrest care, cases with EMS response time of less than 6 mins NOT transported to hospital with specialized post arrest care.   n=12821 EMS CPR prior to Rhythm analysis 0.81 (0.65 - 1.01) 1.35 (1.14 - 1.60)    Epinephrine  0.11 (0.09 - 0.13) 0.33 (0.28 - 0.39)     Advanced airway 1.68 (1.34 - 2.12) 1.78 (1.49 - 2.12)       Transported to hospital capable of specialized post arrest care 4.86 (4.00 - 5.91) 3134.6 (1667 - 5892)      Survival to 6 months after hospital discharge: adjusted OR (95% CI)    3 ROC PRIMED Logistic regression (univariate and multivariable) Bystander attempted resuscitation 1.21 (1.01 – 1.43) 1.10 (1.00 – 1.21) cases that received specialized post-arrest care, cases that DID NOT receive   n=14673 Epinephrine  0.11 (0.09 – 0.14) 0.51 (0.45 – 0.59)    Advanced airway 1.13 (0.87 – 1.47) 3.20 (2.66 – 3.84)   15     interaction between first rhythm and advanced airway placement or epinephrine administration. received  Specialized post arrest care (prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis) 16.41 (13.7 – 19.7) 23.2 (20.1 – 26.7) specialized post arrest care      Survival to hospital discharge adjusted OR (95% CI)    4 ROC Epistry Propensity score matching Epinephrine  0.15 (0.11 - 0.19) 0.50- (0.42 - 0.60) Cases transported to a hospital with specialized post-arrest care, NOT transported to a hospital with specialized post-arrest care.   n=12821   Tracheal intubation 1.97 (1.62 – 2.40) 2.31 (1.92 – 2.77)   16 Chapter 2: Effects of the chain-of-survival interventions and other factors on survival following non-traumatic out of hospital cardiac arrest 2.1 Summary Survival rates following out-of-hospital cardiac arrest remain poor and the effects of the individual interventions in the resuscitation chain remain not yet fully defined. We sought to quantify the potential effects of the chain-of-survival interventions and other factors on survival following out-of-hospital cardiac arrest (OHCA).   We conducted a prospective cohort study of adult (> 19 years old) OHCA patients. Interventions studied included:  Cardio-pulmonary resuscitation (CPR) prior to Emergency Medical Services (EMS) rhythm analysis; advanced airway placement (defined as tracheal intubation, extra-glottic airway, or cricothyrotomy); administration of prehospital epinephrine; and specialized post-arrest care. Stratification and logistic regression were used to estimate adjusted odds ratio (AOR) and 95% confidence intervals (CI) after controlling for confounding and other bias.  The study population included 12,821 non-traumatic OHCA cases, of whom 972 (7.6%) survived to hospital discharge. Factors associated with improved survival included younger age (> 75 years vs. 20–39 years: AOR 0.35, 95% CI 0.25–0.50), public location of arrest (AOR 1.59, 95% CI 1.34–1.90), bystander-witnessed arrest (AOR 2.06, 95% CI 1.71–2.49), ventricular fibrillation/tachycardia as the initial rhythm and shorter response times (> 16 vs. < 6 minutes: AOR 0.30, 95% CI 0.09–0.97). Bystander-attempted resuscitation (AOR 1.30, 95% CI 1.10–1.55) and advanced airway placement (AOR 1.68, 95% CI 1.34–2.12) were variably associated   17 with improved survival, while EMS CPR prior to rhythm analysis improved ROSC rates (AOR 1.35, 95% CI 1.14–1.60) but not survival. Epinephrine use was associated with a substantial reduction in survival, while post-arrest care could not be assessed due to survivor bias.   Age, location of arrest, bystander witness, initial cardiac rhythm, response times, bystander resuscitation and advanced airway placement are associated with improved survival following out-of-hospital cardiac arrest. In addition to patient characteristics, therapies by laypersons, first responder, and advanced EMS providers each contributed to the chances of survival. Epinephrine administration was associated with a decrease in ROSC and a substantial decrease in survival.    2.2 Introduction The chain-of-survival has been the central theoretical framework informing cardiac arrest resuscitation [47] and Emergency Medical Services (EMS) system design in high income nations [48 pp.29-43] since it was first described over 40 years ago [5, 49]. The chain-of-survival has gone through several revisions [4, 5, 8] but the central premise has remained unchanged; namely that rapid, sequenced response to cardiac arrest improves the probability of survival. In 2006, the chain-of-survival consisted of four links [8]: early recognition and system activation, early cardiopulmonary resuscitation (CPR), early defibrillation, and early advanced life support. In 2010 [50, 51], transportation to specialized post-resuscitation care [47] (specialized post-arrest care) was added as a fifth link in the chain.  Research on the individual elements in the chain-of-survival in out-of-hospital cardiac arrest (OHCA) has, in some cases, demonstrated improved survival. Early recognition and activation of   18 the EMS system is important and studies have consistently demonstrated the benefit of bystander-witnessed arrest [52], which shortens both the duration from arrest to EMS notification, and initiation of resuscitative efforts. Evidence for the benefit of a number of advanced cardiac life support (ALS) procedures and drugs is lacking (e.g. epinephrine, [53] antiarrhythmics, [54, 55] and tracheal intubation [28]), leading to questions regarding the value of ALS in OHCA [56]. Beyond this, efforts to determine which, if any, components of the chain-of-survival should be optimized to improve survival have yielded equivocal results (e.g. immediate vs. delayed defibrillation [57], use of an impedance threshold device [58], and continuous versus interrupted chest compressions [59]). Nevertheless, there is evidence suggesting that some EMS systems have achieved improvements by changing their approach to resuscitation [9, 60]. However, survival rates following OHCA remain poor and the effects of the individual components of the chain-of-survival links remain unclear.   The purpose of this study was to quantify the potential effects of the individual interventions in the chain-of-survival and the effects of known predictors of survival, such as age, response time, location of arrest, and bystander participation in the resuscitation.   2.3 Methods  We carried out a cohort study of adult (> 19 years) non-traumatic OHCA cases identified by the Resuscitation Outcomes Consortium (ROC; [61, 62] between December 2005 and May 2007. The ROC is a resuscitation research network comprising 11 academic coordinating centers throughout North America and 264 EMS agencies, which manage and transfer patients with prehospital emergencies to 287 hospitals in Canada and the United States [63].    19  2.3.1 Population and setting  The ROC covers a population of approximately 25 million individuals in mostly urban and metropolitan centers in seven US states and two Canadian provinces [61, 62].  2.3.2 Determinants, outcomes, and covariates  The Utstein reporting template [64] was used to collect information based on a consistent set of data definitions. Standard information and time stamps were collected for each of the following: bystanders (arrest witnessed, cardiopulmonary resuscitation (CPR), attempted automatic external defibrillator (AED) deployment, and defibrillation); Basic Life Support (BLS) activities: (arrival at the scene, EMS CPR by BLS providers, AED rhythm analysis, and AED defibrillation); ALS activities (arrival at the scene, EMS CPR by ALS providers, rhythm analysis, defibrillation, drug therapy, advanced airway1 placement); and arrival at hospital. EMS agencies reported the source of the time stamps and if the time stamps were standardized or synchronized in their system. Trained EMS researchers reconciled the time records [62]. Based on the time stamps, and review of the medical records, trained EMS researchers entered the numbered sequence of specific resuscitation interventions as part of routine data abstraction (Figure 2.1.).   We sought to quantify the association of five prehospital clinical interventions with survival. The prehospital clinical interventions included: bystander attempted resuscitation (i.e. bystander-                                               1 Advanced airway management consisted of 7 non-mutually exclusive categories: supraglottic airway, oral endotracheal tube, nasal endotracheal, cricothyrotomy, continuous positive pressure ventilation, ventilator, and rapid sequence intubation.   20 attempted CPR and deployment or use of a public access defibrillator; execution of EMS CPR prior to EMS defibrillator rhythm analysis, which was derived based on the sequenced time stamp information; placement of an advanced airway (defined as tracheal intubation, placement of an extra-glottic airway or cricothyrotomy); prehospital administration of epinephrine; and administration or availability of post-arrest care, which included prehospital hypothermia, or transportation to a hospital capable of percutaneous coronary intervention (PCI) or having an electrophysiology lab. These interventions represent the components in the chain-of-survival [4].   2.3.3 Outcomes The primary outcome was survival status at hospital discharge. Additional analysis was based on return of spontaneous circulation (ROSC) in any setting (prehospital or in hospital).  2.3.4 Analytic methods  Descriptive statistics including frequencies, proportions, means (standard deviation, SD), and medians (interquartile range, IQR) were calculated. Logistic regression was used to assess the relationship between the clinical interventions and other factors and survival to hospital discharge. Potential confounders were included in the regression model based on univariate relationships with five prehospital interventions and survival status. The effect of excluding potential confounders (on the relationship between the chain-of-survival interventions and survival status) was considered, with a change of  > 10% in the association deemed to be evidence of confounding [65]. Variables assessed as potential confounders included standard Utstein variables [66] and additional system level descriptors including country and ROC site (anonymized).    21 The original analytic approach suggested our analysis should be conducted using the entire cohort. However, upon our initial exploratory data analysis we discovered that transport to hospital and provision of post-arrest care was at least partly determined by the patient’s condition and response to resuscitative efforts. This resulted in a survival bias that artifactually inflated the estimate of the protective effect of post-arrest care. Subsequent analyses estimating the effects of interventions and other factors were performed after stratifying based on provision of post-arrest care (e.g., among patients who received post-arrest care only and separate analysis among those who did not receive post-arrest care). Additional analyses were carried out to address the potential for confounding by indication. Such analyses involved restricting the cohort to patients with a short response time (< 6 minutes). Data management, manipulation, and analysis were performed using IBM SPSS Ver. 23.0 (IBM, Armonk, NY).  2.3.5 Ethics approval  The University of British Columbia, Behavioral Research Ethics Board, provided approval for this study. In addition, the ROC obtained ethics approval from all participating sites [62].  2.3.6 Results There were 28,484 cases of OHCA enrolled in the ROC database during the study period. Of these, 15,397 were excluded from the study because resuscitation was not attempted, EMS witnessed the arrest, the patient was <20 years old, a non-ROC unit responded to the call, or information was missing (Figure 2.2). A total of 12,821 cases of cardiac arrest comprised the study population and, of these, 972 (7.6%) survived to hospital discharge.    22 The median age of the cohort was 66.0 (IQR 25) years, and age groups 40–60, 61–75, and >75 years contained approximately 30% of the cases each (Table 2.1). Nearly two-thirds (65.3), of the cases were male and the median response time was five minutes. Attempted bystander resuscitation was provided in 64.8% of cases, EMS CPR preceded defibrillator rhythm analysis in 82.3%, 80.2% had an advanced airway placed, and 76.5% of the cases received epinephrine. Over a third (36.5%) of the cases were transported to a hospital capable of providing PCI or electrophysiology lab or received prehospital hypothermia (post-arrest care).  Statistically significant differences were found in crude survival rates by age, response time and other Utstein variables (Table 2.2). Survival rates were 9.3% and 10.7% among those aged 20–39 and 40–60 years, respectively, compared with a survival rates of 7.8% among cases aged 61–75 years and 3.8% among those >75 years. An inverse dose-response relationship was found between response time and survival: survival rates among those with a response time of <6, 6–9, 10–16 and >16 minutes were 9%, 5.6%, 4.0% and 2.2%, respectively. ROC Sites 2 and 6 had the highest survival rates (15.5% and 9.5%, respectively). Patients who received bystander attempted CPR had significantly higher survival rates compared with those who did not receive bystander attempted CPR (11.3% vs. 5.6%). Patients who had EMS CPR performed prior to defibrillation analysis (7.4% vs. 8.4%), advanced airway placement (7.4% vs. 8.5%) and receipt of epinephrine (4.4 vs. 17.9%) had lower rates of survival, whereas those who received specialized post-arrest care had higher survival rates (14.5% vs. 3.6%; Table 2.2). The absolute risk difference associated with surviving following receipt of epinephrine was -13.5%, and for advanced airway it was – 1.1%.      23 ROSC rates were substantially higher than survival rates and there were distinct differences between the association patterns of ROSC and survival (Supplemental Table 2.1). The relationship between response time and ROSC was positive, with ROSC rates among patients with response times of <6, 6–9, 10–16 and >16 minutes being 59.0%, 59.1%, 59.2% and 65.2%, respectively. Bystander-attempted resuscitation did not significantly affect ROSC rates (59.0% vs. 59.7%). Although rates of ROSC were significantly lower among those who received EMS CPR prior to rhythm analysis (57.2% vs. 69.0%) and those who received epinephrine (58.5% vs. 61.8%), the relative differences in ROSC rates were smaller than the differences in survival rates. Cases with advanced airway placement had significantly higher rates of ROSC compared with those who did not have advanced airway placement (61.1% vs. 51.8%). Almost all cases (99.7%) who received post-arrest care had ROSC compared with a smaller proportion of those who did not receive post-arrest care (36.0%). The absolute risk difference of achieving ROSC following receipt of epinephrine was – 3.3%, and for advanced airway management it was 3.0%.   Table 2.3 shows unadjusted and adjusted odds ratios expressing the association between interventions and other factors and survival in the study population. Non-public location of arrest (AOR 1.59, 95% CI 1.34–1.90), bystander-witnessed arrest (AOR 2.06, 95% CI 1.71–2.49), ventricular fibrillation/tachycardia as the initial rhythm and specific ROC site (sites 2, 3, 4, 6, 7 and 9) were associated with improved survival. Bystander-attempted resuscitation (AOR 1.30, 95% CI 1.10–1.55) and advanced airway placement (1.68, 95% CI 1.34–2.12) were also positively associated with survival. However, EMS CPR performed prior to rhythm analysis was not significantly associated with survival, while epinephrine use was strongly associated with   24 reduced odds of survival (AOR 0.11, 95% CI 0.09–0.13). The adjusted odds ratio for survival among patients transported to a hospital capable of delivering post-arrest care was 4.86 (95% CI 4.00–5.90).   Table 2.4 provides the results of a similar analysis with ROSC as the outcome. Older age was associated with lower ROSC rates in a dose-response fashion (AOR >75 years vs. 20–39 years 0.58, 95% CI 0.45–0.73). The association between response time and ROSC was unexpected and positive (AOR >16 vs. <6 minutes 1.67, 95% CI 1.07–2.62). Public location of arrest, bystander-witnessed arrest and ventricular fibrillation/tachycardia as the initial rhythm were associated with improved survival. ROC sites that were significantly associated with improved survival (i.e., sites 2, 6 and 7; Table 3) were associated with relatively lower rates of ROSC (Table 4). Bystander attempted CPR was not significantly associated with ROSC, whereas EMS CPR performed prior to rhythm analysis (AOR 1.35, 95% CI 1.14–1.60) and advanced airway placement (AOR 1.78, 95% CI 1.14–1.60) were significantly and positively associated with ROSC. Epinephrine use was negatively associated with ROSC (AOR 0.33, 95% CI 0.28–0.39) and receipt of specialized post-arrest care was very strongly associated with ROSC (AOR 3135, 95% CI 1667–5892).         Analyses restricted to cases who were transported to hospital with ongoing resuscitation significantly attenuated the effect of specialized post-arrest care on survival (adjusted odds ratio 1.64, 95% CI 1.30–2.06; Supplemental Table 2.2).     25 Table 2.5 shows the crude and adjusted odds ratio for survival among the sub-cohort of patients who received post-arrest care. A dose-response relationship was found between age and survival and between response time and survival. Public location of arrest (AOR 1.31, 95% CI 1.05–1.63), bystander-witnessed arrest (AOR 1.88, 95% CI 1.47–2.40) and ventricular fibrillation/tachycardia as the initial rhythm were associated with improved survival. Bystander-attempted resuscitation was associated with survival, but EMS CPR performed prior to rhythm analysis, and advanced airway placement were not significantly associated with survival. Epinephrine use was associated with a significant reduction in survival.   Analyses restricted to cases who did not receive post-arrest care also showed a clear dose-response relationship between age and survival and between response time and survival (Supplemental Table 2.3). Public location of arrest, bystander-witnessed arrest, and ventricular fibrillation/tachycardia as the initial rhythm were associated with improved survival. Specific ROC sites, which were associated with improved survival in the overall study population (i.e., sites 2 and 6), were also associated with lower survival when the analysis was restricted to cases that did not receive post-arrest care. Bystander-attempted resuscitation and EMS CPR performed prior to rhythm analysis were not significantly associated with survival, advanced airway placement was strongly associated with survival (AOR 3.11, 95% CI 2.20–4.39) and epinephrine use was strongly associated with lower survival (AOR 0.08, 95% CI 0.06–0.11).  Supplemental Table 2.4 shows adjusted odds ratios for survival among patients with response times <6 minutes who received post-arrest care. There was a dose-response relationship between age and survival. Public location of the arrest (AOR 1.35, 95% CI 1.04–1.75), bystander-  26 witnessed arrest (AOR 1.87, 95% CI 1.40–2.49), ventricular fibrillation/tachycardia as the initial rhythm and specific ROC sites (e.g., sites 2 and 7) were associated with improved survival. Bystander attempted CPR, EMS CPR prior to defibrillation analysis and advanced airway placement were not significantly associated with survival, while epinephrine use was negatively associated with survival (AOR 0.13, 95% CI 0.10–0.17).  In a final restricted analysis where we examined only patients who had a response time <6 minutes but did not receive post-arrest care, the age-relationship with survival was not significant (Supplemental Table 2.5). Public location of arrest, bystander-witnessed arrest, and ventricular fibrillation/tachycardia as the initial rhythm were associated with improved survival. Specific ROC sites which were associated with improved survival in the overall study population (Table 3) but negatively associated with survival in analysis restricted to cases that did not receive post-arrest care (Supplemental Table 2.3), continued to be associated with lower survival (Supplemental Table 2.5).     2.4 Discussion Our study examined the effects of chain-of-survival interventions and other factors on survival and ROSC among patients with a non-traumatic OHCA. We found younger age, shorter response times, public location of the arrest, bystander-witnessed arrest and ventricular fibrillation/tachycardia as the initial rhythm were consistently associated with improved survival. Use of epinephrine was strongly associated with lower survival, while bystander attempted CPR, EMS CPR prior to defibrillation analysis and advanced airway placement had variable associations with improved survival. The associations between some interventions and other   27 factors and ROSC presented a different pattern from the associations with survival: EMS CPR prior to defibrillation analysis was associated with improved ROSC rates but not with improved survival, and associations between response times and ROSC and between ROC sites and ROSC were reversed as compared with the same associations with survival. The overall results suggest a complex non-intuitive array of effects exists, particularly with regard to elements of the chain-of-survival that are considered to be ‘best practice’ interventions.  Interpreting our results requires a consideration of 3 issues, specifically the potential for survivor bias, confounding by indication and an understanding of the differences between the ROSC and survival associations. Since the decision to transport a patient to hospital for post-arrest care is heavily determined by the patient’s status and response to the initial resuscitative attempts, a substantial survivor bias was introduced in our non-experimental assessment of the effect of post-arrest care. We attempted to mitigate the effect of this bias by quantifying the effects of interventions and other factors separately among patients who had received post-arrest care and among patients who had not received post-arrest care. This stratification addressed survivor bias but precluded a quantification of the effect of post-arrest care.  The strong and consistent finding of a negative effect of epinephrine may suggest confounding by indication. Cases with an early response to resuscitation and early ROSC are less likely to be administered epinephrine, whereas cases showing no response are more likely to receive this intervention. The 9-fold reduction in survival among those administered epinephrine (AOR 0.11, 95% CI 0.09–0.13) compared with the 3-fold reduction in ROSC among patients who received epinephrine (AOR 0.33, 95% CI 0.28–0.39) indicates that a substantial fraction of patients who   28 received epinephrine and achieved ROSC in fact did not survive. We attempted to address such confounding by indication by restricting our analyses to patients with a response time of <6 minutes but this did not change our results.   Bystander-attempted resuscitation clearly has a beneficial effect, being associated with improved survival in the analysis involving all subjects and among the restricted sub-cohort of patients who received post-arrest care. Conversely, EMS CPR performed prior to rhythm analysis was not associated with improved survival in analyses involving all patients, and negatively associated with survival among patients who did not receive post-arrest care. Also, EMS CPR performed prior to rhythm analysis was positively associated with ROSC among all patients. This suggests that the intervention has a transient effect that results in ROSC but not necessarily in improved survival.  Advanced airway placement was associated with improved survival among all patients (68% increase in survival, Table 2.3), among patients transported to hospital with ongoing resuscitation (50% increase in survival, Table 2.4) and among patients who were not transported to a hospital capable of providing post-arrest care (211% improvement in survival). Additionally, advanced airway placement was associated with a 78% increase in ROSC. This consistency between effects on ROSC and survival and the difference in the effect among patient subgroups suggests that advanced airway placement consolidates early resuscitative efforts and may reduce need for post-arrest care.     29 The strengths of our study include a large study size and a population-based cohort. Information was collected through a meticulous system with time stamps and data consistency checks. The study also represents a comprehensive examination of the effects of interventions in the chain-of-survival. Although the effects of some interventions may have been affected by confounding by indication, the plausible effects associated with other factors such as age and response times lend credibility to the findings. Our study represents a careful and comprehensive examination of the effects of chain-of-survival interventions and other factors on ROSC and survival following out-of-hospital cardiac arrest.  Limitations of our study include the non-experimental nature of the design. Effects of interventions are ideally assessed through randomized trials since confounding by indication can bias the evaluation of interventions when the use of that intervention is predicated on the severity of the patient’s condition. We attempted to address this issue by restricting analysis to patients with a response time <6 minutes but our results showed that the effect of epinephrine remained unchanged. The age of our data (2005 – 2007) is a potential limitation, particularly if changes in practice over in the past decade altered the relationship between interventions and outcomes. This appears unlikely given the absence of significant changes in resuscitation algorithms. Our methods provide a robust analysis of practice during the study period, and also provide a template for conducting similar analyses on contemporary data. Other limitations include the retrospective nature of our study design and some lack of detail regarding the sequence of interventions.     30 2.5 Conclusion In summary, our study evaluated the effects of chain-of-survival interventions and other factors associated with out-of-hospital cardiac arrest. We found bystander resuscitation has a positive effect on survival, EMS CPR performed prior to rhythm analysis has a positive effect on ROSC but not survival, and advanced airway placement had a beneficial effect on both ROSC and survival. The effect of some interventions such as post-arrest care could not be evaluated because of survivor bias. In addition to patient characteristics, therapies by laypersons, first responder, and advanced EMS providers each contributed to the chances of survival, supporting the links-in-the-chain model used to conceptualize successful OHCA resuscitation. Epinephrine administration was associated with a decrease in ROSC and a substantial decrease in survival. Further study is required to distinguish ‘best practices’ in OHCA resuscitation and identify reasons for differences in survival across patient populations.  31 Figure 2.1 Resuscitation Outcomes Consortium pre-hospital time record web entry page.      32 Figure 2.2 Subject enrolment details.                       Excluded from analysis: 1. Non-study ROC unit responded to the call:  n=1,277 2. EMS witnessed: n=1,492 3. Resuscitation not attempted:            n=11,392 4. Age <20 years: n=1,236  Missing information on outcomes = 129 Missing information on  Age & Sex = 7  Age = 84 Sex = 17 Missing service level = 29  Available for Analysis Report by ROC site 1 = 996 2 = 1,435 3 = 2,268 4 = 237 5 = 524 6 = 1,777 7 = 942 8 = 595  9 = 1,315 10 = 2,732  Study cohort n=12,821 Patients assessed for cardiac arrest by a participating ROC agency: n=28,484   33 Table 2.1 Study subject demographic and clinical characteristics (n=12,821).  Characteristic Number (%)* Age in years, median (IQR) 66.0 (25.0) Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   927 (7.2) 4,046 (31.6) 3,883 (30.3) 3,965 (30.9)  Patient sex    Male   Female   8,366 (65.3) 4,455 (34.7) Location of arrest   Non-public   Public      10,734 (83.7) 2,078 (16.2)  Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander      Unknown  5,359 (41.8) 6,303 (49.2) 1,159 (9.0)  Cause of arrest assumed cardiac Other, obvious cause  11,882 (92.7) 939 (7.3)  Initial rhythm (EMS**)   VF/VT***   PEA^   Asystole   AED^^ – no shock advised, cannot                 determine, missing   3,094 (24.1) 2,287 (17.8) 5,396 (42.1) 2,044 (15.9)  ROC Site N (%)   1   2   3   4   5   6   7    8     9   10  996 (7.8) 1,435 (11.2) 2,268 (17.7) 237 (1.8) 524 (4.1) 1,777 (13.9) 942 (7.3) 595 (4.9) 1,315 (10.3) 2,732 (21.3)    34 Response time   Median (IQR)   5:00 (2:00) Response time   < 6 minutes   6 – 9 minutes  10 – 15 minutes  ≥ 16 minutes  Unknown   7,804 (60.9) 4,076 (31.8) 676 (5.3) 138 (1.1) 127 (1.0) Bystander-attempted resuscitation      No bystander resuscitation attempted     Bystander resuscitation attempted   8,309 (64.8) 4,512 (35.2) EMS CPR performed prior to rhythm analysis    Yes    No    10,555 (82.3) 2,266 (17.7) Advanced airway placed    Yes    No   10,285 (80.2) 2,536 (19.8) Prehospital epinephrine administered    Yes    No   9,804 (76.5) 3,017 (23.5) Specialized post-arrest care available^*   Yes   No  4,681 (36.5) 8,140 (63.5)   * Unless otherwise specified. **Emergency Medical Services (EMS)  ***VF/VT (ventricular fibrillation/ventricular tachycardia)  ^PEA (pulseless electrical activity)  ^^AED (automatic external defibrillator) ^*Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.   35  Table 2.2. Survival by demographic and clinical characteristics (n=12,821).  Characteristic Patient Survived n (%)* = 972 (7.6) Patient did not survive n (%)* = 11,849 (92.4)  p-value Patient age  Years median (IQR)   60.0 (21.0)  67.0 (24.0)   < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   86 (9.3) 431 (10.7) 303 (7.8) 152 (3.8)  841 (90.7) 3,615 (89.3) 3,580 (92.2) 3,813 (96.2)    < 0.001 Patient sex   Male   Female   698 (8.3) 274 (6.2)  7,668 (91.7) 4,181 (274)  < 0.001 Location of arrest   Non-public   Public     587 (5.5) 385 (18.5)  10,148 (94.5) 1,693 (81.5)  < 0.001 Bystander-witnessed status     Witnessed by bystander     Not witnessed by bystander      Unknown  703 (13.1) 218 (3.5) 51 (4.4)   4,656 (86.9) 6,085 (96.5%) 1,108 (95.6)   < 0.001 Cause of arrest      Assumed cardiac     Other, obvious cause  896 (7.5) 76 (8.1)  10,986 (92.5) 863 (91.9)   0.54 36       Initial rhythm (EMS**)    VF/VT***   PEA^   Asystole   AED^^ – no shock advised/cannot determine   624 (20.2) 128 (5.6) 58 (1.1) 162 (7.9)   2,470 (79.8) 2,159 (94.4) 5,338 (98.9) 1,882 (92.1)    < 0.001 ROC Site n (%)   1   2   3   4   5   6   7    8     9   10  114 (11.4) 222 (15.5) 97 (4.3) 5 (2.1) 44 (8.4) 168 (9.5) 78 (8.3) 49 (8.2) 58 (4.4) 137 (5.0)   882 (88.6) 1,213 (84.5) 2,171 (95.7) 232 (97.9) 480 (91.6) 1,609 (90.5) 864 (91.7) 546 (91.8) 1,257 (95.6) 2,595 (95.0)       < 0.001 Response time   Median (IQR)   4:00 (3.00)  5:00 (2:00)  < 0.001    Response time    <6 minutes 6 – 9 minutes 10 – 15 minutes >16 minutes Unknown   705 (9.0) 228 (5.6) 27 (4.0) 3 (2.2) 9 (7.1)  7,099 (91.0) 3,848 (94.4) 649 (96.0) 135 (97.8) 118 (92.9)   < 0.001  Bystander-attempted resuscitation     No bystander-attempted resuscitation  464 (5.6)  7,845 (94.4)  < 0.001 37     Bystander CPR attempted  508 (11.3) 4,004 (88.7)  EMS CPR performed prior to rhythm analysis    Yes    No   781 (7.4) 191 (8.4)  9,774 (92.6) 2,075 (91.6)  < 0.001 Advanced airway placed ^^^    Yes    No   756 (7.4) 216 (8.5)  9,529 (92.6) 2,320 (91.5)  0.05 Prehospital epinephrine administered    Yes    No   433 (4.4) 539 (17.9)   9,371 (95.6) 2478 (82.1)  < 0.001 Specialized post-arrest care available ^*   Yes   No  680 (14.5) 292 (3.6)   4,001 (92.4) 7,848 (96.4)  < 0.001  * Unless otherwise specified. **Emergency Medical Services (EMS)  ***VF/VT (ventricular fibrillation/ventricular tachycardia)  ^PEA (pulseless electrical activity)  ^^AED (automatic external defibrillator) ^^^ Supraglottic airway was used 1,055 times, tracheal intubation 9,588 times, cricothyrotomy 16 times, and rapid sequence intubation 35 times. Of the cases that had an advanced airway, 96% received only one type of advance airway.  The majority of instances in which rapid sequence intubation and cricothyrotomy were performed had two or more advanced airway interventions were carried out. Data not shown. ^*Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab. 38  Table 2.3. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and survival (n=12,821).  Characteristic Unadjusted  Adjusted OR (95% CI) p-value  OR (95% CI) p-value Bystander-attempted resuscitation     Bystander resuscitation attempted    No bystander resuscitation attempted    2.15 (1.19 – 2.45) Reference  < 0.001  1.30 (1.10 – 1.55) Reference  0.003 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis   0.87 (0.74 – 1.02) Reference   0.09   0.81 (0.65 – 1.01) Reference   0.06 Advanced airway placed     Yes    No      0.85 (0.73 – 1.00) Reference  0.05  1.68 (1.34 – 2.12) Reference  < 0.001 Prehospital epinephrine administered    Yes    No    0.21 (0.19 – 0.24) Reference  < 0.001  0.11 (0.09 – 0.13) Reference  < 0.001 Specialized post-arrest care available*   Yes    No  4.57 (3.96 – 5.27) Reference  < 0.001  4.86 (4.00 – 5.91) Reference   < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)  Reference 1.17 (0.91 – 1.49) 0.83 (0.64 – 1.06) 0.39 (0.30 – 0.51)   0.22 0.14 < 0.001  Reference 0.85 (0.63 – 1.14) 0.59 (0.43 – 0.80) 0.35 (0.25 – 0.50)   0.27 0.001 < 0.001 39   Patient sex   Male   Female   1.39 (1.20 – 1.61) Reference  < 0.001  0.97 (0.81 – 1.16) Reference  0.77 Location of arrest   Non-public   Public     Reference 3.93 (3.43 – 4.52)   < 0.001   Reference 1.59 (1.34 – 1.90)   < 0.001  Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander     Unknown  4.22 (3.60 – 4.93) Reference 1.29 (0.94 – 1.75)  < 0.001  0.12  2.06 (1.71 – 2.49) Reference 1.14 (0.77 – 1.68)   < 0.001  0.50 Cause of arrest      Assumed cardiac      Other, obvious cause     0.93 (0.73 – 1.18) Reference  0.54  0.77 (0.57 – 1.06) Reference  0.11 Initial rhythm (EMS**)   VF/VT***   PEA^   Asystole   AED^^ – no shock advised/cannot determine   Reference 0.24 (0.19 – 0.29) 0.04 (0.03 – 0.06) 0.34 (0.28 – 0.41)   < 0.001 < 0.001 < 0.001  Reference 0.27 (0.22 – 0.35) 0.08 (0.06 – 0.12) 0.37 (0.30 – 0.47)   < 0.001 < 0.001 < 0.001 ROC Site    1   2   3   4  2.45 (1.89 – 3.18) 3.47 (2.77 – 4.34) 0.85 (0.65 – 1.10) 0.41 (0.17 – 1.01)  < 0.001 < 0.001 0.22 0.05  1.09 (0.78 – 1.53) 1.62 (1.20 – 2.17) 0.65 (0.48 – 0.89) 0.09 (0.03 – 0.23)  0.6 0.001 < 0.001 < 0.001 40    5   6   7    8     9   10 1.74 (1.22 – 2.47) 1.98 (1.57 – 2.50) 1.71 (1.28 – 2.28) 1.70 (1.21 – 2.39) 0.87 (0.64 – 1.20) Reference  0.002 < 0.001 < 0.001 0.002 0.40   0.69 (0.45 – 1.07) 1.53 (1.15 – 2.02) 1.65 (1.15 – 2.38) 0.95 (0.63 – 1.44) 0.40 (0.26 – 0.60) Reference  0.10 0.003 0.01 0.81 < 0.001  Response time  < 6 minutes 6 – 9 minutes 10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.60 (0.51 – 0.70) 0.42 (0.28 – 0.62) 0.22 (0.07 – 0.70) 0.77 (0.39 – 1.52)    < 0.001 < 0.001 0.01 < 0.45  Reference 0.74 (0.61 – 0.90) 0.52 (0.33 – 0.80) 0.30 (0.09 – 0.97) 1.35 (0.61 – 2.98)   0.001 0.003 0.04 0.46  *Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab. **Emergency Medical Services (EMS)  ***VF/VT (ventricular fibrillation/ventricular tachycardia)  ^PEA (pulseless electrical activity)  ^^AED (automatic external defibrillator)       41   Table 2.4. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and return of spontaneous circulation (ROSC) (n=12,281).  Characteristic Unadjusted Adjusted OR (95% CI) p-value  AOR (95% CI) p-value Bystander-attempted resuscitation     Bystander resuscitation attempted    No bystander resuscitation attempted    1.03 (0.96 – 1.11) Reference  0.43  0.98 (0.88 – 1.10) Reference  0.77 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis     0.60 (0.55 – 0.66) Reference   < 0.001   1.35 (1.14 – 1.60) Reference   0.001 Advanced airway placed    Yes    No      1.46 (1.34 – 1.60) Reference  < 0.001  1.78 (1.14 – 1.60) Reference  < 0.001 Prehospital epinephrine administered    Yes    No    0.87 (0.80 – 0.94) Reference  0.001  0.33 (0.28 – 0.39) Reference  < 0.001 Specialized post-arrest care available   Yes   No   638.5 (370 – 1,103) Reference  < 0.001  3134.6 (1,667 – 5,892) Reference   < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years)  Reference 1.05 (0.91 – 1.22)   0.51  Reference 0.79 (0.62 – 1.01)   0.06 42  Older (60 – 74 years) Elderly (≥ 75 years)  0.95 (0.82 – 1.10) 0.71 (0.62 – 0.83) 0.48 < 0.001 0.67 (0.53 – 0.85) 0.58 (0.45 – 0.73) 0.001 < 0.001 Patient sex   Male   Female   1.16 (1.08 – 1.25) Reference  < 0.001  1.01 (0.90 – 1.13) Reference  0.89  Location of arrest   Non-public   Public     Reference 3.19 (2.85 – 3.58)    < 0.001  Reference 2.24 (1.90 – 2.64)   < 0.001 Bystander-witnessed status Witnessed by bystander   Not witnessed by bystander   Unknown  2.52 (2.34 – 2.72) Reference 4.04 (3.49 – 4.69)   < 0.001  < 0.001  1.62 (1.44 – 1.81) Reference 1.18 (0.87 – 1.61)  < 0.001  0.28 Cause of arrest      Assumed cardiac     Other, obvious cause     1.20 (1.05 – 1.37) Reference  0.008  1.07 (0.87 – 1.33) Reference  0.51 Initial rhythm (EMS*)   VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot determine  Reference 0.59 (0.52 – 0.67) 0.22 (0.20 – 0.25) 0.58 (0.51 – 0.65)    < 0.001 < 0.001 < 0.001  Reference 0.65 (0.55 – 0.78) 0.26 (0.22 – 0.30) 0.48 (0.41 – 0.57)   < 0.001 < 0.001 < 0.001 ROC Site   1   2   3  1.22 (1.06 – 1.41) 0.81 (0.71 – 0.92) 1.42 (1.27 – 1.59)  0.007 0.001 < 0.001  0.22 (0.17 – 0.29) 0.02 (0.01 – 0.04) 1.06 (0.92 – 1.21)  < 0.001 < 0.001 0.42 43    4   5   6   7    8     9   10 5.33 (3.70 – 7.69) 5.36 (4.16 – 6.91) 1.01 (0.90 – 1.14) 0.49 (0.42 – 0.57) 1.80 (1.49 – 2.16) 23.87 (17.78 – 32.07) Reference < 0.001 < 0.001 0.81 < 0.001 < 0.001 < 0.001 0.73 (0.40 – 1.33) 1.10 (0.76 – 1.57) 0.71 (0.61 – 0.82) 0.05 (0.04 – 0.08) 0.31 (0.22 – 0.44) 2.41 (1.53 – 3.80) Reference  0.30 0.62 < 0.001 < 0.001 < 0.001 < 0.001 Response time     < 6 minutes    6 – 9 minutes    10 – 15 minutes    ≥ 16 minutes    Unknown   Reference 1.00 (0.93 – 1.08) 1.02 (0.87 – 1.20) 1.30 (0.91 – 1.85) 1.75 (1.19 – 2.59)   0.96 0.83 0.14 0.005  Reference 1.18 (1.05 – 1.32) 1.45 (1.18 – 1.80) 1.67 (1.07 – 2.62) 0.81 (0.45 – 1.47)   0.005 0.001 0.03 0.49     44  Table 2.5. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients transported to hospital capable of providing specialized post-arrest care* (n=4681).  Characteristic Unadjusted Adjusted  OR (95% CI) p-value  OR (95% CI) p-value Bystander-attempted resuscitation     Bystander resuscitation attempted    No bystander resuscitation attempted    2.39 (2.02 – 2.81) Reference  < 0.001  1.13 (1.01– 1.58) Reference  0.04 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis     1.38 (1.14 – 1.69) Reference    0.001   0.81 (0.61 – 1.07) Reference    0.14 Advanced airway placed    Yes    No      0.46 (0.37 – 0.57) Reference  < 0.001  0.82 (0.59 – 1.14) Reference  0.24 Prehospital epinephrine administered    Yes    No    0.13 (0.11 – 0.16) Reference  < 0.001  0.13 (0.10 – 0.17) Reference  < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   Reference 1.03 (0.77 – 1.37) 0.74 (0.55 – 0.99) 0.39 (0.28 – 0.54)    0.85 0.04 < 0.001  Reference 0.68 (0.47 – 0.98) 0.42 (0.29 – 0.62) 0.29 (0.19 – 0.43)   0.04 < 0.001 < 0.001  45  Patient sex   Male   Female   1.31 (1.10 – 1.56) Reference  0.003  0.96 (0.76 – 1.20) Reference  0.71 Location of arrest   Non-public   Public     Reference 2.74 (2.30 – 3.25)   < 0.001   Reference 1.31 (1.05 – 1.63)   0.02  Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander     Unknown  2.96 (2.43 – 3.60) Reference 0.63 (0.45 – 0.89)  < 0.001  0.008  1.88 (1.47 – 2.40) Reference 1.11 (0.72 – 1.72)   < 0.001  0.63 Cause of arrest      Assumed cardiac     Other, obvious cause      0.77 (0.58 – 1.04) Reference  0.09  0.73 (0.49 – 1.09) Reference  0.12 Initial rhythm (EMS*)    VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot                   determine   Reference 0.24 (0.19 – 0.31) 0.07 (0.05 – 0.10) 0.45 (0.36 – 0.57)   < 0.001 < 0.001 < 0.001  Reference 0.26 (0.19 – 0.34) 0.12 (0.09 – 0.17) 0.51 (0.38 – 0.69)   < 0.001 < 0.001 < 0.001 ROC Site n (%)   1   2   3   4   5  1.53 (1.03 – 2.26) 3.09 (2.15 – 4.42) 0.53 (0.34 – 0.94) 0.18 (0.07 – 0.46) 0.74 (0.47 – 1.17)  0.04 < 0.001 0.01 0.006 0.20  1.44 (0.89 – 2.33) 2.90 (1.87 – 4.51) 0.53 (0.31 – 0.88) 0.12 (0.04 – 0.32) 0.86 (0.50 – 1.50)  0.14 < 0.001 0.02 < 0.001 0.60 46    6   7    8     9   10 1.87 (1.22 – 2.86) 2.02 (1.34 – 3.06) 0.94 (0.60 – 1.47) 0.30 (0.20 – 0.46) Reference  0.004 0.001 0.77 < 0.001   1.48 (0.88 – 2.49) 2.82 (1.70 – 4.67) 1.16 (0.68 – 1.99) 0.41 (0.24 – 0.68) Reference  0.14 < 0.001 0.58 0.001  Response time  < 6 minutes 6 – 9 minutes 10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.57 (0.47 – 0.70) 0.26 (0.13 – 0.52) 0.25 (0.06 – 1.02) 0.62 (0.30 – 1.90)    < 0.001 < 0.001 0.053 0.21  Reference 0.72 (0.56 – 0.92) 0.32 (0.14 – 0.74) 0.46 (0.10 – 2.06) 1.48 (0.62 – 3.55)   0.01 0.01 0.31 0.38  * Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.   47  Supplemental Table 2.1. Return of spontaneous circulation (ROSC) rates by demographic and clinical characteristics (n=12,821).  Characteristic     Return of Spontaneous Circulation n (%) = 7598 (59.3) No Return of Spontaneous Circulation n (%) = 5223 (40.7)  p-value Bystander-attempted resuscitation n (%)     No bystander resuscitation attempted     Bystander resuscitation attempted   4,903 (59.0) 2,695 (59.7)  3,406 (41.0) 1,817 (40.3)  0.43 EMS CPR performed prior to rhythm analysis    Yes    No   6,035 (57.2) 1,563 (69.0)   4,520 (42.8) 703 (31.0)  < 0.001 Advanced airway placed    Yes    No   6,285 (61.1) 1,313 (51.8)  4,000 (38.9) 1,223 (48.2)  < 0.001 Prehospital epinephrine administered    Yes    No   5,732 (58.5) 1,866 (61.8)  4,072 (41.5) 1,151 (38.2)  0.001 Specialized post-arrest care available   Yes   No  4,668 (99.7) 2,930 (36.0)   13 (0.3) 5210 (64.0)  < 0.001 Patient age  Years median (IQR)  65.0 (24.0)  69.0 (24.0)   < 0.001 48     Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   573 (61.8) 2,548 (63.0) 2,351 (60.5) 2,126 (53.6)  354 (38.2) 1,498 (37.0) 1,532 (39.5) 1,839 (46.4)   < 0.001 Patient sex   Male   Female   5,046 (60.5) 2,534 (56.9)  3,302 (39.5) 1,921 (43.1)  < 0.001  Location of arrest   Non-public   Public     5,934 (55.3) 1,658 (79.8)  4,800 (44.7) 420 (20.2)  < 0.001 Bystander-witnessed status     Witnessed by bystander     Not witnessed by bystander     Unknown  2,977 (47.2) 3,713 (69.3) 908 (78.3)   3,326 (52.8) 1,646 (30.7) 251 (21.7)   < 0.001 Cause of arrest      Assumed cardiac      Other, obvious cause    7,080 (59.6) 518 (55.2)   4,802 (40.4) 421 (44.8)    0.01 Initial rhythm (EMS*)    VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot determine   2,392 (77.3) 1,528 (66.8) 2,325 (43.1) 1,353 (66.2)  702 (22.7) 759 (33.2) 3071 (56.9) 691 (33.8)   < 0.001 49  ROC Site      1   2   3   4   5   6   7    8     9   10  567 (56.9) 671 (46.8) 1375 (60.6) 202 (85.2) 447 (85.3) 930 (52.3) 327 (34.7) 393 (66.1) 1,266 (96.3) 1,420 (52.0)   429 (43.1) 764 (53.2) 893 (39.4) 35 (14.8) 77 (14.7) 847 (47.7) 615 (65.3) 202 (33.9) 49 (3.7) 1,312 (48.0)       < 0.001  Response time   Median (IQR)   5.0 (2.0)  5.0 (2.0)  0.57  Response time  < 6 minutes 6 – 9 minutes 10 – 15 minutes ≥ 16 minutes Unknown   4,607 (59.0) 2,408 (59.1) 402 (59.2) 90 (65.2) 91 (71.7)  3,197 (41.0) 1,668 (40.9) 274 (40.5) 48 (34.8) 36 (28.3)    0.04  * BLS (basic life support) ** ALS (advanced life support) ^ BMV (bag-mask ventilation) ^^CPR (cardiopulmonary resuscitation) *^ AED (automatic external defibrillator) **^ EMS (emergency medical services)   50  Supplemental Table 2.2. Logistic regression model (restricted cohort) showing unadjusted and adjusted association between interventions and other factors and survival among patients transported to hospital with ongoing resuscitation or ROSC (n=7612).  Characteristic Unadjusted  Adjusted  OR (95% CI) p-value  OR (95% CI) p-value Bystander-attempted resuscitation     Bystander resuscitation attempted    No bystander resuscitation attempted    2.22 (1.94 – 2.55) Reference  < 0.001  1.26 (1.06 – 1.51) Reference  0.01 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis     1.07 (0.90 – 1.27) Reference    0.43   0.80 (0.64 – 1.00) Reference    0.05 Advanced airway placed    Yes    No      0.69 (0.59 – 0.82) Reference  < 0.001  1.50 (1.19 – 1.89) Reference  0.001 Prehospital epinephrine administered    Yes    No    0.20 (0.18 – 0.23) Reference  < 0.001  0.13 (0.11 – 0.16) Reference  < 0.001 Specialized post-arrest care available   Yes   No   1.54 (1.33 – 1.78) Reference  < 0.001  1.64 (1.30 – 2.06) Reference  < 0.001 Patient age  Adult (20 – 39 years)  Reference    Reference   51  Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)  1.15 (0.89 – 1.48) 0.84 (0.65 – 1.08) 0.44 (0.33 – 0.58)  0.28 0.17 < 0.001 0.86 (0.64 – 1.17) 0.60 (0.43 – 0.82) 0.37 (0.26 – 0.52) 0.35 0.002 < 0.001 Patient sex   Male   Female   1.32 (1.14 – 1.53) Reference  < 0.001  1.01 (0.84 – 1.21) Reference  0.92 Location of arrest   Non-public   Public     Reference 2.75 (2.39 – 3.17)   < 0.001   Reference 1.49 (1.25 – 1.78)   < 0.001  Bystander-witnessed status     Witnessed by bystander     Not witnessed by bystander      Unknown  2.96 (2.52 – 3.47) Reference 0.75 (0.55 – 1.03)  < 0.001  0.08  1.84 (1.52 – 2.23) Reference 1.07 (0.73 – 1.58)   < 0.001  0.73 Cause of arrest      Assumed cardiac      Other, obvious cause     0.84 (0.65 – 1.08) Reference  0.18  0.71 (0.52 – 0.98) Reference  0.04 Initial rhythm (EMS*)    VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot determine   Reference 0.26 (0.21 – 0.32) 0.07 (0.06 – 0.10) 0.39 (0.32 – 0.47)   < 0.001 < 0.001 < 0.001  Reference 0.29 (0.23 – 0.36) 0.12 (0.09 – 0.16) 0.41 (0.33 – 0.52)   < 0.001 < 0.001 < 0.001 ROC Site    1  2.36 (1.80 – 3.09)  < 0.001  2.04 (1.43 – 2.92)  < 0.001 52    2   3   4   5   6   7    8     9  10 4.63 (3.65 – 5.87) 0.71 (0.54 – 0.93) 0.24 (0.10 – 0.58) 1.02 (0.71 – 1.46) 2.07 (1.62 – 2.64) 2.94 (2.16 – 4.00) 1.33 (0.94 – 1.88) 0.45 (0.33 – 0.62) Reference  < 0.001 0.01 0.002 0.92 < 0.001 < 0.001 0.11 < 0.001   3.70 (1.43 – 2.92) 0.79 (0.58 – 1.07) 0.17 (0.06 – 0.43) 1.16 (0.74 – 1.81) 1.86 (1.40 – 2.48) 3.75 (2.50 – 5.63) 1.74 (1.12 – 2.70) 0.64 (0.42 – 0.99) Reference  < 0.001 0.12 < 0.001 0.52 < 0.001 < 0.001 0.01 0.04  Response time category < 6 minutes 6 – 9 minutes 10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.58 (0.50 – 0.68) 0.40 (0.23 – 0.60) 0.19 (0.06 – 0.61) 0.60 (0.30 – 1.20)   < 0.001 < 0.001 0.005 0.15  Reference 0.73 (0.60 – 0.88) 0.43 (0.28 – 0.68) 0.25 (0.08 – 0.84) 1.38 (0.63 – 3.01)   0.001 < 0.001 0.03 0.42    53  Supplemental Table 2.3. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients who were NOT transported to a hospital capable of providing specialized post-arrest care* (n=8140).  Characteristic Unadjusted Adjusted  OR (95% CI) p-value  OR (95% CI) p-value Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted    1.94 (1.53 – 2.45) Reference  < 0.001  1.19 (0.89 – 1.58) Reference  0.24 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis     0.73 (0.53 – 1.00) Reference    0.048   0.73 (0.49 – 1.09) Reference    0.12 Advanced airway placed    Yes    No      0.80 (0.62 – 1.04) Reference  0.10  3.11 (2.20 – 4.39) Reference  < 0.001 Prehospital epinephrine administered    Yes    No    0.19 (0.15 – 0.24) Reference  < 0.001  0.08 (0.06 – 0.11) Reference  < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   Reference 1.85 (1.07 – 3.19) 1.53 (0.89 – 2.66) 0.69 (0.39 – 1.24)    0.03 0.13 0.22  Reference 1.30 (0.71 – 2.38) 0.96 (0.52 – 1.78) 0.51 (0.27 – 0.98)   0.40 0.90 0.04 54  Patient sex   Male   Female   1.68 (1.28 – 2.20) Reference  < 0.001  1.09 (0.80 – 1.49) Reference  0.59 Location of arrest   Non-public   Public     Reference 4.77 (3.73 – 6.10)   < 0.001   Reference 2.30 (1.70 – 3.05)   < 0.001  Bystander-witnessed status n (%)    Witnessed by bystander     Not witnessed by bystander      Unknown  4.71 (3.59 – 6.18) Reference 0.92 (0.37 – 2.30)  < 0.001  0.86  2.02 (1.48 – 2.76) Reference 0.62 (0.23 – 1.71)   < 0.001  0.36 Cause of arrest      Assumed cardiac n (%)     Other, obvious cause     1.25 (0.77 – 2.02)  0.37  0.81 (0.46 – 1.41)  0.45 Initial rhythm (EMS*) n (%)   VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot determine   Reference 0.21 (0.14 – 0.30) 0.03 (0.02 – 0.05) 0.25 (0.18 – 0.35)   < 0.001 < 0.001 < 0.001  Reference 0.34 (0.22 – 0.50) 0.06 (0.03 – 0.10) 0.22 (0.15 – 0.33)   < 0.001 < 0.001 < 0.001 ROC Site n (%)   1   2   3   4   5   6  1.04 (0.64 – 1.68) 0.10 (0.03 – 0.31) 0.81 (0.58 – 1.14) 0.00  0.70 (0.25 – 1.92) 1.86 (1.40 – 2.47)  0.88 < 0.001 0.23 1.00 0.49 < 0.001  1.37 (0.77 – 2.42) 0.13 (0.04 – 0.43) 0.85 (0.58 – 1.24) 0.00 0.75 (0.25 – 2.29) 1.78 (1.26 – 2.52)  0.28 0.001 0.39 1.00 0.61 0.001 55    7    8     9   10 0.16 (0.06 – 0.43) 0.40 (0.15 – 1.11) 0.50 (0.12 – 2.06) Reference  < 0.001 0.08 0.34   0.35 (0.12 – 1.00) 0.97 (0.32 – 2.95) 1.41 (0.29 – 6.84) Reference  0.049 0.96 0.67  Response time  < 6 minutes 6 – 9 minutes 10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.86 (0.66 – 1.11) 0.96 (0.60 – 1.56) 0.27 (0.04 – 1.91) 0.46 (0.06 – 3.33)    0.24 0.88 0.19 0.44  Reference 0.77 (0.57– 1.03) 0.60 (0.35 – 1.04) 0.19 (0.02 – 1.45) 0.72 (0.09 – 5.93)   0.08 0.07 0.11 0.80  * Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.         56  Supplemental Table 2.4. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients with a response time of less than 6 minutes who were transported to hospital capable of providing specialized post arrest care (n=3094).  Characteristic Unadjusted   Adjusted   OR (95% CI) p-value  OR (95% CI) p-value Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted    2.20 (1.18 – 2.65) Reference  < 0.001  1.13 (0.87 – 1.47) Reference  0.35 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis     1.45 (1.14 – 1.83) Reference    0.002   0.88 (0.63 – 1.23) Reference    0.45 Advanced airway placed    Yes    No      0.41 (0.32 – 0.53) Reference  < 0.001  0.78 (0.52 – 1.16) Reference  0.22 Prehospital epinephrine administered    Yes    No    0.12 (0.10 – 0.15) Reference  < 0.001  0.13 (0.10 – 0.17) Reference  < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   Reference 1.13 (0.81 – 1.57) 0.76 (0.54 – 1.07) 0.37 (0.26 – 0.57)    0.47 0.11 < 0.001  Reference 0.72 (0.47 – 1.10) 0.40 (0.25 – 0.62) 0.28 (0.17 – 0.45)   0.13 < 0.001 < 0.001 57  Patient sex   Male   Female   1.42 (1.16 – 1.75) Reference  0.001  1.03 (0.79 – 1.34) Reference  0.84 Location of arrest   Non-public   Public     Reference 2.79 (2.29 – 3.40)   < 0.001   Reference 1.35 (1.04 – 1.75)   0.02  Bystander-witnessed status n (%)    Witnessed by bystander     Not witnessed by bystander     Unknown  2.95 (2.35 – 3.71) Reference 0.69 (0.47 – 1.01)  < 0.001  0.06  1.87 (1.40 – 2.49) Reference 1.06 (0.65 – 1.73)   < 0.001  0.82 Cause of arrest      Assumed cardiac n (%)     Other, obvious cause     0.76 (0.54 – 1.06)  0.11  0.69 (0.43 – 1.08)  0.11 Initial rhythm (EMS*) n (%)   VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot                                        determine   Reference 0.23 (0.17 – 0.31) 0.07 (0.05 – 0.11) 0.48 (0.37 – 0.63)   < 0.001 < 0.001 < 0.001  Reference 0.23 (0.16 – 0.32) 0.13 (0.08 – 0.19) 0.52 (0.36 – 0.75)   < 0.001 < 0.001 < 0.001  ROC Site n (%)   1   2   3   4   5   6  1.73 (1.02 – 2.91) 3.24 (1.97 – 5.33) 0.66 (0.36 – 1.19) 0.22 (0.06 – 0.75) 0.80 (0.45 – 1.45) 2.15(1.20 – 3.83)  0.04 < 0.001 0.17 0.02 0.46 0.01  2.19 (1.16 – 4.13) 4.03 (2.21 – 7.36) 0.88 (0.44 – 1.74) 0.20 (0.06 – 0.75) 1.36 (0.68 – 2.72) 1.84 (0.91 – 3.70)  0.02 < 0.001 0.71 0.02 0.39 0.09 58    7    8     9   10 2.05 (1.19 – 3.52) 1.61 (0.89 – 2.90) 0.40 (0.23 – 0.68) Reference  0.01 0.12 0.001   4.15 (2.16 – 7.99) 1.93 (0.94 – 3.94) 0.68 (0.35 – 1.34) Reference  < 0.001 0.07 0.27    59  Supplemental Table 2.5. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients with a response time of less than 6 minutes who were NOT transported to a hospital capable of providing specialized post-arrest care (n=4710).  Characteristic Unadjusted Adjusted  OR (95% CI) p-value  OR (95% CI) p-value Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted    1.62 (1.20 – 2.19) Reference  0.002  0.96 (0.66 – 1.39) Reference  0.82 EMS CPR performed prior to rhythm analysis    CPR prior to rhythm analysis    CPR not prior to rhythm analysis    0.57 (0.39 – 0.83) Reference   0.004  0.56 (0.34– 0.93) Reference    0.03 Advanced airway placed    Yes    No      0.77 (0.55 – 1.07) Reference  0.12  3.48 (2.22 – 5.46) Reference  < 0.001 Prehospital epinephrine administered    Yes    No   0.16 (0.13 – 0.22) Reference  < 0.001  0.07 (0.05 – 0.11) Reference  < 0.001 Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74 years) Elderly (≥ 75 years)   Reference 1.63 (0.83 – 3.20) 1.56 (0.79 – 3.06) 0.68 (0.33 – 1.39)    0.15 0.20 0.29  Reference 1.08 (0.49 – 2.35) 0.92 (0.42 – 2.01) 0.48 (0.21 – 1.10)    0.86 0.83 0.08 Patient sex     60    Male   Female  1.72 (1.22 – 2.42) Reference 0.002 1.01 (0.67 – 1.51)  Reference 0.97 Location of arrest   Non-public   Public     Reference 5.33 (3.92 – 7.27)   < 0.001   Reference 2.38 (1.63 – 3.46)    < 0.001  Bystander-witnessed status n (%)    Witnessed by bystander     Not witnessed by bystander     Unknown  5.12 (3.62 – 7.24) Reference 1.07 (0.38 – 2.99)  < 0.001  0.90  2.06 (1.38 – 3.10) Reference 0.65 (0.20 – 2.12)   < 0.001  0.47 Cause of arrest      Assumed cardiac n (%)     Other, obvious cause     1.38 (0.72 – 2.64) Reference  0.33  0.94 (0.45 – 1.98) Reference  0.87 Initial rhythm (EMS*) n (%)   VF/VT**   PEA^   Asystole   AED^^ – no shock advised/cannot   determine   Reference 0.18 (0.11 – 0.29) 0.03 (0.02 – 0.06) 0.26 (0.17 – 0.41)   < 0.001 < 0.001 < 0.001  Reference 0.29 (0.17 – 0.50) 0.07 (0.03 – 0.13) 0.22 (0.13 – 0.37)   < 0.001 < 0.001 < 0.001  ROC Site n (%)   1   2   3   4   5   6   7   0.96 (0.55 – 1.67) 0.12 (0.04 – 0.39) 0.76 (0.50 – 1.12) 0.00 0.43 (0.10 – 1.80) 2.00 (1.37 – 2.93) 0.12 (0.04 – 0.38)  0.87 < 0.001 0.21 1.00 0.25 < 0.001 < 0.001  1.27 (0.67 – 2.52) 0.19 (0.05 – 0.64) 0.79 (0.49 – 1.29) 0.00 0.42 (0.09 – 1.98) 1.73 (1.09 – 2.76) 0.31 (0.09 – 1.04)  0.44 0.008 0.35 1.00 0.27 0.02 0.06 61    8     9   10 0.14 (0.02– 0.99) 0.58 (0.14 – 2.41) Reference  0.048 0.45   0.28 (0.03 – 2.51) 1.31 (0.25– 6.83) Reference  0.25 0.75     62  Chapter 3: Survival following non-traumatic out-of-hospital cardiac arrest: Do prehospital clinical interventions modify the prognostic effect of initial rhythm? 3.1 Summary Survival following out-of-hospital cardiac arrest (OHCA) remains poor and the effects of individual clinical interventions require further evaluation. We examined the effects of standard interventions on survival specifically to assess potential modification of the effect of the initial rhythm by epinephrine administration and advanced airway placement.   We carried out a cohort study of adult (>19 years) non-traumatic OHCA cases and quantified the associations between patient and emergency medical services (EMS) characteristics, and interventions including advanced airway placement, administration of epinephrine, and specialized post-arrest care on return of spontaneous circulation (ROSC) and survival. Restriction and logistic regression were used to estimate adjusted odds ratios (AOR) and 95% confidence intervals (CI) expressing the effect of interventions and other factors on survival, and potential modification of the effect of initial rhythm was tested using interaction terms.  The study population included 14,673 cases of non-traumatic cardiac arrest, of whom 3,713 (25.3%) had ROSC in the Emergency Department and 1,122 (7.6%) survived at least 6 months after hospital discharge. Factors associated with return of spontaneous 63  circulation were bystander-witnessed arrest, bystander-attempted resuscitation, and initial rhythm. Factors associated with survival were younger age, public arrest location, bystander-witnessed arrest, faster EMS response time, and initial rhythm. Advanced airway placement was not associated with survival, and administration of prehospital epinephrine was associated with lower survival (AOR 0.11, 95% CI 0.09–0.14). Among patients who received specialized post-arrest care, epinephrine administration modified the effect of initial rhythm: receipt of epinephrine was associated with a 14-fold decrease in the adjusted odds of survival if the patient was in asystole, while epinephrine administration was associated with a 5.8-fold decrease in the adjusted odds of survival if the case was in ventricular fibrillation (p-value for interaction <0.01).  Younger age, public location, bystander-witnessed arrest, faster EMS response time, and ventricular fibrillation/tachycardia (as an initial rhythm) positively influence survival following out-of-hospital cardiac arrest, whereas epinephrine administration is negatively associated with survival. Epinephrine administration also modified the relationship between initial rhythm and survival by substantially decreasing survival among those with asystole and moderately decreasing survival among those with ventricular fibrillation/tachycardia.   3.2 Introduction  Emergency medical services (EMS) systems were designed to respond to and treat life threatening injuries and illness, including cardiac arrest. In the initial development of EMS systems, much of the focus was on treating cardiac arrest. The chain-of-survival, a 64  5-mode framework of interconnected community and prehospital clinical interventions intended to improve survival from out-of-hospital cardiac arrest (OHCA), developed in parallel with EMS systems and has informed and supported much of the organization of care and treatment for OHCA over the past 40 years [48, 49]. However, evidence for the effectiveness of the clinical components in the chain-of-survival remains unclear.    In previous work conducted by our group [67, 68], we confirmed the beneficial survival effects of younger age, public location, bystander witness, shorter response times, and bystander-attempted resuscitation in OHCA. Additionally, we found a beneficial effect of advanced airway management (defined as interventions which include supraglottic airways, and tracheal intubation). However we were unable to assess the effect of specialized post-arrest care due to survivor bias, and our assessment of the effect of epinephrine administration was potentially compromised by confounding by indication [69]. Controversy continues to surround the role of both epinephrine [34, 36, 70, 71] and advanced airway use [26, 72-74], and recently launched randomized trials may help address this [75].    As efforts to improve survival following out-of-hospital (OHCA) cardiac arrest survival continue, it behooves EMS system designers and administrators to have clarity regarding which parts of the system are beneficial. The purpose of this study was to determine whether the standard clinical interventions including epinephrine and advanced airway placement modify the effect of initial (first documented EMS) cardiac rhythm and survival. 65   3.3 Methods  We carried out a secondary analysis of adult (>19 years) non-traumatic OHCA cases who were initially enrolled in a Resuscitation Outcomes Consortium (ROC) trial studying the effect of an impedance valve and early vs. delayed rhythm analysis called ROC PRIMED [76, 77]. This study was carried out in 11 academic coordinating centers across the United States and Canada [61].   3.3.1 Population  OHCA cases were obtained from a population of approximately 25 million people, predominantly in urban and metropolitan centers, in seven US states and two Canadian provinces. Over 18,000 OHCA cases were screened and 14,673 were enrolled.   3.3.2 Determinants and outcomes  The determinants of interest included patient characteristics (gender, age, and initial rhythm), EMS system and community factors (public location of arrest, bystander attempt at resuscitation, and EMS response time) and interventions (epinephrine, advanced airway placement, and provision of specialized post-arrest care). The primary outcome was survival at 6-month follow up after hospital discharge. Return of spontaneous circulation (ROSC) in the emergency department was a secondary outcome. Initial rhythm was established via EMS defibrillator or automatic external defibrillator (AED) tracings and recorded as ventricular fibrillation/ventricular tachycardia (VF/VT), pulseless electrical activity (PEA), asystole, perfusing, “AED-no shock, No-strip”, and 66  ‘cannot determine’. The initial rhythm was a composite that included an EMS rhythm recorded approximately ten seconds after pad placement or at the time of the first EMS shock analysis, depending on the trial arm to which the case was randomized.  3.3.3 Analytic methods  Descriptive statistics including frequencies, proportions, means (standard deviation, SD), and medians (interquartile range, IQR) were calculated. Logistic regression was used to assess the relationship between clinical interventions and other factors and survival to six months after hospital discharge. Potential confounders were included in the regression model based on their effect on relationships with prehospital interventions and survival status: a change of  >10% in the association between an intervention and survival was deemed a priori to be evidence of confounding [65]. Variables assessed as potential confounders encompassed standard Utstein variables [66], including age, sex, arrest location, cause of the arrest, and initial rhythm, and additional system level descriptors including ROC site. Associations were quantified using adjusted odds ratios (AOR) with 95% confidence intervals (CI) and 2-sided p-values.  The initial analytic approach involved the entire study cohort. However, our previous work and initial analyses showed that transport to hospital and provision of specialized post-arrest care (defined as receipt of hypothermia treatment, fibrinolysis or diagnostic cardiac catheterization), which were at least partly determined by the patient’s condition and response to resuscitative efforts, resulted in a survival bias that distorted the survival effect of specialized post-arrest care. Therefore, analyses estimating the effects of 67  interventions and other factors were also performed after stratifying based on whether patients received, or did not receive, specialized post-arrest care.  The potential modification of the effect of the initial rhythm on survival by chain-of-survival interventions was assessed using interaction terms. Models exploring interactions included product terms between initial rhythm (VF/VT, PEA and “AED-no shock, No-strip” with asystole as the reference group) and epinephrine administration, and between initial rhythm and advanced airway placement. Model fit was assessed via hierarchical models, and interaction terms were retained if the p-value was <0.10 [78]. Data management and analysis were performed using IBM SPSS Ver. 23.0 (IBM, Armonk, NY).  Ethics approval: The University of British Columbia Behavioral Research Ethics Board provided approval for this study. In addition, the ROC obtained ethics approval from all participating sites for the original data collection.  3.4 Results There were 18,034 OHCA cases screened by the ROC PRIMED study (Figure 3.1.) Of these, 3,361 (18.5%) were excluded because the patient was not enrolled in ROC PRIMED (n=1324, 7.3%), or they did not meet enrolment criteria for this study (not in cardiac arrest, n=43; no resuscitation was attempted, n=21; EMS witnessed arrest, n=1886; age less than 20 years, n=72. A further 15 cases were excluded because of missing information (1.0%). Of the 14,673 cases of adult, non-traumatic, out-of-hospital, 68  non-EMS witnessed cardiac arrest (OHCA) included in the study population, 3,713 (25.1%) had ROSC in the Emergency Department (Supplemental Table 3.1) and 1,122 (7.6%) were alive at 6 months after discharge from hospital (Table 3.1).  The median age of the cohort was 68.0 (IQR 25.0) years, and approximately 65% were male (Table 3.1). There were statistically significant differences in crude survival rates by age, response time, ROC site, and important Utstein variables (Table 3.2). Table 3.3 shows the unadjusted and adjusted odds ratios for the associations between Utstein variables, EMS clinical and system interventions, and survival. Age, public location, bystander-witnessed arrest, initial rhythm, bystander-attempted resuscitation, epinephrine administration and specialized post-arrest care were significantly associated with survival. The strongest associations were the negative association between epinephrine administration and survival (AOR 0.11, 95% CI 0.09–0.14) and the positive association between specialized post-arrest care and survival (AOR 16.4, 95% CI 13.7–19.7). The absolute risk difference associated with survival following receipt of epinephrine was – 1.8%, and for advanced airway it was 9.3%.    Some associations with ROSC were similar to those with survival, although others were not (Table 3.4). The association between age and ROSC was significantly different from the relationship between age and survival (e.g. ≥75 years ROSC AOR 1.37, 95% CI 1.11–1.69 vs. survival AOR 0.18, 95% CI 0.13–0.25). Other factors whose associations with ROSC differed from their associations with survival included male sex (ROSC AOR 0.71, 95% CI 0.65–0.79 vs. survival AOR 0.98, 95% CI 0.81–1.18), public location of 69  arrest (ROSC AOR 0.74, 95% CI 0.65–0.85 vs. survival AOR 1.46, 95% CI 1.21–1.75), initial rhythm (VF/VT ROSC AOR 1.83, 95% CI 1.61–2.09 vs. survival AOR 8.00, 95% CI 6.00–10.6), advanced airway placement (ROSC AOR 3.20, 95% CI 2.66–3.84 vs. survival AOR 1.13, 95% CI 0.87–1.47), prehospital epinephrine administration (ROSC AOR 0.51, 95% CI 0.45–0.59 vs. survival AOR 0.11, 95% CI 0.09–0.14), specialized post-arrest care (ROSC AOR 23.2, 95% CI 20.1–26.7 vs. survival AOR 16.4, 95% CI 13.7–19.7) and ROC site (Tables 3.3 and 3.4). The absolute risk difference of achieving ROSC following receipt of epinephrine was -15.7%, and for advanced airway it was -1.8%.    Restriction of the cohort to the subset of cases who received specialized post-arrest care showed a graded reduction in the AOR for survival with increasing age and response times (Table 3.5). Public location (AOR 1.76, 95% CI 1.38–2.25) and bystander-witnessed arrest (AOR 2.02, 95% CI 1.56–2.63) were associated with improved survival. The effect of bystander-attempted resuscitation was not significant (AOR 1.18, 95% CI 0.93–1.49). Among cases who did not receive epinephrine, those who had an initial rhythm of asystole (reference group) had survival rates that were not significantly different from those whose initial rhythm was PEA (AOR 1.06, 95% CI 0.46–2.48). However, among cases who did not receive epinephrine, those with an initial rhythm of VF/VT had a 3-fold higher odds of survival compared with those whose initial rhythm was asystole (AOR 3.32, 95% CI 1.63–6.75).   70  The effect of initial rhythm was modified by epinephrine administration. Those whose initial rhythm was asystole and did not receive epinephrine had substantially better survival (AOR 1.00) than those whose initial rhythm was asystole and received epinephrine (AOR 1 x 0.07 = 0.07; a 14-fold decrease in survival; Table 3.5). Those whose initial rhythm was VF/VT and did not receive epinephrine (AOR 3.32 x 1 = 3.32) had significantly better odds of survival compared with those whose initial rhythm was VF/VT and received epinephrine (AOR 3.32 x 0.07 x 2.45 = 0.57; Table 3.5). There was no evidence of effect modification between advanced airway placement and of initial rhythm.   Supplemental Table 3.2 provides results for the cohort of cases who had ROSC at ED and shows associations that were generally similar to the full cohort with regard to age, public location, and witnessed arrest. Among cases who had ROSC at ED arrival, the AOR for survival given advanced airway placement was 0.46 (95% CI 0.32–0.65), the AOR for epinephrine administration was 0.12 (95% CI 0.10–0.15) and the AOR for specialized post-arrest care was 3.30 (95% CI 2.65–4.10). Supplemental Table 3.3 provides associations between interventions and other factors and survival for cases who did not receive specialized post-arrest care, including those for whom resuscitation was terminated in the field. Associations between age, public location, and initial rhythm and survival in this cohort were similar to the associations found the in full cohort (except for EMS response time which was not associated with survival).  71  3.5 Discussion Our results suggest that survival following cardiac arrest is associated with a number of patient characteristics and EMS system factors. We found improved survival was associated with younger age, shorter response time, public location, bystander-witnessed arrest, and ventricular fibrillation/tachycardia as the initial rhythm. Of note, prehospital epinephrine administration had a strong negative association with survival, while bystander resuscitation, EMS response times, and advanced airway placement had variable associations with survival. The negative association between epinephrine administration and survival was substantially stronger than the association between epinephrine administration and ROSC, suggesting a small and negative short-term effect and a much larger negative effect on longer-term survival. Finally, prehospital epinephrine administration modified the effect of initial cardiac rhythm on survival: asystole was associated with a substantial decrease in survival if epinephrine was administered, while ventricular fibrillation/tachycardia was associated with a relatively smaller decrease in survival if epinephrine was administered.  Our results underscore that an initial rhythm of VF/VT, a rhythm that is amenable to defibrillation, is a determinant of survival in out-of-hospital cardiac arrest [79]. Further, our results confirm the known beneficial role of bystander witness and public location of cardiac arrest with regard to survival irrespective of other treatments or circumstances. The gradients in effects by age and EMS response times on survival were all in the expected direction and this strengthens the plausibility of our findings.   72  This study confirms results of previous work by our group in a different population (Chapter 2; [67, 68]) and shows similar findings with regard to clinical interventions. Additionally, we examined potential modification of the effect of the initial rhythm by two important and controversial clinical interventions: epinephrine administration and advanced airway placement.   Strikingly, this analysis identified 5.8-fold lower odds of survival among those with VF/VT who received epinephrine compared with those with VF/VT who did not receive epinephrine. The statistically significant difference in the effect of epinephrine among those with asystole as the initial rhythm (AOR 0.07) and those with VF/VT as the initial rhythm (AOR 0.57) suggests that epinephrine has a negative effect on both those with asystole and those with VF/VT, although the negative effect is much larger among those with asystole. Epinephrine administered to cases of out-of-hospital cardiac arrest appears to be harmful, and the harm appears to be greatest among those with asystole who already have a poor prognosis compared with those who have VF/VT as the initial rhythm.    The role of epinephrine in out-of-hospital cardiac arrest has been examined in previous studies [36-38, 71, 80]. Three systematic reviews and meta-analyses [36-38] found between a 2- and 3-fold increase in the odds of ROSC, but varying effects on survival to discharge from hospital including null effects [36, 38], and a halving of the odds of neurologically intact survival [38]. A 2014 French study by Dumas et al.[80] found that epinephrine was associated with increased mortality, and that the negative effect of epinephrine showed a strong dose-response based on volume administered. Additionally, 73  the effect of epinephrine differed among those with a shockable vs. non-shockable rhythm (AOR 0.26, 95% CI 0.16–0.43 for shockable rhythm and AOR 0.42, 95% CI 0.22–0.80 for a non-shockable rhythm) and these findings are consistent with the effect modification seen in our study. A Japanese cohort study by Machida et al. [71] and a randomized trial by Jacobs et al. [34] found no short-term or long-term survival benefit for epinephrine. Post-hoc analysis conducted by Jacobs et al. intended to isolate the effect of epinephrine based on shockable and non-shockable rhythms demonstrated differential effects on ROSC (non-shockable rhythms OR 6.9, 95% CI 2.6–18.4 vs. shockable rhythms OR 2.4, 95% CI 1.2–4.5)[34], no significant effect on survival to hospital discharge. Our study corroborates these finding regarding the differential effect of epinephrine based on initial rhythm while simultaneously indicating that epinephrine has a harmful, rather than a beneficial, effect.  A recent systematic review and meta-analysis of tracheal intubation vs. supraglottic airway use in OHCA [28] found a beneficial effect on ROSC, admission to hospital and neurologically intact survival, but not on survival to hospital discharge. We found a similarly mixed result, with a beneficial effect on ROSC (AOR 3.20 95% CI 2.66–3.84) while analyses stratified by receipt of specialized post-arrest care (AOR 0.72 95% CI 0.45–1.16) vs. lack of specialized post-arrest care (AOR 1.77 95% CI 1.26–2.49), survival among cases that achieved ROSC (AOR 0.46 95% CI 0.32–0.65), and no interaction between advanced airway and initial rhythm with regard to survival. These findings suggest complex effects depending on the clinical circumstances affecting the management of airway in cardiac arrest. The close temporal relationship between ROSC 74  and airway management, and lack of available time stamps prevented us from determining the sequence of airway management and ROSC. Among cases who had specialized post-arrest care, we found a null effect from advanced airway management. However, for cases without access to specialized post-arrest care, we found advanced airway management was beneficial. This has not been described previously and may indicate that decisions surrounding advanced airway management require more nuanced algorithms and complex decision-making in the field depending on both the patient’s clinical response, and availability of post-arrest resources at the destination hospital. Finally, our finding of a null result for the interaction between initial rhythm and advanced airway management suggests advanced airway management does not influence prognosis based on initial rhythm, simplifying the potential parameters required for resuscitation decision-making.   Our results have a number of implications for both further research and OHCA management. It is hoped that the ongoing effort to evaluate the effectiveness of epinephrine in OHCA with the launch of the Paramedic-2 trial [75] will shed light on modification of the effect of epinephrine by the initial rhythm. Evidence from this and prior research on interventions in the chain-of-survival should prompt a review of the role of epinephrine by resuscitation councils, and potentially an amendment of current resuscitation guidelines.   Our study has both strengths and weaknesses that merit consideration. The completeness and accuracy of the data in this study was assured by the standardized protocols used in 75  the ROC PRIMED randomized trial. Trial algorithms also ensured optimal administration of the chain-of-survival interventions. Potential weaknesses include the non-experimental nature of the study design. The age of our data also merits consideration, although we are unaware of any substantive changes to published resuscitation guidelines since the data were collected. Our study was retrospective and as a result we were not able to explore the relationship between timing of EMS CPR and rhythm analysis. Finally, we were unable to evaluate the role of specialized post-arrest care as contrasts between those who did and those who did not receive specialized post-arrest care were compromised by survival bias.   3.6 Conclusion In summary, our study evaluated the effects of several key clinical elements in chain-of-survival interventions and other factors associated with out-of-hospital cardiac arrest. We found a consistent association between younger age, public location of arrest, and bystander witness to the arrest and survival to hospital discharge. Epinephrine administration was associated with poorer survival and its use modified the effect of initial rhythm on survival. Advanced airway placement, on the other hand, did not modify the effect of initial rhythm on survival, which may simplify both clinical application and the design of future randomized clinical trials. Further research is needed to elucidate best practices in OHCA and identify the impact of specialized post-arrest care on survival.      76  Figure 3.1 Subject enrolment details.                                  Patients screened for enrolment in ROC PRIMED randomized clinical trials n=18,034 Excluded from analysis: 1. Not enrolled, no data collected: n=1,324 2. Case not in cardiac arrest: n=43 3. EMS witnessed arrest: n=1,886 4. No attempted resuscitation: n=21 5. Age < 20 years: n=72  Missing information on outcomes: n=13 Missing information on sex/gender: n=2  Available for Analysis Report by ROC site 1 = 226 2 = 680 3 = 935 4= 2,377 5 = 481 6 = 949 7 = 649 8 = 2,125 9 = 4,596 10 = 1,655  Study cohort n=14,673 77  Table 3.1.  Study subjects’ demographic and clinical characteristics (n=14,673).  Characteristic Number (%)* Age in years, median (IQR) 68.0 (25.0) Age in years **    Adult (20 – 39 years)    Middle age (40 – 59 years)    Older (60 – 74 years)    Elderly >75     Unknown   974 (6.6) 3,867 (26.4) 4,248 (29.0) 5,549 (37.8) 35 (0.2)   Patient sex   Male   Female   9,464 (64.5) 5,209 (35.5) Public arrest location Non-public arrest location 2,247 (15.3) 12,426 (84.7)  Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander     Unknown  6,320 (43.1) 7,867 (53.6) 489 (3.3)  Cause of arrest assumed cardiac Other, obvious cause  13,978 (95.3) 695 (4.7)  Initial rhythm    VF/VT ^   PEA^^   Asystole   Perfusing   AED^^^ – no shock advised   Cannot determine   Rhythm analysis missing   3,352 (22.8) 2,704 (18.4) 6,513 (44.4) 25 (0.2) 1,337 (9.4) 49 (0.3) 693 (4.7) Response time   Median (IQR)   5:47 (1.74) Response time      < 6 minutes   6 – 9 minutes   10 – 15 minutes    ≥ 16 minutes    Unknown  8,422 (57.4) 4,979 (33.9) 721 (4.9) 105 (0.7) 446 (3.0) 78  Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted    Unknown/not noted   7,864 (53.6) 6,115 (41.7) 694 (4.7) Advanced airway placed    Yes    No   12,594 (85.8) 2,079 (14.2) Prehospital epinephrine  Yes   No   11,908 (81.2) 2,765 (18.8) Specialized post-arrest care provided*^   Yes   No    2,006 (13.7)  12,667 (86.3)  ROC site n (%)   1   2   3   4   5    6   7   8    9  10   226 (1.5) 680 (4.6) 935 (6.4) 2,377 (16.2) 481 (3.3) 949 (6.5) 649 (4.4) 2,125 (15.5) 4,596 (31.3) 1,655 (11.3)   *Unless otherwise specified. ** Exact age of cases older than 89 years of age not provided (n=851). All cases older than 90 years were imputed with a random uniform number between 0 and 9 plus 90.   ^ VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^ PEA (pulseless electrical activity)  ^^^ AED (automatic external defibrillator) *^ Specialized post-arrest care: administration prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis.  79  Table 3.2. Survival by demographic and clinical characteristics (n=14673).  Characteristic Patient Survived n (%)*=1,122 (7.6) Patient did not survive n (%)*=13,551 (92.4) p-value Patient age  Years median (IQR)   60.0 (19.0)  69.0 (26.0)   < 0.001 Age in years **    Adult (20 – 39 years)    Middle age (40 – 59 years)    Older (60 – 74 years)    Elderly >75     Unknown Missing   120 (12.3) 426 (11.0) 406 (9.6) 170 (3.1) 0 (0.0)  854 (87.7) 3,441 (89.0) 3,842 (90.2) 5379 (96.6) 35 (100.0)    < 0.001 Patient sex   Male   Female   818 (8.6) 304 (5.8)  8,646 (91.4) 4,905 (94.2)  < 0.001 Location of arrest   Non-public    Public     683 (5.5) 439 (19.5)  11,743 (94.5) 1,808 (80.5)  < 0.001 Bystander-witnessed status     Witnessed by bystander     Not witnessed by bystander     Unknown  849 (13.4) 247 (3.1) 26 (5.3)   5,471 (86.6) 7,620 (96.9) 460 (94.7)   < 0.001 Cause of arrest     80      Assumed cardiac     Other, obvious cause    1,083 (7.7) 39 (5.6)  12,895 (92.3) 656 (94.4)   0.04 Initial rhythm    VF/VT**   PEA^   Asystole   Perfusing   AED^^ – no shock advised   Cannot determine   Rhythm analysis missing   743 (22.2) 149 (5.5) 71 (1.1) 16 (64.0) 40 (3.0) 6 (12.2) 97 (14.0)   2,609 (77.8) 2,555 (94.5) 6,442 (98.9) 9 (36.0) 1,297 (97.0) 43 (87.8) 596 (86.0)     < 0.001 Response time   Median (IQR)   5:00 (2:21)  5:30 (2:43)  < 0.001  Response time         < 6 minutes     6 – 9 minutes     10 – 15 minutes     ≥ 16 minutes     Unknown   748 (8.9) 310 (6.2) 33 (4.6) 3 (2.9) 28 (6.3)  7,674 (91.1) 4,669 (93.8) 688 (95.4) 102 (97.1) 418 (93.7)   < 0.001  Bystander-attempted resuscitation     No bystander-attempted resuscitation    Bystander resuscitation attempted    Unknown   443 (5.6) 642 (10.5) 37 (7.6)  7,421 (94.4) 5,473 (89.5) 657 (94.7)   <0.001 Advanced airway placed    Yes    No   930 (7.4) 192 (9.2)  11,664 (92.6) 1,887 (90.8)  0.003 81  Prehospital epinephrine administration  Yes   No   557 (4.7) 565 (20.4)   11,351 (95.3) 2,200 (79.6)  < 0.001 Specialized post-arrest care provided^^^   Yes   No  824 (41.1) 298 (2.4)   1,182 (58.9) 12,369 (97.6)  < 0.001 ROC site n (%)   1   2   3   4   5    6   7   8    9  10  11 (4.9) 29 (4.3) 71 (7.6) 147 (6.2) 23 (4.8) 112 (11.8) 43 (6.6) 313 (14.7) 220 (4.8) 153 (9.2)   215 (95.1) 1651 (95.7) 864 (92.4) 2,230 (93.8) 458 (95.2) 837 (88.2) 606 (93.4) 1,812 (85.3) 4,376 (95.2) 1,502 (90.8)       < 0.001  * Unless otherwise specified. **VF/VT (ventricular fibrillation/ventricular tachycardia)  ^PEA (pulseless electrical activity)  ^^AED (automatic external defibrillator) ^^^ Specialized post-arrest care: administration prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis. 82 Table 3.3. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and survival (n=14,673).  Characteristic Unadjusted  Adjusted OR (95% CI) p-value  OR (95% CI) p-value Age in years **    Adult (20 – 39 years)    Middle age (40 – 59 years)    Older (60 – 74 years)    Elderly (>75 years)     Unknown   Reference 0.88 (0.71 – 1.09) 0.75 (0.61 – 0.93) 0.23 (0.18 – 0.29) Undefined   0.25 0.01 < 0.001   Reference 0.59 (0.43 – 0.80) 0.39 (0.29 – 0.54) 0.18 (0.13 – 0.25) Undefined   0.001 0.001 < 0.001  Patient sex   Male   Female   1.53 (1.33 – 1.75) Reference  < 0.001  0.98 (0.81 – 1.18) Reference  0.84 Location of arrest   Non-public   Public     Reference 4.18 (3.67 – 4.75)   < 0.001   Reference 1.46 (1.21 – 1.75)   < 0.001  Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander     Unknown  4.79 (4.14 – 5.54) Reference 1.74 (1.15 – 2.64)  < 0.001  0.01  2.21 (1.83 – 2.68) Reference 1.41 (0.80 – 2.48)   < 0.001  0.24 Cause of arrest      Assumed cardiac      Other, obvious cause     0.71 (0.51 – 0.98) Reference  0.04  0.78 (0.52 – 1.19) Reference  0.25 Initial rhythm    VF/VT*   PEA**   Asystole  25.84 (20.17 – 33.10) 5.29 (3.97– 7.05) Reference  < 0.001 < 0.001   8.00 (6.0 – 10.63) 3.45 (2.51 – 4.74) Reference  < 0.001 < 0.001   83   Perfusing rhythm   AED^ – no shock advised   Cannot determine   Rhythm analysis missing  161.3 (69.0 –377.2) 2.80 (1.89 – 4.14) 12.66 (5.22 – 30.69) 14.77 (10.75 – 20.28) < 0.001 < 0.001 0.001 < 0.001 66.1 (25.7 – 170.2) 2.14 (1.37 – 3.30) 8.90 (2.8 – 28.3) 3.59 (2.41 – 5.36)  < 0.001 0.001 < 0.001 < 0.001 Response time category    < 6 minutes    6 – 9 minutes   10 – 15 minutes    ≥ 16 minutes    Unknown   Reference 0.68 (0.59 – 0.78) 0.49 (0.34 – 0.70) 0.30 (0.10 – 0.95) 0.69 (0.47 – 1.02)    < 0.001 < 0.001 0.04 0.06  Reference 0.86 (0.71 – 1.04) 0.57 (0.37 – 0.90) 0.36 (0.08 – 1.58) 0.83 (0.48 – 1.41)   0.12 0.02 0.18 0.48 Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted   Unknown    1.97 (1.73 – 2.23) Reference 0.94 (0.67 – 1.33)  < 0.001  0.74  1.21 (1.01 – 1.43) Reference 1.33 (0.81 – 1.43)  0.04  0.26 Advanced airway placed     Yes    No      0.78 (0.67 – 0.92) Reference  0.003  1.13 (0.87 – 1.47) Reference  0.38 Prehospital epinephrine administered    Yes    No    0.19 (0.17 – 0.22) Reference  < 0.001  0.11 (0.09 – 0.14) Reference  < 0.001 Specialized post-arrest care provided^^    Received    Did not receive  28.94 (25.2 – 33.5) Reference  < 0.001  16.41 (13.7 – 19.7) Reference   < 0.001 ROC site    1   2   3  1.02 (0.55 – 1.89) 0.89 (0.60 – 1.32) 1.64 (1.24 – 2.16)  0.96 0.55 0.001  0.90 (0.42 – 1.95) 2.01 (1.20 – 3.38) 2.90 (1.92 – 4.39)  0.79 0.008 < 0.001  84   4   5    6   7   8    9  10 1.31 (1.06 – 1.63) 1.00 (0.64 – 1.55) 2.66 (2.10 – 3.38) 1.14 (1.01 – 1.98) 3.44 (2.87 – 4.12) Reference 2.03 (1.64 – 2.51) 0.01 1.00 < 0.001 0.045 <0.001  < 0.001 1.38 (1.03 – 1.84) 0.90 (0.50 – 1.63) 1.40 (1.00 – 1.96) 2.98 (1.90 – 4.66) 2.56 (1.95 – 3.36) Reference 1.98 (1.48 – 2.65)  0.03 0.73 0.054 < 0.001 < 0.001  < 0.001   *VF/VT (ventricular fibrillation/ventricular tachycardia)  ** PEA (pulseless electrical activity) ^ AED (automatic external defibrillator) ^^ Specialized Post-arrest Care: administration prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis.    85 Table 3.4. Logistic regression model showing unadjusted and adjusted associations between interventions and other factors and return of spontaneous circulation (ROSC) (n=14,673).  Characteristic Unadjusted Adjusted OR (95% CI) p-value  AOR (95% CI) p-value Patient age  Adult (20 – 39 years) Middle age (40 – 59 years) Older (60 – 74) Elderly (≥ 75 years)  Unknown age   Reference 1.14 (0.95 – 1.31) 1.27 (1.08 – 1.50) 0.11 (0.94 – 1.30) Undefined   0.20 0.003 0.22  Reference 0.95 (0.76 – 1.18) 1.10 (0.89 – 1.36) 1.37 (1.11 – 1.69) Undefined   0.64 0.39 0.004  Patient sex   Male   Female   0.93 (0.86 – 1.00) Reference  0.06  0.71 (0.65 – 0.79) Reference  < 0.001  Location of arrest   Non-public   Public     Reference 1.62 (1.47 – 1.78)    < 0.001  Reference 0.74 (0.65 – 0.85)   < 0.001 Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander     Unknown  2.80 (2.59 – 3.02) Reference 0.92 (0.72– 1.18)   < 0.001  0.52  1.89 (1.71 – 2.09) Reference 0.94 (0.68 – 1.28)  < 0.001  0.68 Cause of arrest      Assumed cardiac     Other, obvious cause     0.87 (0.73 – 1.04) Reference  0.13  1.31 (1.04 – 1.65) Reference  0.02 Initial rhythm    VF/VT*   PEA**   Asystole  4.37 (3.98 – 4.84) 2.96 (2.66 – 3.29) Reference  < 0.001 < 0.001   1.83 (1.61 – 2.09) 2.27 (2.00 – 2.57) Reference  < 0.001 < 0.001   86   Perfusing rhythm   AED^ – no shock advised   Cannot determine   Rhythm analysis missing 71.91 (16.9 – 305.5) 1.64 (1.41 – 1.91) 1.41 (0.68 – 2.91) 3.64 (3.07 – 4.31)  < 0.001 < 0.001 0.36 < 0.001 81.4 (18.4 – 360.1) 1.48 (1.24 – 1.76) 1.53 (0.65 – 3.58) 1.46 (1.14 – 1.88) < 0.001 < 0.001 0.33 0.003 Response time category    < 6 minutes    6 – 9 minutes    10 – 15 minutes    ≥ 16 minutes    Unknown   Reference 0.90 (0.83 – 0.97) 0.78 (0.65 – 0.94) 0.74 (0.46 – 1.19) 0.86 (0.69 – 1.08)   0.01 0.01 0.21 0.19  Reference 0.96 (0.86 – 1.06) 0.82 (0.66 – 1.03) 1.03 (0.60 – 1.78) 1.05 (0.80 – 1.39)   0.42 0.09 0.91 0.73  Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted   Unknown    1.42 (1.32 – 1.53) Reference 0.86 (0.71 – 1.04)  < 0.001  0.12  1.10 (1.00 – 1.21) Reference 1.02 (0.79 – 1.32)  0.053  0.89 Advanced airway placed     Yes    No   2.19 (1.92 – 2.49) Reference  < 0.001  3.20 (2.66 – 3.84) Reference  < 0.001 Prehospital epinephrine administered    Yes    No    0.70 (0.64 – 0.77) Reference  0.001  0.51 (0.45 – 0.59) Reference  < 0.001 Specialized post-arrest care provided^^   Post-arrest care provided   Post-arrest care not provided   26.96 (23.72 – 30.63) Reference  < 0.001  23.2 (20.1 – 26.7) Reference   < 0.001 ROC site   1   2   3  0.76 (0.53 – 1.10) 0.61 (0.48 – 0.77) 1.45 (1.23 – 1.70)  0.15 0.001 < 0.001  0.70 (0.46 – 1.07) 0.83 (0.63 – 1.10) 1.80 (1.44 – 2.24)  0.10 0.19 < 0.001  87   4   5    6   7   8    9  10 1.00 (0.87 – 1.12) 1.17(0.93 – 1.46) 2.18 (1.87– 2.53) 1.35 (1.12 – 1.64) 2.45 (2.19 – 2.75 Reference 2.01 (1.77 – 2.28) 0.84 0.18 < 0.001 0.002 < 0.001  < 0.001 0.97 (0.83 – 1.12) 0.95 (0.71 – 1.27) 1.58 (1.30 – 1.92) 1.99 (1.58 – 2.49) 1.57 (1.34 – 1.85) Reference 2.45 (2.11 – 2.86)  0.67 0.74 < 0.001 < 0.001 < 0.001  < 0.001  * VF/VT (ventricular fibrillation/ventricular tachycardia)  ** PEA (pulseless electrical activity) ^ AED (automatic external defibrillator) ^^ Specialized Post-arrest Care: administration of prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis.     88 Table 3.5. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients who received specialized post-arrest care (hypothermia, fibrinolysis or diagnostic cardiac catheterization) (n = 2,006).  Characteristic Unadjusted Adjusted OR (95% CI) p-value  OR (95% CI) p-value Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74)   Elderly (≥ 75 years)   Unknown Age   Reference 0.90 (0.63 – 1.26) 0.76 (0.53 – 1.07) 0.24 (0.16 – 0.35) No cases   0.51 0.12 < 0.001  Reference 0.67 (0.43 – 1.05) 0.45 (0.29 – 0.70) 0.15 (0.09 – 0.25) No cases   0.08 < 0.001 < 0.001  Patient sex   Male   Female   1.63 (1.33 – 1.99) Reference  < 0.001  1.17 (0.90 – 1.53) Reference  0.23 Location of arrest   Non-public   Public     Reference 2.88 (2.37 – 3.50)   < 0.001   Reference 1.77 (1.38 – 2.26)   < 0.001  Bystander-witnessed status    Witnessed by bystander     Not witnessed by bystander     Unknown  2.84 (2.30 – 3.51) Reference 2.34 (1.21 – 4.52)  < 0.001  0.01  2.04 (1.57 – 2.65) Reference 2.53 (1.06 – 6.04)   < 0.001  0.04 Cause of arrest     Assumed cardiac     Other, obvious cause   0.38 (0.22 – 0.66) Reference  0.001  0.47 (0.24 – 0.93) Reference  0.03 Initial rhythm   VF/VT*   PEA**  10.11 (7.0 – 14.5) 2.28 (1.47 – 3.54)  < 0.001 < 0.001  3.32 (1.63 – 6.75) 1.06 (0.46 – 2.48)  0.001 0.89  89   Asystole   Perfusing   AED^ – no shock advised   Cannot determine   Rhythm analysis missing  Reference Undefined 1.70 (0.91 – 3.18) 11.55 (1.9 – 71.4) 5.09 (3.27 – 7.94)    0.10 0.01 < 0.001 Reference Undefined 1.32 (0.62 – 2.84) 2.47 (0.27 – 22.6) 3.65 (2.09 – 6.39)   0.47 0.42 < 0.001 Interactions (initial rhythm * epinephrine)   VF/VT * epinephrine    PEA * epinephrine   1.22 (1.01 – 1.46) 0.24 (0.17 – 0.35)   0.04 < 0.001   2.45 (1.20 – 5.03) 2.20 (0.88 – 5.49)   0.01 0.09 Response time category   < 6 minutes   6 – 9 minutes   10 – 15 minutes   ≥ 16 minutes   Unknown   Reference 0.75 (0.62 – 0.92) 0.40 (0.23 – 0.68) 0.51 (0.10 – 2.64) 0.94 (0.53 – 1.64)    0.01 0.001 0.42 0.82  Reference 0.86 (0.66 – 1.12) 0.43 (0.22 – 0.85) 0.40 (0.05 – 3.17) 1.02 (0.49 – 2.12)   0.31 0.02 0.38 0.96 Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted   Unknown    1.50 (1.25 – 1.80) Reference 1.37 (0.77 – 2.42)  < 0.001  0.28  1.18 (0.93 – 1.49) Reference 1.82 (0.81 – 4.09)  0.17  0.15 Advanced airway placed    Yes    No      0.24 (0.16 – 0.34) Reference  < 0.001  0.72 (0.45 – 1.16) Reference  0.17 Prehospital epinephrine administered    Yes    No    0.14 (0.11 – 0.18) Reference  < 0.001  0.07 (0.04 – 0.14) Reference  < 0.001 ROC site n (%)   1   2  1.26 (0.47 – 3.39) 1.35 (0.67 – 2.71)  0.65 0.40  1.23 (0.35 – 4.33) 1.80 (0.71 – 4.55)  0.75 0.22  90   3   4   5    6   7   8    9  10  1.34 (0.88 – 2.02) 1.70 (1.23 – 2.36) 0.50 (0.25 – 0.82) 1.41 (1.01 – 1.95) 1.61 (0.91 – 2.85) 1.33 (1.03 – 1.71) Reference 1.33 (0.96 – 1.85)  0.17 0.001 0.009 0.04 0.11 0.03  0.08  2.16 (1.21 – 3.85) 1.48 (0.97 – 2.24) 0.49 (0.23 – 1.08) 1.37 (0.91 – 2.07) 3.02 (1.46 – 6.25) 2.30 (1.61 – 3.28) Reference 1.41 (0.93 – 2.14)  0.01 0.07 0.08 0.14 0.003 < 0.001  0.10  * VF/VT (ventricular fibrillation/ventricular tachycardia)  ** PEA (pulseless electrical activity) ^ AED (automatic external defibrillator)    91 Supplemental Table 3.1. Return of spontaneous circulation (ROSC) at emergency department by demographic and clinical characteristics (n=14,673).  Characteristic     Return of spontaneous circulation (ROSC) n (%) = 3,713 (25.3) No return of spontaneous circulation (ROSC) n (%) = 10,960 (74.7)  p-value Patient age  Years median (IQR)   68.0 (24.0)   68.0 (25.0)   0.83  Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Unknown   230 (23.6) 957 (24.7) 1,161 (27.3) 1,365 (24.6) 0 (0.0)  744 (76.4) 2,910 (75.3) 3,087 (72.7) 4,184 (75.4) 35 (100.0)   < 0.001 Patient sex   Male   Female   2,347 (24.8) 1,366 (26.2)  7,117 (75.2) 3,843 (73.8)  0.06  Location of arrest   Non-public   Public     2,959 (23.8) 754 (33.6)  9,467 (76.2) 1,493 (66.4)  < 0.001 Bystander-witnessed status     Witnessed by bystander     Not witnessed by bystander     Unknown  2,300 (36.4) 1,336 (17.0) 77 (15.7)   4,020 (63.6) 6,531 (83.0) 409 (84.2)    < 0.001 Cause of arrest      Assumed cardiac   3,554 (25.4)  10,424 (74.6)    92     Other, obvious cause   159 (22.9)  536 (77.1)  0.13 Initial rhythm    VF/VT*   PEA**   Asystole   Perfusing   AED^ – no shock advised   Cannot determine   Rhythm analysis missing   1,382 (41.2) 868 (32.1) 898 (13.8) 23 (92.0) 279 (20.8) 9 (18.4) 255 (36.8)  1,970 (58.8) 1,836 (67.9) 5,615 (86.2) 2 (8.0) 1,059 (79.2) 40 (81.6) 438 (63.2)     < 0.001 ROC site   1   2   3   4   5    6   7   8    9  10  36 (15.9) 89 (13.1) 247 (26.4) 468 (19.7) 108 (22.5) 333 (35.1) 163 (25.1) 804 (37.8) 914 (19.9) 551 (33.3)   190 (84.1) 591 (86.9) 688 (73.6) 1,909 (80.3) 373 (77.5) 616 (64.9) 486 (74.9) 1,321 (62.2) 3,682 (80.1) 1,104 (66.7)      < 0.001  Response time   Median (IQR)   5:18 (2:42)  5:30 (2:42)  0.02  Response time  < 6 minutes 6 – 9 minutes 10 – 15 minutes ≥ 16 minutes Unknown   2,220 (26.4) 1,208 (24.3) 158 (21.9) 22 (21.0) 105 (23.5)   6,202 (73.6) 3,771 (75.7) 563 (78.1) 83 (79.0) 341 (76.5)     0.007  93 Bystander-attempted resuscitation     No bystander-attempted resuscitation    Bystander resuscitation attempted    Unknown   1,779 (22.6) 1,795 (29.4) 139 (20.0)  6,085 (77.4) 4,320 (70.6) 555 (80.0)   < 0.001 Advanced airway placed    Yes    No   3,411 (27.1) 302 (14.5)  9,183 (72.9) 1,777 (85.5)  < 0.001 Prehospital epinephrine administered    Yes    No   2,857 (24.0) 856 (31.0)  9,051 (76.0) 1909 (69.0)  0.001 Specialized post-arrest care provided ^^   Yes   No  1,680 (83.7) 2,033 (16.0)   326 (16.3) 10,634 (84.0)  < 0.001  * VF/VT (ventricular fibrillation/ventricular tachycardia)  ** PEA (pulseless electrical activity) ^ AED (automatic external defibrillator) ^^ Specialized post-arrest care: administration prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis.    94 Supplemental Table 3.2. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between of interventions and other factors and survival among patients with ROSC at arrival in the emergency department (n=3,713).  Characteristic Unadjusted  Adjusted OR (95% CI) p-value  OR (95% CI) p-value Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Unknown   Reference 0.74 (0.56 – 0.99) 0.50 (0.38 – 0.67) 0.14 (0.10 – 0.19) No cases   0.04 < 0.001 < 0.001  Reference 0.55 (0.37 – 0.80) 0.32 (0.22 – 0.47) 0.12 (0.08 – 0.17) No cases   0.002 < 0.001 < 0.001 Patient sex   Male   Female   1.78 (1.52 – 2.08) Reference  < 0.001  1.13 (0.91 – 1.39) Reference  0.27 Location of arrest   Non-public   Public     Reference 3.76 (3.18 – 4.44)   < 0.001   Reference 1.55 (1.24 – 1.94)   < 0.001  Bystander-witnessed status     Witnessed by bystander     Not witnessed by bystander     Unknown  2.57 (2.17 – 3.04) Reference 2.30 (1.39 – 3.80)  < 0.001  0.001  1.78 (1.43 – 2.21) Reference 1.99 (0.99 – 4.01)   < 0.001  0.054 Cause of arrest      Assumed cardiac      Other, obvious cause     0.80 (0.55 – 1.16) Reference  0.24  0.77 (0.47 – 1.26) Reference  0.30 Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted   Unknown   1.61 (1.39 – 1.87) Reference 1.14 (0.77 – 1.70)  < 0.001  0.52  1.03 (0.84 – 1.26) Reference 1.44 (0.79 – 2.61)  0.78  0.24  95  Initial rhythm   VF/VT*   PEA**   Asystole   Perfusing   AED^ – no shock advised   Cannot determine   Rhythm analysis missing   12.19 (9.26 – 16.06) 2.49 (1.82 – 3.40) Reference 29.79 (11.79-75.03) 1.94 (1.26 – 2.99) 3.72 (0.76 – 18.29) 7.36 (5.13 – 10.55)  < 0.001 < 0.001  < 0.001 0.003 0.11 < 0.001   5.41 (3.90 – 7.48) 1.78 (1.24 – 2.55) Reference 7.69 (2.74 – 21.55) 1.27 (0.76 – 2.13) 2.65 (0.32 – 22.11) 2.51 (1.60 – 3.96)  < 0.001 0.002  < 0.001 0.36 0.37 < 0.001 Response time category   < 6 minutes   6 – 9 minutes   10 – 15 minutes   ≥ 16 minutes   Unknown   Reference 0.70 (0.60 – 0.82) 0.47 (0.31 – 0.72) 0.28 (0.05 – 0.98) 0.75 (0.48 – 1.18)   < 0.001 < 0.001 0.05 0.21  Reference 0.91 (0.73 – 1.14) 0.76 (0.44 – 1.31) 0.45 (0.08 – 2.45) 1.03 (0.56 – 1.90)   0.40 0.32 0.36 0.91 Advanced airway placed     Yes    No      0.25 (0.20 – 0.32) Reference  < 0.001  0.46 (0.32 – 0.65) Reference  < 0.001 Prehospital epinephrine administered    Yes    No    0.12 (0.10 – 0.15) Reference  < 0.001  0.12 (0.10 – 0.15) Reference  < 0.001 Specialized post-arrest care provided^^   Post-arrest care provided   Post-arrest care not provided   5.19 (4.43 – 6.10) Reference  < 0.001  3.30 (2.65 – 4.10) Reference  < 0.001 ROC site   1   2   3  1.38 (0.66 – 2.91) 1.04 (0.62 – 1.76) 1.37 (0.99 – 1.88)  0.40 0.88 0.01  1.27 (0.45 – 3.59) 1.67 (0.83 – 3.37) 2.20 (1.36 – 3.55)  0.65 0.17 0.001  96   4   5    6   7   8    9  10  1.28 (0.99 – 1.66) 0.82 (0.49 – 1.36) 1.59 (1.20 – 2.10) 1.13 (0.76 – 1.67) 2.17 (1.76 – 2.69) Reference 1.14 (0.89 – 1.47) 0.06 0.44 0.001 0.54 < 0.001  0.30 1.22 (0.86 – 1.74) 0.70 (0.34 – 1.41) 1.12 (0.76 – 1.64) 2.12 (1.23 – 3.66) 2.47 (1.81 – 3.37) Reference 1.35 (0.96 – 1.90) 0.27 0.32 0.58 0.007 < 0.001 0.04 0.08  * VF/VT (ventricular fibrillation/ventricular tachycardia) ** PEA (pulseless electrical activity) ^ AED (automatic external defibrillator) ^^ Specialized post-arrest care: administration prehospital hypothermia, diagnostic cardiac catheterization, or fibrinolysis.    97 Supplemental Table 3.3. Logistic regression model (restricted cohort) showing unadjusted and adjusted associations between interventions and other factors and survival among patients who did NOT receive specialized post-arrest care (n=12,667).  Characteristic Unadjusted Adjusted OR (95% CI) p-value  OR (95% CI) p-value Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Unknown   Reference 0.64 (0.44 – 0.93) 0.46 (0.32 – 0.68) 0.29 (0.20 – 0.43) Undefined   0.02 < 0.001 < 0.001  Reference 0.50 (0.32 – 0.76) 0.31 (0.20 – 0.49) 0.19 (0.12 – 0.30) Undefined   0.001 < 0.001 < 0.001 Patient sex   Male   Female   1.01 (0.79 – 1.28) Reference  0.95  0.81 (0.63 – 1.06) Reference  0.13 Location of arrest   Non-public   Public     Reference 2.22 (1.70 – 2.92)   < 0.001   Reference 0.98 (0.72 – 1.35)   0.90  Bystander-witnessed status n (%)    Witnessed by bystander     Not witnessed by bystander      Unknown  3.11 (2.43 – 3.99) Reference 1.56 (0.78 – 3.12)  < 0.001  0.20  2.41 (1.81 – 3.22) Reference 0.93 (0.40 – 2.17)   < 0.001  0.86 Cause of arrest      Assumed cardiac n (%)     Other, obvious cause     1.65 (1.07 – 2.54) Reference  0.02  1.14 (0.68 – 1.91) Reference  0.63 Initial rhythm    VF/VT*   PEA**   Asystole  10.59 (7.22–15.52) 6.38 (4.26 – 9.54) Reference  < 0.001 < 0.001   8.87 (5.82 -13.53) 5.07 (3.30 – 7.78) Reference  < 0.001 <0.001   98   Perfusing   AED^ – no shock advised   Cannot determine   Rhythm analysis missing  281.7 (114.2–694.7) 3.41 (2.00 – 5.82) 13.25 (3.91–44.87) 6.48 (3.59–11.69) < 0.001 < 0.001 < 0.001 < 0.001 79.51 (29.5– 214.0) 2.56 (1.46 – 4.48) 15.65 (4.27- 57.38) 2.97 (1.54 – 5.74) < 0.001 0.001 < 0.001 0.001 Response time category   < 6 minutes   6 – 9 minutes   10 – 15 minutes   ≥ 16 minutes   Unknown   Reference 0.85 (0.66 – 1.09) 0.91 (0.54 – 1.56) 0.40 (0.06 – 2.85) 0.59 (0.26 – 1.35)    0.20 0.74 0.36 0.21  Reference 0.86 (0.64 – 1.14) 0.72 (0.40 – 1.30) 0.35 (0.05 – 2.66) 0.67 (0.28 – 1.61)    0.30 0.28 0.31 0.38  Bystander-attempted resuscitation    Bystander resuscitation attempted   No bystander resuscitation attempted   Unknown    1.72 (1.36 – 2.18) Reference 1.29 (0.75 – 2.21)  < 0.001  0.36  1.28 (0.98 – 1.67) Reference 1.07 (0.54 – 2.10)  0.07  0.87 Advanced airway placed    Yes    No      0.49 (0.38 – 0.64) Reference  < 0.001  1.77 (1.26 – 2.49) Reference  0.001 Prehospital epinephrine administered    Yes    No   0.13 (0.10 – 0.16) Reference  < 0.001  0.08 (0.06 – 0.11) Reference  < 0.001 ROC site   1   2   3   4   5    6   7  1.40 (0.50 – 3.89) 1.59 (0.88 – 2.87) 1.98 (1.20 – 3.25) 1.39 (0.93 – 2.08) 1.44 (0.68 – 3.04) 1.51 (0.85 – 2.68) 2.23 (1.30 – 3.82)  0.52 0.12 0.007 0.11 0.34 0.16 0.003  0.64 (0.21 – 1.91) 2.15 (1.10 – 4.20) 3.63 (1.98 – 6.65) 1.18 (0.77 – 1.83) 2.19 (0.93 – 4.98) 1.23 (0.65 – 2.33) 3.04 (1.65 – 5.60)  0.42 0.03 < 0.001 0.45 0.06 0.52 < 0.001  99   8    9  10 2.97 (2.06– 4.28) Reference 3.02 (2.09 – 4.38)  < 0.001  < 0.001  2.94 (1.92 – 4.52) Reference 2.72 (1.80 – 4.11)  < 0.001  < 0.001   * VF/VT (ventricular fibrillation/ventricular tachycardia) ** PEA (pulseless electrical activity) ^ AED (automatic external defibrillator)    100 Chapter 4: Effect of epinephrine and tracheal intubation in out-of-hospital cardiac arrest: A propensity score analysis 4.1 Summary The effect of epinephrine and tracheal intubation (ETI) in out-of-hospital cardiac arrest (OHCA) resuscitation has not been adequately studied and remains unclear. We estimated the effect of prehospital epinephrine and ETI in out-of-hospital cardiac arrest resuscitation while accounting for patient characteristics and other factors.  We conducted a retrospective cohort study of non-traumatic OHCA cases enrolled by the Resuscitation Outcomes Consortium. Outcome measures included survival to hospital discharge and return of spontaneous circulation (ROSC) upon arrival to the emergency department. Propensity score matched analyses were performed based on the propensity to receive epinephrine or ETI. Additional analyses were performed while stratifying on availability of specialized post-arrest care to control for survival bias.  Of the 28,484 cases available for study, 12,821 met inclusion criteria. Propensity score matching resulted in treated and untreated arms for the epinephrine and ETI contrasts with satisfactory overlap with regard to propensity scores. Receipt of epinephrine was associated with a propensity score adjusted odds ratio (OR) of 0.15 (95% confidence interval (CI) 0.11–0.19) for survival, while ETI was associated with an OR of 1.97 (95% CI 1.62–2.40) for survival. The propensity score adjusted OR expressing the association between epinephrine and ROSC was 0.50, 95% CI 0.42–0.60, while the OR for ETI and  101 ROSC was 2.31, 95% CI 1.92–2.77. Stratification based on post-arrest care availability yielded similar results.  Receipt of epinephrine following OHCA is associated with a substantial increase in mortality, while ETI is associated with increased survival. This calls into question current resuscitation guidelines, and highlights the need for randomized controlled trials to definitively assess the efficacy of epinephrine and ETI in out-of-hospital cardiac arrest.  4.2 Background   Clinical treatments for non-traumatic out-of-hospital cardiac arrest (OHCA) are guided by resuscitation guidelines generated by the International Liaison Committee on Resuscitation (ILCOR), the American Heart Association (AHA), and national resuscitation councils that propose strategies and care based on the available evidence. Since 2005 [10-15], guidelines have been reviewed and updated every 5 years and published collaboratively. Given the cost and complexity of performing OHCA research, much of the evidence reviewed during these update cycles has been based on observational studies and, as a result, potentially compromised because of residual bias.  The effects of epinephrine and tracheal intubation (ETI) remain unclear owing to the lack of high quality randomized trial evidence. In the 2015 AHA resuscitation update, it was noted that no high-quality studies had compared ETI with bag-mask or bag-mask and supraglottic airway support, while recommendations related to epinephrine were based on one small study which found favourable effects on return of spontaneous circulation (ROSC) and short-term survival [19, 34]. In both instances, AHA recommendations  102 provided a weak endorsement of these treatment modalities while calling for randomized clinical trials to provide more definitive evidence. Currently clinicaltrials.gov lists 158 studies involving epinephrine in OHCA, and 38 studies involving intubation. Two clinical trials [45, 46] are currently enrolling patients in studies of ETI vs. supraglottic airways. Despite bag-valve mask being used in 16% of cases [81], demonstrating that it is possible to manage the airway in OHCA in a non-invasive manner, no studies are ongoing or registered to examine the effect of bag-valve mask ventilation. For both epinephrine and ETI, there is a substantial gap in the ongoing effort to understand the benefit of these clinical interventions in the context of other recommended interventions.   Non-experimental methods for determining efficacy of therapy may play a role in providing clarity on these important issues. Propensity score methods, including propensity score matching, are well-accepted means of unconfounding relationships based on observational data [82]. We estimated the effects of prehospital epinephrine and ETI in OHCA resuscitation while accounting for patient characteristics and other factors using propensity score analysis methods.   4.3 Methods The study population included a cohort of adult (>19 years) non-traumatic OHCA cases identified by the Resuscitation Outcomes Consortium (ROC) and enrolled in the Epistry registry [61, 62] between December 2005 and May 2007. Briefly, the ROC is an international resuscitation research network of 10 research sites and one coordinating  103 center funded to conduct large randomized clinical trials in cardiac arrest and trauma resuscitation [62, 83].   4.3.1 Population and Setting  The ROC Epistry registry covered a population of approximately 25 million individuals in mostly urban and metropolitan centers in seven US states and two Canadian provinces.   4.3.2 Determinants, outcomes, and covariates  The ROC used the Utstein reporting template [64] to collect information based on a consistent set of data definitions. Standard information and time stamps were collected for each of the following: bystanders (arrest witnessed, cardiopulmonary resuscitation (CPR), attempted automatic external defibrillator (AED) deployment, and defibrillation); Basic Life Support (BLS) activities (arrival at the scene, EMS CPR by BLS providers, AED rhythm analysis, and AED defibrillation); ALS activities (arrival at the scene, EMS CPR by ALS providers, rhythm analysis, defibrillation, drug therapy, and ETI including endotracheal and nasal tracheal intubation); and arrival at hospital. EMS agencies reported the source of the time stamps, and if the time stamps were standardized or synchronized in their system. Trained EMS researchers reconciled the time records [62]. Based on the time stamps, and review of the medical records, trained EMS researchers entered the numbered sequence of specific resuscitation interventions as part of routine data abstraction (Figure 2.1).   104 The effects of prehospital epinephrine and ETI were estimated within a model that included clinical elements within the chain of survival [4], namely,: bystander attempted resuscitation, including bystander CPR, and deployment or use of a public access defibrillator; performance of EMS CPR prior to EMS rhythm analysis; placement of ETI; administration of prehospital epinephrine; and availability of specialized post-arrest care, including prehospital hypothermia, or transport to a hospital capable of delivering percutaneous coronary intervention (PCI) or having an electrophysiology lab.  4.3.3 Outcomes  The primary outcome was survival to hospital discharge and the secondary outcome was ROSC upon arrival to the emergency department.    4.3.4 Analytic methods  Descriptive statistics, including frequencies, proportions, means (standard deviation, SD), and medians (interquartile range IQR) were calculated. Potential confounders were identified from a list of candidate variables (12), which included population demographics, EMS system factors, and standard Utstein variables [66]. Additional subgroup analyses were carried out following stratification by specialized post-arrest care in order to address survival bias (12). Survival bias in this context refers to the artefactual protective effect of post-arrest care that occurs because patients eligible for specialized post-arrest care in hospital are those who have survived the initial episode of cardiac arrest and been transported to hospital. By performing our analysis of prehospital  105 epinephrine and ETI within subgroups stratified by post-arrest care, we eliminated potential survival bias associated with post-arrest care services. Propensity score matched analytic cohorts were created for assessing the effects of both prehospital epinephrine and ETI. Propensity to receive either prehospital epinephrine or ETI was assessed by logistic regression with all candidate variables in the model. To maximize the analytic cohort size, continuous variables were recoded into categorical variables, with indicator (dummy) variables created for missing values. We used caliper matching to create our cohorts of treated and untreated study subjects. Caliper matching involves matching treated and untreated study subjects using a percentage of the propensity score of the treated subjects as a matching yardstick (or caliper).  We started at 20% of the propensity score and reduced the width of the caliper in an iterative manner until a satisfactory overlap was achieved [84, 85]. All data management, manipulation, and analysis were performed using IBM SPSS Version 24.0 (IBM, Armonk, NY).  4.3.5 Ethics approval  The University of British Columbia Behavioral Research Ethics Board provided approval for this study. In addition, the ROC obtained ethics approval from all participating sites for the original data collection and subsequent secondary analyses [62].   4.4 Results The ROC Epistry registry enrolled 28,484 cases of OHCA during the study period. A total of 15,397 cases were excluded because resuscitation was not attempted, EMS witnessed the arrest, the patient was <20 years old, or because data were missing since a  106 non-ROC unit responded to the call first (Figure 2.2). The final study population included 12,821 cases of non-traumatic OHCA of whom 972 (7.6%) survived to hospital discharge.     Differences in the characteristics of all subjects in the study population and those who received epinephrine and had ETI are shown in Table 4.1. Epinephrine was used more frequently in older OHCA cases compared with younger cases (24.8% vs. 22.4%), females compared with males (24.5% vs. 23.0), cases found to have an obvious cause (30.1% vs. 23.0), patients treated in Canada (26.8% vs. 19.9%), and in those transported to hospital NOT offering specialized post-arrest care (26.8% vs. 17.9%). ETI was performed more frequently in patients with VF/FT as the initial rhythm (81.1% vs. 77.1%), while epinephrine was administered more frequently for cases in asystole (80.8% vs. 74.8%).  Epinephrine and ETI were commonly co-administered, with nearly 76.5% of cases receiving both epinephrine and ETI; approximately 16.7 % of cases received one intervention, but not the other.    Survivors were younger (median age 60.0 years vs. 67.0 years), and more likely to have been female (8.3% vs. 6.2%). They were also more likely to have arrested in a public location (18.5% vs. 5.5%), had a witnessed arrest (13.1% vs. 3.6%), occurred in the United States (9.4% vs. 5.9%), had VF/VT as the initial rhythm (20.2% survival), bystander-attempted resuscitation (11.3% vs. 5.6%), had not received epinephrine (17.9% vs. 4.4%), and had been transported to a hospital with specialized post-arrest care capability (14.5% vs. 3.6%) (Table 4.2). Among those who received epinephrine, crude  107 survival was 4.4%, whereas this rate was 7.5% among those who received ETI (compared with a 7.6% rate in the overall cohort). There was a substantial crude survival advantage among cases who received ETI but not epinephrine (37.3%) and this was not observed among cases who received epinephrine but not ETI (4.6%).   Supplemental Tables 4.1 and 4.2 display associations between patient characteristics and other factors and odds ratios for receiving epinephrine and ETI in the full cohort and in the propensity score matched cohorts. The epinephrine propensity matched cohort included 3,688 cases of OHCA, 1,844 in each arm (including 61% of the epinephrine untreated cases in the full cohort). The ETI propensity matched cohort included 3,768 cases, 1,884 in each arm (including 58.1% of the untreated cases in the full cohort). In both the epinephrine and ETI propensity score matched cohorts, high levels of overlap were achieved for the chain-of-survival clinical treatments of bystander-attempted resuscitation, EMS CPR prior to rhythm analysis, specialized post-arrest care, and epinephrine administration and ETI (Figure 4.1).  Table 4.3 displays unadjusted odds ratios (OR) in the full cohort, and the ORs from the propensity score matched analysis for epinephrine administration and ETI placement for the primary outcome of survival at hospital discharge and also for ROSC present at presentation to the emergency department. For epinephrine, the OR for survival and ROSC upon presentation to the emergency department in the propensity score matched analyses were lower than the unadjusted OR (survival OR 0.15, 95% CI 0.11–0.19 vs.  108 0.21, 95% CI 0.19–0.24, respectively; and OR for ROSC 0.50, 95% CI 0.42–0.60 vs. 0.86, 95% CI 0.77–0.96, respectively).    Results of stratified analyses based on availability of specialized post-arrest care are also shown in Table 4.3 and Supplemental Tables 4.3 to 4.6. Propensity score matched OR of survival following epinephrine administration in the stratified subpopulations were mostly similar to the propensity score matched OR of survival in the full unstratified cohort with OR varying from 0.14 to 0.22. Similarly, results of the propensity score matched analyses for the effect of epinephrine on ROSC were similar in the stratified and unstratified cohorts (with OR varying from 0.46 to 0.52). However, there were some differences in the results of analyses on the effect of ETI stratified by availability of post-arrest care and the results of analyses based on the full cohort. Nevertheless, all estimates showed a beneficial effect of ETI on both survival and ROSC, although OR ranged from 1.11 to 3.24 for survival and from 1.63 to 3.32 for ROSC.       4.5 Discussion Our study examined the effects ETI and epinephrine, two specific clinical interventions that are central to the chain of survival and current OHCA resuscitation practice, in multifactorial models that included other clinical interventions. In the study cohort, ETI and epinephrine were frequently administered together. The risk of death was elevated nearly 7 times as a result of receiving epinephrine, while the odds of ROSC at the emergency department was reduced by half among those who received epinephrine. ETI was associated with a near doubling of survival, with a similar increase associated with  109 ROSC. Stratification by availability of specialized post-arrest care resulted in approximately similar effects for epinephrine and ETI with regard to both survival and ROSC.   The data used in this study were collected as part of an observational study of OHCA resuscitation [62] in the absence of any experimental intervention.  They represent the experience of OHCA resuscitation across a large segment of the North American population. Since OHCA resuscitation is based on guidelines, with clinical interpretation and adaptation for clinical contingencies, these data include potential artifacts resulting from the clinical flow that occurs during resuscitation, including the clinician’s response to clinical information. Perhaps the most important bias to consider in this context is confounding by indication. By including the main clinically prognostic variables, including other interventions associated with the chain-of-survival, and propensity score methods, we attempted to mitigate any bias arising from such confounding.  Additionally, based on previous work by our group that identified an existing survival bias [68], we stratified our results by availability of specialized post-arrest care. Support for the validity of our study findings is also forthcoming from the survival rates associated with various patient characteristics and other factors evaluated in our analysis. Thus, patient age, public location of arrest, bystander-witnessed arrest, bystander resuscitation, and response times all showed expected patterns with regards to survival.      Differences between population characteristics in our study vs previously published  110 research on both ETI and epinephrine administration make comparisons between our findings and previous studies challenging. In our ETI analysis, we compared ETI to bag-mask ventilation or supraglottic airway placement, thereby contrasting currently used advanced airway techniques vs. basic airway techniques. In a 2014 systematic review and meta-analysis, Fouche et al.[86] sought to establish the effect of advanced airway management to basic airway management in OHCA. Their study population included OHCA cases among adults and cases less than 18 years of age and both traumatic and non-traumatic arrest cases. Their study found decreased short- and longer-term survival as a result of treatment with advanced airway interventions compared with basic airway interventions. Fouche et al. suggested that confounding by indication resulting from advanced airways being placed in patients with worse prognostic factors likely biased their results. A systematic review by Benoit et al.[28] of ETI vs. supraglottic airways, and a propensity score matched analysis by McMullan et al.[87] found favourable survival resulting from ETI placement compared with supraglottic airway.   Patanwala et al.[53] conducted a systematic review of the effect of epinephrine after cardiac arrest and concluded that epinephrine administration in OHCA was not associated with survival. Two propensity score matched analyses [80, 88] of non-traumatic OHCA demonstrated survival and ROSC effects similar to those found in our analysis, while a propensity score matched analysis involving a basic life support system in Japan found favourable survival resulting from epinephrine administration.    111 Differences in the effects of epinephrine and ETI on ROSC and survival also lead to challenges in the interpretation of our results. Effects of epinephrine were substantially more negative with regard to survival than with regard to ROSC. Similarly, ETI had a slightly greater beneficial effect on ROSC than on survival. This suggests the short-term effects of these interventions on ROSC are somewhat more positive than the slightly longer-term effects on survival. More importantly, our application of propensity score analysis methods did not reverse the strongly negative effect associated with epinephrine use. In fact, the effect of epinephrine as assessed using propensity score methods in this study was similar to the effect assessed using standard regression techniques (12).       Propensity score methods, while widely accepted in the social sciences and epidemiologic studies, remain uncommon in OHCA resuscitation research. These methods are particularly useful in clinical situations where randomized trials are difficult to carry out. In the one RCT of either epinephrine or ETI comparing intervention to placebo published to date [34], the planned patient accrual schedule was never achieved, underscoring the substantial challenges involved in studying OHCA using standard experimental designs. In contrast, causal inference methods are available for use with observational data, and propensity score methods are among the most accepted of such approaches [89]. However, it is recognized that propensity score methods can only address confounding by indication if the strength of the indication can be adequately quantified.      112 In the most current resuscitation guidelines [18, 19], epinephrine continues to be endorsed as a beneficial treatment modality, while ETI is cautiously included. Our results, augmented by those of Dumas et al [80], several systematic reviews [36, 38], and the randomized trial by Jacobs et al [34], all suggest that the cautious endorsement is better directed towards epinephrine. Additionally, ETI deserves a stronger endorsement given the findings of Voss et al [81] (which show that over 86% of patients require advanced airway management to protect the airway or ensure adequate ventilation/ oxygenation), and of systematic reviews (which show favorable survival for ETI over SGA and basic airway management [28, 87]). Although evidence-based guidelines are ideally supported by the highest-level evidence (viz., by meta-analysis of high quality RCTs), the absence of such evidence requires that evidence-based guidelines are based on robust evidence from the meta-analysis of high quality observational studies.  The strengths of our study include its size and population-based provenance. The meticulous data collection systems in place helped to reduce misclassification bias. Including the clinical elements of the chain of survival provided a more complete context for modeling the effect of epinephrine and ETI. Our propensity score analysis methodology, which ensured that there was overlap between the treated and untreated groups in terms of propensity for treatment served to minimize confounding. Even though we used different covariates in propensity matching process, our results were mostly consistent with the results of other studies [80].     113 Limitations of our study include the non-experimental design, although it is noteworthy that our results are similar to the only randomized trial of epinephrine vs. placebo to date [34]. By stratifying our analysis on specialized post-arrest care, we further limited the potential for survival bias. The age of our data (2005 – 2007) may be considered a potential limitation as changes in practice could lead to a reduction in the current relevance of our findings. Of note, the American Heart Association reported that the 2011 Resuscitation Outcomes Consortium survival from a similarly defined cohort was approximately 10% [2]. While survival in a similar and more recent cohort shows improvement, the recommendation for administration of epinephrine and provision of ETI have remained in the resuscitation guidelines from 2005 to 2015, and practice as it relates to these elements of OHCA resuscitation has not changed. The availability and type of specialized post-arrest care has changed substantially over time, and in some locations includes the availability of specialized interventions such as extracorporeal membrane oxygenation (ECMO). However, our stratified analysis, which controlled for availability of such care, demonstrated similar effects for epinephrine regardless of the availability of post-arrest care. 4.6 Conclusion  Our study showed that epinephrine has a modest negative effect on ROSC and a strong negative effect on survival, whereas ETI has a beneficial effect on both ROSC and survival. This calls into question current resuscitation guidelines, and highlights the need  for randomized controlled trials to definitively assess the efficacy of epinephrine and ETI in out-of-hospital cardiac arrest.   114  Figure 4.1. Box plots of propensity scores (raw vs. matched) for epinephrine and tracheal intubation treated cases and controls. 115  Table 4.1 Distribution of patient characteristics and other factors among those who did not receive epinephrine and endotracheal intubation (ETI).  Characteristic Number (%)*    Epinephrine ETI Full cohort  Not treated Not treated Cases N (%) 12,821 (100) 3,017 (23.5) 3,242 (25.3) Age in years, median (IQR*) 66.0 (25.0) 67.0 (25.0) 66.0 (23.0) Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)    927 (7.2) 4,046 (31.6) 3,883 (30.3) 3,965 (30.9)  208 (22.4) 913 (22.6) 913 (23.5) 983 (24.8)  242 (26.1) 1,027 (25.4) 973 (25.1) 1,000 (25.2) Sex  Male gender  Female gender   8,366 (65.3) 4,455 (34.7)   1,924 (23.0) 1,093 (24.5)   1,167 (26.2) 2,075 (24.8) Location of arrest    Non-public     Public    10,743 (83.8) 2,078 (16.2)   2,480 (23.1) 537 (25.8)   459 (22.1) 2,783 (25.9)  Bystander-witnessed status    Witnessed     Not witnessed    5,359 (41.8) 7,462 (58.2)  1,250 (23.3) 1,767 (23.7)  2,100 (28.1) 1,142 (21.3) Classification of arrest cause      Assumed cardiac      Other, obvious cause  11,882 (92.7) 939 (7.3)  2,734 (23.0) 283 (30.1)  2,973 (25.0) 269 (28.6) 116    Country    Canada    United States  6,777 (52.9) 6,044 (47.1)  1,815 (26.8) 1,202 (19.9)   1,794 (26.5) 1,448 (24.0) Initial rhythm    VF/VT ***   PEA ^   Asystole    AED ^^– no shock advised/unknown   3,094 (24.1) 2,287 (17.8) 5,396 (42.1) 2,044 (15.9)  708 (22.9) 429 (18.8) 1,037 (19.2) 843 (41.2)  584 (18.9) 419 (18.3) 1,361 (25.2) 879 (43.0) Response time   Median (IQR)   5:00 (2:00)  5:00 (3:00)  5:00 (3:00) Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown     7,804 (60.9) 4,076 (31.8) 676 (5.3) 138 (1.1) 127 (0.9)  1,749 (22.4) 971 (23.8) 221 (32.7) 42 (30.4) 34 (26.8)  1,852 (23.7) 1,072 (26.3) 228 (33.7) 59 (42.8) 31 (24.4) Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   8,309 (64.8) 4,512 (35.2)   1,972 (23.7) 1,045 (23.2)  2,226 (26.8) 1,016 (22.5) EMS CPR/Rhythm analysis sequence     CPR preceded rhythm analysis     CPR did not precede rhythm analysis   2,266 (17.7) 10,555 (82.3)  2,381 (23.5) 536 (23.7)   670 (29.6) 2,572 (24.4) Tracheal intubation    117     Yes    No  3,242 (25.3) 9,579 (74.7) 872 (9.1) 2,145 (66.2)            - 3,242 (25.3) Prehospital epinephrine administered    Yes    No   9,804 (76.5) 3,017 (23.5)             - 3017 (23.5)  1,097 (11.2) 2,145 (71.1) Specialized post-arrest care provided^^^   Yes   No    4,681 (36.5) 8,140 (63.5)   837 (17.9) 2,180 (26.8)  2,370 (29.1) 872 (18.6)  *Unless otherwise noted. ** Interquartile range (IQR). ***VF/VT (ventricular fibrillation/ventricular tachycardia)  ^PEA (pulseless electrical activity)  ^^AED (automatic external defibrillator) ^^^ Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.118  Table 4.2 Survival rates by patient characteristics, interventions, and other factors.  Characteristic Full cohort Received Epinephrine n=9,804 Received ETI n=9,579  Number (%)* Did not survive Survived Survived Survived Cases N (%) 11,849 (92.4) 972 (7.6%) 433 (4.4) 722 (7.5) Age in years, median (IQR)** 67.0 (28.0) 60.0 (21.0) 58 (23.0) 60.0 (21.0) Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)    841 (90.7) 3,615 (89.3) 3,580 (92.2) 3,813 (96.2)  86 (9.3) 431 (10.7) 303 (7.8) 152 (3.8)  53 (7.4) 187 (6.0) 126 (4.2) 67 (2.2)  70 (10.2) 312 (10.3) 227 (7.8) 113 (3.8) Patient sex  Male  Female   4,181 (93.3) 7,668 (91.7)  274 (6.2) 698 (8.3)  300 (4.7) 133 (4.0)   209 (6.4) 513 (8.2) Location of arrest   Non-public   Public      10,156 (94.5) 1,693 (81.5)   587 (5.5) 385 (18.5)   286 (3.5) 147 (9.5)   449 (5.6) 273 (16.9)  Bystander-witnessed status    Witnessed     Not witnessed   4,656 (86.9) 7,193 (96.4)   703 (13.1) 269 (3.6)   305 (7.4)  128 (2.2)  530 (12.6) 192 (3.6) Classification of arrest cause     Assumed cardiac     Other, obvious cause  11,849 (92.4) 863 (91.9)  896 (7.5) 76 (8.1)  409 (4.5) 24 (3.7)  668 (7.5) 54 (8.1) 119   Country    Canada    United States   6,375 (94.1) 5474 (90.6)  402 (5.9) 570 (9.4)  151 (3.0) 282 (5.8)  289 (5.8) 433 (9.4) Initial rhythm    VF/VT***   PEA^   Asystole   AED^^ – no shock advised/unknown   2,470 (79.8) 2,159 (94.4) 5,338 (98.9) 1,882 (92.1)  624 (20.2) 128 (5.6) 58 (1.1) 162 (7.9)  297 (12.4) 58 (3.1) 43 (1.0) 35 (2.9)   485 (19.3) 105 (5.6) 51 (1.5) 81 (6.9) Response time   Median (IQR)   5:00 (2:00)  4:00 (3:00)  4:00 (3:00)  4:00 (3:00) Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown   7,099 (91.0) 3,848 (94.4) 649 (96.0) 135 (97.8) 118 (92.9)  705 (9.0) 228 (5.6) 27 (4.0) 2 (2.2) 9 (7.1)  311 (5.1) 109 (3.5) 8 (1.8) 1 (1.0) 4 (4.3)  533 (9.0) 168 (5.6) 13 (2.9) 2 (1.8) 6 (6.2) Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   7,845 (94.4) 4,004 (88.7)   464 (5.6) 508 (11.3)   216 (3.4) 217 (6.3)  341 (5.6) 381 (10.9) EMS CPR/Rhythm analysis sequence    CPR preceded rhythm analysis    CPR did not precede rhythm analysis    9,774 (92.6) 2,075 (91.9)    781 (7.4) 191 (8.4)    74 (4.3) 359 (4.4)   114 (7.1) 608 (7.6)  120  Tracheal intubation performed    Yes    No   8,857 (92.5) 2,992 (92.3)   722 (7.5) 250 (7.7)  36 (3.3) 397 (4.6)  722 (7.5)        -  Prehospital epinephrine administered    Yes    No   9,371 (95.5) 2,478 (82.1)  433 (4.4) 539 (17.9)  433 (4.4)        -   397 (4.6) 325 (37.3) Specialized post-arrest care provided ^^^   Yes   No     4,001 (85.5) 7,848 (96.4)    680 (14.5) 292 (3.6)   103 (1.7) 330 (8.6)   198 (3.4) 524 (13.8)  *Unless otherwise noted. ** Interquartile range (IQR). ***VF/VT (ventricular fibrillation/ventricular tachycardia)  ^PEA (pulseless electrical activity)  ^^AED (automatic external defibrillator) ^^^ Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.121   Table 4.3. Logistic regression model showing unadjusted and propensity score matched effect of epinephrine and tracheal intubation (ETI) on survival at discharge from hospital and return of spontaneous circulation present at emergency department among different populations.  Intervention   Population Outcome Unadjusted Propensity score matched OR (95% CI) p-value  OR (95% CI)  p-value Epinephrine   All patients Survival 0.21 (0.19 – 0.24) <0.001 0.15 (0.11 – 0.19) < 0.001 ROSC present at ED 0.86 (0.77 – 0.96) 0.006 0.50 (0.42 – 0.60) < 0.001 Post-arrest care available Survival 0.13 (0.11 – 0.16) <0.001 0.22 (0.17 – 0.29) < 0.001 ROSC present at ED 0.54 (0.46 – 0.63) < 0.001 0.52 (0.42 – 0.65) < 0.001 Post-arrest care not available  Survival 0.19 (0.15 – 0.24) <0.001 0.14 (0.08 – 0.23) < 0.001 ROSC present at ED 0.96 (0.81 – 1.14) 0.65 0.46 (0.34 – 0.63) < 0.001 ETI    All patients Survival 0.98 (0.84 – 1.13) 0.75 1.97 (1.62 – 2.40) < 0.001 ROSC present at ED 2.42 (2.11 – 2.77) < 0.001 2.31 (1.92 – 2.77) < 0.001 Post-arrest care available Survival 0.73 (0.60 – 0.89) 0.75 1.11 (0.87 – 1.42) 0.42 ROSC present at ED 1.84 (1.53 – 2.21) < 0.001 1.63 (1.29 – 2.05) < 0.001 Post-arrest care not available  Survival 0.86 (0.67 – 1.11) 0.24 3.24 (2.22 – 4.74) < 0.001 ROSC present at ED 2.44 (1.97 – 3.01) < 0.001 3.32 (2.39 – 4.61) < 0.001 122   Supplemental Table 4.1. Unadjusted and propensity score matched odds ratios of receiving epinephrine, and standardized difference and variance ratio of propensity score matched cases*.  Characteristic Unadjusted Odds Ratios (95% CI) p-value Propensity Score Matched OR (95% CI) p-value Standardized difference in PS matched groups Variance Ratio Cases N (%) 12,821 (100) Untreated cases  n=3017  n=3,688 1844 per arm    Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)    Reference 0.90 (0.72 – 1.12) 0.83 (0.67 – 1.04) 0.72 (0.58 – 0.91)    0.35 0.12 0.005   Reference 1.03 (0.78 – 1.35) 1.09 (0.82 – 1.43) 1.04 (0.79 – 1.71)    0.86 0.57 0.79   -0.020 0.048 -0.008   0.903 1.009 0.972  Patient sex  Male   Female    1.18 (1.06 – 1.31) Reference   0.004  1.07 (0.93 – 1.23) Reference  0.34  0.024   1.022   Location of arrest   Non-public    Public    Reference 0.71 (0.61 – 0.82)    < 0.001  Reference 0.97 (0.81 – 1.17)    0.74    0.008    0.993  Bystander-witnessed status    Witnessed     Not witnessed   0.88 (0.79 – 0.99) Reference   0.03  1.06 (0.91 – 1.22) Reference   0.45   0.036   0.992  Classification of arrest cause       Assumed cardiac  1.74 (1.43 – 2.12)  < 0.001  1.13 (0.88 – 1.46)  0.33  0.012  1.136 123        Other, obvious cause  Reference  Reference     Country    Canada    United States  Reference 1.30 (1.16 – 1.47)   < 0.001   Reference 1.01 (0.87 – 1.18)    0.87    0.016    1.000  Initial rhythm    VF/VT ^   PEA^^   Asystole   AED^^^ – no shock advised/unknown   0.54 (0.47 – 0.63) 0.75 (0.64 – 0.88) Reference 0.42 (0.37 – 0.49)  < 0.001 < 0.001  < 0.001  1.06 (0.88 – 1.27) 1.07 (0.88 – 1.30) Reference 1.11 (0.91 – 1.34)   0.57 0.50  0.31   0.044 0.032  0.064  0.958 0.967  1.043  Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes >16 minutes Unknown   Reference 1.08 (0.93 – 1.21) 0.81 (0.65 – 1.02) 1.31 (0.83 – 2.08) 0.65 (0.40 – 1.08)   0.19 0.07 0.24 0.10  Reference 1.03 (0.89 – 1.19) 0.95 (0.70 – 1.29) 1.15 (0.64 – 2.05) 0.78 (0.39 – 1.55)   0.72 0.74 0.65 0.48    0.016 -0.052 0.128 -0.238   0.991 1.064 0.923 1.250  Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   Reference 0.89 (0.79 – 0.99)    0.04  Reference 0.93 (0.81 – 1.07)    0.94    -0.048    1.022  EMS CPR/Rhythm analysis sequence    CPR preceded rhythm analysis    CPR did NOT precede rhythm                                                  analysis   Reference 0.72 (0.71 – 0.94)  0.005  Reference 1.01 (0.85 – 1.20)    0.94    0.000    0.994  Tracheal intubation    Yes  20.8 (18.6 – 23.2)  < 0.001  0.98 (0.85 – 1.12)  0.73  0.000  1.000 124      No  Reference Reference     Specialized post-arrest care provided*^   Yes   No    1.25 (1.09 – 1.42) Reference   0.001  1.12 (0.95 – 1.31) Reference   0.19   0.080   0.966   * Standard deviation of the means of the natural log transformed predicted values of receiving epinephrine = 0.49. Propensity score matching caliper distance = 0.007. ^VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^PEA (pulseless electrical activity)  ^^^AED (automatic external defibrillator) *^ Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.   125   Supplemental Table 4.2. Unadjusted and propensity score matched odds ratios of receiving tracheal intubation, and standardized difference and variance ratio of propensity score matched cases*.  Characteristic Unadjusted Odds Ratios (95% CI) p-value Propensity Score Matched Or (95% CI) p-value Standardized difference in PS matched groups Variance Ratio Cases N (%) 12,821 (100) Untreated cases n=3,242  n=3,768  1,884 per arm    Patient age    Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Reference 0.98 (0.79 – 1.20) 1.08 (0.87 – 1.33) 1.20 (0.97 – 1.48)   0.82 0.50 0.09   Reference 0.96 (0.74 – 1.25) 0.97 (0.74 – 1.27) 1.04 (0.79 – 1.36)    0.75 0.97 0.79    -0.028 -0.016 0.040    1.013 1.010 0.976 Sex  Male gender  Female gender   0.95 (0.86 – 1.06) Reference   0.37  0.97 (0.85 – 1.12) Reference  0.70  -0.008  0.991 Location of arrest   Non-public      Public  Reference 1.23 (1.06 – 1.43)   0.007  Reference 1.07 (0.90 – 1.29)   0.44   0.072   0.946  Bystander-witnessed status    Witnessed     Not witnessed   1.28 (1.15 – 1.44) Reference  0.007  1.10 (0.96 – 1.27) Reference   0.18  0.064  0.984 Classification of arrest cause       Assumed cardiac  0.82 (0.67 – 1.00)  0.048  0.85 (0.67 – 1.09)  0.21  -0.012  0.883 126        Other, obvious cause  Reference  Reference Country    Canada    United States  Reference 0.67 (0.60 – 0.76)    < 0.001  Reference 0.98 (0.84 – 1.14)   0.80    0.012   1.000 Initial rhythm    VF/VT ^   PEA^^   Asystole   AED^^^ – no shock advised/unknown   1.54 (1.33 – 1.79) 1.51 (1.29 – 1.76) Reference 0.67 (0.58 – 0.77)  < 0.001 < 0.001  < 0.001  0.98 (0.81 – 1.17) 1.13 (0.93 – 1.36) Reference 0.96 (0.79 – 1.16)   0.79 0.22  0.67   -0.008 0.132  -0.040   1.005 0.903  1.036  Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.88 (0.78 – 0.98) 0.81 (0.65 – 1.01) 0.41 (0.27 – 0.63) 1.24 (0.72 – 2.13)    0.02 0.06 < 0.001 0.44  Reference 0.97 (0.84 – 1.13) 0.97 (0.71 – 1.31) 1.04 (0.58 – 1.86) 1.41 (0.68 – 2.89)    0.72 0.82 0.90 0.35   -0.028 -0.044 0.044 0.328    1.014 1.044 0.923 0.778 Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   Reference 1.37 (1.22 – 1.53)    < 0.001  Reference 1.06 (0.92 – 1.23)   0.39   0.048     0.978 EMS CPR/Rhythm analysis sequence     CPR preceded rhythm analysis      CPR did not precede rhythm analysis   1.53 (1.34 – 1.75) Reference  < 0.001   1.05 (0.89 – 1.14) Reference   0.60  0.004  1.012 Prehospital epinephrine administered       127      Yes    No  20.8 (18.6 – 23.1) Reference < 0.001 1.04 (0.90 – 1.19) Reference 0.61 0.000 1.000 Specialized post-arrest care provided*^   Yes   No    1.76 (1.55 – 1.99) Reference   < 0.001   1.12 (0.96 – 1.31) Reference   0.16  0.060  0.984 * Standard deviation of the mean of the natural log transformed predicted values of receiving tracheal intubation = 0.14.  Propensity score caliper distance = 0.03 ^VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^PEA (pulseless electrical activity)  ^^^AED (automatic external defibrillator) *^ Specialized post-arrest care: prehospital hypothermia, transport to hospital with electrophysiology lab, or transport to hospital with a cardiac catheter lab.   128   Supplemental Table 4.3. Unadjusted and propensity score matched odds ratios of receiving epinephrine for cases transported to hospital capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*.  Characteristic Unadjusted Odds Ratios (95% CI) p-value Propensity Score Matched Or (95% CI) p-value Standardized difference in PS matched groups Variance Ratio Cases N (%) n= 4,681 Untreated cases n= 837  n=1,514  757 per arm    Patient age **   Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Reference 1.06 (0.77 – 1.46) 1.03 (0.74 – 1.42) 1.17 (0.84 – 1.64)    0.71 1.03 1.17  Reference 0.76 (0.51 – 1.12) 0.75 (0.50 – 1.12) 0.71 (0.47 – 1.08)   0.17 0.16 0.11    -0.012 -0.020 -0.052    1.004 1.014 1.034 Sex  Male gender  Female gender   1.24 (1.05 – 1.48) Reference   0.14  0.98 (0.79 – 1.22) Reference  0.85  -0.012  0.987 Location of arrest   Non-public     Public      Reference 0.67 (0.56 – 0.81)     < 0.001  Reference 0.97 (0.77 – 1.14)    0.82   -0.028    1.020  Bystander-witnessed status    Witnessed     Not witnessed   0.69 (0.58 – 0.83) Reference  < 0.001  1.02 (0.81 – 1.18) Reference   0.89  -0.020  1.000 Classification of arrest cause       129         Assumed cardiac      Other, obvious cause  1.52 (1.11 – 2.07) Reference  0.009 1.07 (0.72 – 1.60) Reference 0.73 -0.004 0.972 Country    Canada    United States  Reference 1.50 (1.24 – 1.81)    < 0.001  Reference 0.94 (0.74 – 1.19)   0.12   0.080   0.996 Initial rhythm    VF/VT ^   PEA^^   Asystole   AED^^^ – no shock advised/unknown   0.40 (0.31 – 0.50) 0.47 (0.36 – 0.50) Reference 0.29 (0.22 – 0.37)  < 0.001 < 0.001  < 0.001  0.91 (0.67 – 1.13) 0.96 (0.69 – 1.33) Reference 0.98 (0.71 – 1.34)  0.53 0.80  0.88   -0.004 -0.044  0.008   1.017 0.994  0.989   Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown   Reference 1.25 (1.04 – 1.51) 1.31 (0.83 – 2.07) 3.01 (1.03 – 8.77) 0.98 (0.50 – 1.95)    0.02 0.25 0.04 0.96  Reference 0.89 (0.70 – 1.13) 0.99 (0.55 – 1.78) 0.22 (0.02 – 2.01) 0.76 (0.28 – 2.06)    0.35 0.10 0.18 0.59   -0.064 0.044 -1.500 -0.252    1.043 0.935 5.000 1.333 Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   Reference 0.86 (0.72 – 1.03)    < 0.001  Reference 0.96 (0.76 – 1.19)    0.66   -0.040   0.102 EMS CPR/Rhythm analysis sequence     CPR preceded rhythm analysis      CPR did not precede rhythm analysis  0.83 (0.68 – 1.01) Reference  0.06   0.99 (0.77 – 1.26) Reference  0.91  0.000  1.000 Tracheal intubation       130    Yes   No  5.02 (4.19 – 6.02) Reference < 0.001 1.00 (0.80 – 1.16) Reference 0.97 0.000 1.000  * Standard deviation of the mean of the natural log transformed predicted value of receiving epinephrine for cases transported to hospital capable of delivering specialized post-arrest care = 0.21. Propensity score matching caliper distance = 0.01 ^VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^PEA (pulseless electrical activity)  ^^^AED (automatic external defibrillator)   131   Supplemental Table 4.4. Unadjusted and propensity score matched odds ratios for epinephrine for cases that were not transported to hospitals capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*.  Characteristic Unadjusted Odds Ratios (95% CI) p-value Propensity Score Matched Or (95% CI) p-value Standardized difference in PS matched groups Variance Ratio Cases N (%) n= 8,140 Untreated cases n= 2,180  n=1,878  939 per arm    Patient age **   Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Reference 0.74 (0.54 – 1.02) 0.65 (0.47 – 0.89) 0.48 (0.35 – 0.67)    0.06 0.008 < 0.001  Reference 1.07 (0.72 – 1.59) 0.98 (0.65 – 1.48) 1.06 (0.71 – 1.60)    0.76 0.93 0.77    0.036 -0.044 0.012    0.982 1.024 0.991  Sex  Male gender  Female gender   1.15 (0.98 – 1.33) Reference   0.08  1.10 (0.91 – 1.34) Reference  0.32  0.036  1.036 Location of arrest   Non-public     Public      Reference 0.77 (0.61 – 0.97)    0.03  Reference 1.15 (0.86 – 1.55)    0.36   0.156    0.871  Bystander-witnessed status    Witnessed     Not witnessed   1.11 (0.94 – 1.30) Reference  0.21  1.10 (0.89 – 1.35) Reference   0.39  -0.020  1.000 Classification of arrest cause       132         Assumed cardiac      Other, obvious cause  1.97 (1.50 – 2.58) Reference  < 0.001 1.02 (0.72 – 1.46) Reference 0.90 0.004 0.946 Country    Canada    United States  Reference 1.14 (0.96 – 1.35)    0.12  Reference 1.18 (0.96 – 1.45)   0.58   -0.012   0.980 Initial rhythm    VF/VT ^   PEA^^   Asystole   AED^^^ – no shock advised/unknown   0.58 (0.47 – 0.71) 0.89 (0.72 – 1.12) Reference 0.39 (0.32 – 0.47)  < 0.001 0.32  < 0.001  1.08 (0.83 – 1.40) 1.20 (0.90 – 1.59) Reference 0.99 (0.74 – 1.32)    0.60 0.22  0.92   0.064 0.152  -0.076   0.954 0.875 1.060  Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.92 (0.79 – 1.08) 0.60 (0.45 – 0.80) 0.96 (0.53 – 1.73) 0.47 (0.22 – 1.04)    0.30 < 0.001 0.88 0.06  Reference 1.07 (0.87 – 1.31) 1.05 (0.68 – 1.62) 1.42 (0.61 – 3.29) 1.27 (0.34 – 4.80)   0.54 0.84 0.42 0.73    0.008 0.000 0.262 0.225    0.995 1.000 0.786 0.800  Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   Reference 0.96 (0.82 – 1.12)    0.58  Reference 1.01 (0.83 – 1.23)   0.91   0.040     0.983 EMS CPR/Rhythm analysis sequence     CPR preceded rhythm analysis      CPR did not precede rhythm analysis  1.03 (0.84 – 1.28) Reference  0.77   0.81 (0.61 – 1.06) Reference   0.12  -0.028  0.846 133   Tracheal intubation  Yes   No   47.0 (40.6 – 54.6) Reference  < 0.001  0.97 (0.79 – 1.17) Reference  0.72  0.000  1.000  * Standard deviation of the mean of the natural log transformed predicted values of receiving epinephrine = 0.21.  Propensity score matching caliper distance = 0.007 ^VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^PEA (pulseless electrical activity)  ^^^AED (automatic external defibrillator)    134   Supplemental Table 4.5. Unadjusted and propensity score matched odds ratios of receiving tracheal intubation for cases transported to hospital capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*.  Characteristic Unadjusted Odds Ratios (95% CI) p-value Propensity Score Matched Or (95% CI) p-value Standardized difference in PS matched groups Variance Ratio Cases N (%) n= 4,681 Untreated cases n= 872  n=1,694 847 per arm    Patient age **   Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Reference 0.83 (0.61 – 1.12) 0.91 (0.67 – 1.25) 1.05 (0.76 – 1.46)   0.22 0.57 0.76   Reference 0.84 (0.59 – 1.22) 0.84 (0.58 – 1.22) 0.88 (0.60 – 1.29)    0.36 0.36 0.5    -0.028 -0.028 0.024   1.013 1.019 0.984  Sex  Male gender  Female gender   0.95 (0.80 – 1.13) Reference   0.76  0.89 (0.73 – 1.10) Reference  0.29  -0.044  0.966 Location of arrest   Non-public   Public      Reference 1.04 (0.85 – 1.27)    0.74  Reference 1.01 (0.79 – 1.29)    0.94   -0.016    1.012  Bystander-witnessed status    Witnessed     Not witnessed   1.53 (1.28 – 1.83) Reference  < 0.001  1.11 (0.89 – 1.38) Reference   0.36  0.044  0.983 Classification of arrest cause       135         Assumed cardiac      Other, obvious cause  0.74 (0.53 – 1.02) Reference  0.07 1.02 (0.69 – 1.50) Reference 0.94 0.000 1.000 Country    Canada    United States  Reference 0.54 (0.44 – 0.66)    < 0.001  Reference 0.98 (0.75 – 1.27)   0.87   0.000   1.007 Initial rhythm    VF/VT ^   PEA^^   Asystole   AED^^^ – no shock advised/unknown   1.39 (1.11 – 1.74) 1.47 (1.16 – 1.87) Reference 0.52 (0.42 – 0.65)  0.004 0.001  < 0.001  0.93 (0.71 – 1.23) 0.96 (0.72 – 1.29) Reference 0.83 (0.64 – 1.07)   0.62 0.79  0.16   0.000 0.048  -0.132   1.000 0.958  1.086   Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.83 (0.69 – 0.99) 0.77 (0.51 – 1.16) 0.52 (0.25 – 1.07) 1.40 (0.67 – 2.92)    0.04 0.21 0.08 0.37  Reference 0.96 (0.78 – 1.20) 0.85 (0.51 – 1.41) 1.03 (0.47 – 2.26) 0.73 (0.27 – 1.95)    0.74 0.53 0.94 0.53   -0.016 -0.152 0.000 -0.358    1.010 1.176 1.000 1.500  Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   Reference 1.27 (1.06 – 1.53)    0.01  Reference 1.00 (0.80 – 1.25)   0.97    0.004     0.995 EMS CPR/Rhythm analysis sequence     CPR preceded rhythm analysis      CPR did not precede rhythm analysis  1.58 (1.28 – 1.83) Reference  < 0.001   1.04 (0.84 – 1.28) Reference   0.75  0.012  1.009 136   Prehospital epinephrine administered    Yes    No   5.08 (4.24 – 6.01) Reference  < 0.001  1.11 (0.90 – 1.38) Reference  0.32  0.040  1.026 * Standard deviation of the mean of the natural log transformed predicted value of receiving tracheal intubation for cases transported to hospital capable of delivering specialized post-arrest care = 0.22. Propensity score matching caliper distance = 0.04. ^VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^PEA (pulseless electrical activity)  ^^^AED (automatic external defibrillator)   137   Supplemental Table 4.6. Unadjusted and propensity score matched odds ratios for tracheal intubation for cases that were not transported to hospitals capable of providing specialized post-arrest care, and standardized difference and variance ratio of propensity score matched cases*.   Characteristic Unadjusted Odds Ratios (95% CI) p-value Propensity Score Matched Or (95% CI) p-value Standardized difference in PS matched groups Variance Ratio Cases N (%) n= 8,140 Untreated cases n= 2,370  n=1,890 945 per arm    Patient age **   Adult (20 – 39 years)   Middle age (40 – 59 years)   Older (60 – 74 years)   Elderly (≥ 75 years)   Reference 1.18 (0.89 – 1.58) 1.35 (1.00 – 1.81) 1.59 (1.18 – 2.14)    0.26 0.05 0.002  Reference 0.83 (0.56 – 1.22) 0.83 (.056 – 1.14) 0.89 (0.60 – 1.33)    0.34 0.36 0.57   -0.028 -0.040 0.028    1.014 1.024 0.981 Sex  Male gender  Female gender   0.96 (0.83 – 1.11) Reference   0.58  0.96 (0.79 – 1.16) Reference  0.66  -0.008  0.991 Location of arrest   Non-public   Public      Reference 1.30 (1.04 – 1.63)    0.02  Reference 1.24 (0.93 – 1.65)    0.15   0.152    0.886  Bystander-witnessed status    Witnessed     Not witnessed   1.06 (0.91 – 1.24) Reference  0.45  0.93 (0.76 – 1.14) Reference   0.49  -0.024  1.004 Classification of arrest cause       138         Assumed cardiac      Other, obvious cause  0.79 (0.61 – 1.04) Reference  0.09 0.95 (0.67 – 1.14) Reference  0.95 -0.008 0.992 Country    Canada    United States  Reference 0.78 (0.66 – 0.91)    0.002  Reference 1.05 (0.81 – 1.29)   0.66   0.016   0.996 Initial rhythm    VF/VT ^   PEA^^   Asystole   AED^^^ – no shock advised/unknown   1.64 (1.33 – 2.02) 1.44 (1.16 – 1.78) Reference 0.84 (0.69 – 1.02)  < 0.001 0.001  0.08  1.05 (0.81 – 1.35) 1.13 (0.85 – 1.50) Reference 0.96 (0.72 – 1.27)   0.74 0.39  0.76   0.000 0.100  -0.036    1.000 0.921  1.032  Response time    < 6 minutes   6 – 9 minutes   10 – 15 minutes ≥ 16 minutes Unknown   Reference 0.98 (0.84 – 1.14) 0.98 (0.74 – 1.30) 0.44 (0.25 – 0.79) 1.19 (0.51 – 2.76)    0.82 0.98 0.005 0.69  Reference 1.08 (0.88 – 1.33) 1.07 (0.71 – 1.60) 1.98 (0.84 – 4.75) 1.99 (0.49 – 8.04)    0.44 0.75 0.13 0.34   0.028 0.020 0.638 0.709   0.986 0.981 0.500 0.500  Bystander-attempted resuscitation     No bystander resuscitation attempted    Bystander resuscitation attempted   Reference 1.32 (1.14 – 1.54)    < 0.001  Reference 0.98 (0.81 – 1.20)   0.86   0.016     0.996 EMS CPR/Rhythm analysis sequence     CPR preceded rhythm analysis      CPR did not precede rhythm analysis   1.24 (1.02 – 1.54) Reference   0.03    0.88 (0.77 – 1.26) Reference   0.34   -0.020   0.906 139    Prehospital epinephrine administered    Yes    No   47.1 (40.6 – 54.6) Reference  < 0.001  0.99 (0.81 – 1.20) Reference  0.90  0.000  1.000  * Standard deviation of the mean of the natural log transformed predicted values of receiving tracheal intubation for cases not transported to a hospital capable of delivering specialized post-arrest care = 0.77. Propensity score matched caliper distance set at 0.007. ^VF/VT (ventricular fibrillation/ventricular tachycardia)  ^^PEA (pulseless electrical activity)  ^^^AED (automatic external defibrillator)   140   Chapter 5: Discussion and Implications The purpose of this chapter is to synthesize the overall findings of this thesis, discuss its implications and highlight the unique contribution of this work, and its implications, to the existing body of knowledge.  5.1 Key findings This dissertation examined the effect of clinical interventions associated with the chain-of-survival in adult non-traumatic OHCA. The specific objectives included quantifying the effects of individual clinical interventions represented in the chain-of-survival, assessing potential modification of the effect of epinephrine administration and advance airway placement on survival by initial rhythm, and estimation of the causal effect of epinephrine and ETI using a propensity score matched analysis (in order to address confounding by indication).  In summary, the studies confirmed consistent positive associations between younger age, shorter response time, public location, bystander witnessed arrest, and ventricular fibrillation/tachycardia as the first rhythm and ROSC and survival. We also found a consistently negative association between epinephrine and ROSC and survival. Epinephrine administration modified the survival effect of initial rhythm. However, we did not find a consistently positive or negative association between advanced airway interventions and survival. We attempted to assess the effect of the availability of specialized post-arrest care and the delivery of post-arrest care (including control of the body temperature, and identification and treatment of acute coronary syndrome) which was included in the resuscitation guidelines in 2010 (90), but were unable to do so due to 141   survivor bias. We also explored the role of some novel aspects of the chain of survival such as EMS CPR performed prior to EMS rhythm analysis (EMS CPR represents the clinical operationalization of the three-phase model of cardiac arrest [79, 90] and requires a sequenced response to resuscitation practice). Our findings showed that the effect of “priming the pump”, that is ensuring a short period of EMS CPR prior to defibrillation, is negatively associated with survival when the ambulance response time is 6 minutes or less and the patient is not transported to a hospital able to provide specialized post-arrest care. In an effort to address the potential for confounding by indication, we conducted a matched propensity score analysis and confirmed a moderate beneficial effect of ETI on ROSC and survival and a moderately strong negative effect of epinephrine on ROSC and a very strong negative effect of epinephrine on survival to discharge from hospital.        5.2 Analytic approach Resuscitation research as a discipline has historically focused on hypothesizing and testing individual causal effects. Also, analytic heuristics, such as the Utstein reporting template [64], encourage a uniform method of capturing and reporting variables that has been widely adopted as the standard means of covariate inclusion in analytic studies. Despite its ubiquity, the Utstein template variables explain only a modest amount of the difference in survival among ROC sites.  The amount of difference in survival explained by the Utstein variables is reduced further when the subpopulation of bystander witnessed arrest is examined [66]. Nevertheless, isolated evaluation of individual clinical interventions has to be abandoned in favour of a focus on the chain of survival as a complete set of clinical interventions. Results of regression models vary 142   depending on the variable set included, and isolated modeling on chain-of-survival interventions risks producing results that are less meaningful and possibly uninterpretable. Our modeling of the set of the chain-of-survival interventions showed that the effects of known determinants of survival such as age, response time, first rhythm, and public location were congruent with findings from previously studies. This confirmed that our statistical models functioned in a manner that was consistent with what has been previously demonstrated to have an impact on survival.    5.3 Role of advanced airway management and tracheal intubation  Only one randomized controlled trial [30] has compared ETI with bag-mask ventilation in OHCA and that non-inferiority trial failed to demonstrate non-inferiority or inferiority with regard to the outcome of favourable neurological function. Four systematic reviews [26-29] have produced conflicting results, which may be a consequence of variations in the questions posed, study populations and other characteristic, the underlying physiologic mechanisms of oxygenation in OHCA (including the potential for hyperoxemia and dyscarbia resulting from overly vigorous ventilation efforts, both of which are harmful) or greater concentration on performing chest compressions and delaying ETI until a later stage in the resuscitation effort [91]. Advanced airway management, including ETI is indicated in cardiac arrest to protect the airway from aspiration, and provide adequate oxygenation and ventilation. However, what actually constitutes optimal oxygenation and ventilation during resuscitation is unknown and the requirements may change over the course of the resuscitation itself, for example, by mechanically inducing hypotension during the pre-ROSC phase [92], and/or inducing 143   reperfusion injury [91] in the post-ROSC phase. The paradoxical results for advanced airway management reported in Chapter 2 may be evidence of confounding by indication. Those most at risk OHCA cases may have received advanced airway management and this may have resulted in reduced relative odds of ROSC (AOR for ROSC at ED among cases receiving advanced airway management 0.46, 95% CI 0.32-0.65). However, analysis stratified by receipt of specialized post-arrest care showed that among those who did receive post-arrest care, the odds of survival contrasting those who received advanced airway management were not significantly different from the odds of survival among those who did not receive advanced airway management (AOR 0.72, 95% CI 0.45-1.16). On the other hand, among those who did not receive specialized post-arrest care, the relative odds of survival among those receiving advanced airway were substantially higher (AOR 1.77, 95% CI 1.26-2.49). Assuming that the sickest cases were selected for intubation and did not achieve ROSC in the prehospital environment, and the decision to carry out hospital transfer likely reflected the clinical judgment of the paramedical personnel who believed that the patient warranted additional specialized services if they were to survive and the provision of specialized post-arrest care may have reduced the mortality effect of hyperoxia [93]. Complicating the interpretation of the findings further is the lack of data associated with the timing of airway management and ROSC. It is therefore not possible to fully ascertain the cause-effect relationship, if any, between airway management and ROSC in the absence of detailed information of the indication for the placement of an advanced airway (e.g., following the guidelines, protecting the airway from regurgitation, or responding to end tidal CO2 information that can only be captured after ROSC).  144   Given our findings and the numerous clinical decisions made during and after resuscitation,  future studies will likely benefit from considering more closely the post-arrest care that is provided. If post-arrest care produces a survival benefit (and especially considering that EMS systems are now considering including ECMO as part of post arrest care [94, 95]), and if clinical decisions by paramedical personnel regarding transport options after ROSC affect survival, then controlling for this important determinant will be required in studies on OHCA.   5.4 Role of epinephrine  In this investigation, epinephrine had a consistent, negative effect on both ROSC and survival.  Although epinephrine has been included in resuscitation guidelines since the 1970’s [44], there has only been one RCT comparing epinephrine vs placebo [34] and this failed to demonstrate an improvement in survival. A number of recently published systematic reviews [36-38] report improved odds of ROSC with no effect on survival or neurologically intact survival, confirming a differential effect of epinephrine on ROSC compared with survival. An increasing body of research is beginning to shed new light on the survival benefit of epinephrine [38-42, 53, 80, 88, 96, 97], specifically that its effects are short term, and time dependent meaning that earlier and smaller doses result in better outcomes than do later and larger doses.       Epinephrine has several physiologically beneficial actions in cardiac arrest including its α-adrenergic stimulating effect resulting in increased heart rate, ventricular contractility, and peripheral vasoconstriction. However this leads to increased myocardial oxygen consumption [98].  Neset et al (96) described increased cardiac rhythm transitions following receipt of 145   epinephrine as compared with patients who did not receive epinephrine. Such rhythm transitions more often than not resulted in an increased frequency of transition into refractory VF, which is not a beneficial outcome. These findings are supported by electro-physiologic research showing that epinephrine has a varying effect on different cardiac rhythms [99]. Our results and those of much of the published literature are consistent with these findings which reveal a temporary effect of epinephrine and the mechanisms that underlie potential adverse effects.    Given the evolving conflicting evidence surrounding the role of epinephrine in OHCA resuscitation and the scanty experimental foundation for its use [99], a number of avenues for further inquiry need to be explored for definitely determining the effect of this commonly used drug. There is certainly a need to understand the effect that epinephrine has on the post ROSC heart. Similar paradoxical increases in ROSC followed by no benefit on survival, and compromised neurological status [100] have been observed with other interventions used during OHCA (for example, mechanical compared to manual chest compression CPR) and this has led to a cautionary approach limiting their use. Similar caution should be applied to the use of epinephrine.    5.5 A note on confounding by indication in non-experimental research Confounding by the indication can intractably bias non-experimental studies on the efficacy of interventions [69]. This is because the indication (which is closely related to but not synonymous with the severity of the underlying clinical condition) cannot be measured with complete accuracy and hence cannot be controlled through epidemiologic or statistical means. It is for this 146   reason that efficacy of interventions is best assessed through rigorously randomized controlled trials.  In our studies on the efficacy of interventions used in OHCA, especially the efficacy of epinephrine, we used various methods to control confounding by indication including stratification, regression, and propensity score adjustment. However, as mentioned, these methods cannot guarantee that residual confounding by indication was eliminated. Thus matched propensity score analysis cannot fully adjust for an indication which cannot be accurately measured [69]. Nevertheless, our finding of the differential effects of epinephrine on ROSC vs survival, the strong negative effects of epinephrine on survival, and the modification of the effect of epinephrine on survival by initial rhythm cast doubt on any argument that these effects are entirely accounted for by confounding by indication. Previous studies which have shown improved survival following early vs late administration of epinephrine [42] also suggest that epinephrine is not reserved for the most severe cases and hence unlikely to be intractably linked with confounding by indication.         5.6 Strengths and limitations There are several strengths to this research. The analyses contained in this dissertation were performed on data collected from a multi-centre academic prehospital resuscitation research network that used standard data dictionaries and data collection processes to ensure uniformity of the data collected across sites. Standard data definitions, use of electronic data collection where possible for first cardiac rhythm, and time stamps reduced the potential for misclassification of 147   information. Additionally, the clinical population and communities represented in these data are substantial and cover a vast network in North America. Both the quality of the data and large footprint covered by the sites speaks to be the internal validity and generalizability of the results.  Additionally, we relied to two datasets collected by the ROC affording us the opportunity to study key questions in an initial cohort, test the findings from this study on a second cohort, and then confirm the findings on the original dataset using a different and rigorous methodology.  This research has some limitations. The observational nature of the studies mean that bias and confounding are difficult to eliminate completely. However, we worked to carefully identify sources of bias and to address confounding using appropriate analyses. Some of the complex   results that emerged from the studies described in Chapters 2 and 3 suggest that confounding by indication likely remains at least in residual form [69]. We used several different techniques to reduce the effect of such bias including stratification, regression and propensity score adjustment. The effects of known determinants of survival including age, response time, and bystander witness status were consistent and expected both in direction and magnitude and this lends credibility to the methods used in our studies.    5.7 Implications This dissertation research has several implications for practice and further research. First, in light of the evolving and conflicting evidence regarding the effects of epinephrine and advanced airway management on survival following OHCA, resuscitation guidelines should reevaluate the central role of these intervention in resuscitation. A reduction in the emphasis on airway 148   management occurred with the shift from A-B-C (airway, breathing, circulation) in the resuscitation algorithm to C-A-B (circulation, airway, breathing), and this emphasized the importance of chest compressions rather than “securing the airway”. There is a similar need to reevaluate the central role of epinephrine in OHCA resuscitation. Randomized trial evidence on the efficacy of epinephrine for improving survival following OHCA would provide definitive evidence and clarity with regard to a critical element in the chain of survival.   There is also a need to further investigate the complex relationship between ROSC and survival, with greater attention being paid to understanding the common determinants of ROSC, survival, and favourable neurological status. Studies including ours have reported improved ROSC but reduced survival (or neurologically intact survival), indicating that epinephrine may be helping to establish one of the conditions required for survival, while simultaneously creating other conditions that decrease survival. An important question arising from this is: Is it possible to reduce the cardiac instability that results from the administration of epinephrine, and if so how can this be achieved?  Additionally, as noted by Paradis and Koscove in the early 1990’s [99], cardiac rhythms represent a heterogeneous collection of pathologic phenomena. Unfortunately, it remains the case that little is known or understood about the varied pathology at issue in any given instance of OHCA, as cardiac arrest manifests with a uniform clinical presentation despite a range of etiologies, including coronary vascular occlusion, asphyxiation, poisoning, and electrophysiological disturbances. Without a better understanding of the etiology of the various 149   types of arrest, the differential effects of varied potentially beneficial treatments may not be applied to the specific populations who would benefit most from their application.  5.7 Postscript Perkins et al published the results of their placebo controlled randomized controlled trial of epinephrine [75, 101] in the 23 August 2018 issue of the New England Journal of Medicine, two weeks prior to the scheduled oral defence of this dissertation.  The following is a brief discussion of their study.  The study enrolled 8,014 patients, and observed the following: the rate of survival at 30 days post-discharge for the epinephrine group was 3.2% and for the placebo group it was 2.4%. However there was no evidence of a significant difference in the proportion of patients with a favourable neurological outcome at hospital discharge (2.2% vs 1.9%), and there was an increase in severe neurological impairment at discharge in the epinephrine group (31.0% vs 17.8%). The number needed to treat (NNT) associated with receipt of epinephrine (NNT = 112), was substantially higher than that for early recognition of arrest (NNT = 11), bystander CPR (NNT = 15), and early defibrillation (NNT = 5). The authors argued that ongoing efforts to increase public participation in resuscitation is likely to yield substantially greater benefit with less harm than epinephrine administration.  While these results do demonstrate a statistically significant but small survival benefit for epinephrine, the investigators argued that the increase in neurological impairment among those 150   who received epinephrine, as well as the substantial burden placed on the providers and population to attain a single survivor may be clinically inadvisable. Additionally, there are limitations to this study that warrant the conduct of additional randomized controlled trials on the issue of epinephrine efficacy and safety.  The investigators acknowledged that patients prefer long-term neurologically intact survival rather than just survival, and so a study that is powered to investigate differences in neurologically intact survival is warranted. The Perkins et al study also included a mixed population of traumatic and non-traumatic cardiac arrest, despite the known lack of effect of epinephrine in traumatic arrest. Further studies on epinephrine should be restricted to non-traumatic out of hospital cardiac arrest.  Finally, the investigators noted that they planned and powered the study with the expectation of 6% survival, while they observed only a 2% survival which was unexpected. In summary, further randomized trials of epinephrine are required if use of this drug is to be evidence-based.      151   References 1. Heart and Stroke Foundation of Canada, e-pulse: National Electronic Newsletter, H.a.S.F.o. Canada, Editor. 2012. 2. Mozaffarian, D., et al., Heart disease and stroke statistics--2015 update: a report from the American Heart Association. Circulation, 2015. 131(4): p. e29-322. 3. Go, A.S., et al., Heart disease and stroke statistics--2014 update: a report from the American Heart Association. Circulation, 2014. 129(3): p. e28-e292. 4. Jacobs, I., et al., The chain of survival. Annals of Emergency Medicine, 2001. 37(4): p. S5-S16. 5. Newman, M., Chain of Survival: Concept Takes Hold. Jems, 1989. 14(8): p. 11 - 13. 6. Cummins, R.O., et al., Improving survival from sudden cardiac arrest: the "chain of survival" concept. A statement for health professionals from the Advanced Cardiac Life Support Subcommittee and the Emergency Cardiac Care Committee, American Heart Association. Circulation, 1991. 83(5): p. 1832-47. 7. Guidelines for cardiopulmonary resuscitation and emergency cardiac care. Emergency Cardiac Care Committee and Subcommittees, American Heart Association. Part IX. Ensuring effectiveness of communitywide emergency cardiac care. JAMA, 1992. 268(16): p. 2289-95. 8. Nolan, J., J. Soar, and H. Eikeland, The chain of survival. Resuscitation, 2006. 71(3): p. 270-271. 9. Bass, R.R.B., J.H.; Delbridge T.R.; Gunderson, M.R., Medical Oversight of EMS. Emergency Medical Services - Clincial Practice and Systems Practice, ed. O.C.R. Cone DC, Fowler RL. Vol. 2. 2009, Dubuque, IA: Kendall Hunt Publishing Co. 10. ECC Committee, S. and A. Task Forces of the American Heart, 2005 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2005. 112(24 Suppl): p. IV1-203. 11. Nolan, J. and C. European Resuscitation, European Resuscitation Council guidelines for resuscitation 2005. Section 1. Introduction. Resuscitation, 2005. 67 Suppl 1: p. S3-6. 12. Field, J.M., et al., Part 1: executive summary: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2010. 122(18 Suppl 3): p. S640-56. 13. Nolan, J.P., et al., European Resuscitation Council Guidelines for Resuscitation 2010 Section 1. Executive summary. Resuscitation, 2010. 81(10): p. 1219-76. 14. Monsieurs, K.G., et al., European Resuscitation Council Guidelines for Resuscitation 2015: Section 1. Executive summary. Resuscitation, 2015. 95: p. 1-80. 15. Neumar, R.W., et al., Part 1: Executive Summary: 2015 American Heart Association Guidelines Update for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2015. 132(18 Suppl 2): p. S315-67. 16. Perkins, G.D., et al., European Resuscitation Council Guidelines for Resuscitation 2015: Section 2. Adult basic life support and automated external defibrillation. Resuscitation, 2015. 95: p. 81-99. 152   17. Kleinman, M.E., et al., Part 5: Adult Basic Life Support and Cardiopulmonary Resuscitation Quality: 2015 American Heart Association Guidelines Update for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2015. 132(18 Suppl 2): p. S414-35. 18. Soar, J., et al., European Resuscitation Council Guidelines for Resuscitation 2015: Section 3. Adult advanced life support. Resuscitation, 2015. 95: p. 100-47. 19. Link, M.S., et al., Part 7: Adult Advanced Cardiovascular Life Support: 2015 American Heart Association Guidelines Update for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2015. 132(18 Suppl 2): p. S444-64. 20. Salmen, M., G.A. Ewy, and C. Sasson, Use of cardiocerebral resuscitation or AHA/ERC 2005 Guidelines is associated with improved survival from out-of-hospital cardiac arrest: a systematic review and meta-analysis. BMJ Open, 2012. 2(5). 21. Perkins, G.D., P. Cottrell, and S. Gates, Is adrenaline safe and effective as a treatment for out of hospital cardiac arrest? Bmj, 2014. 348: p. g2435. 22. Benger, J., Commentary to intubation strategy in cardiac arrest victims. Trends in Anaesthesia and Critical Care, 2015. 5(5): p. 134-5. 23. Wang, H.E., et al., Intubation on trial. JEMS, 2015. 40(8): p. 40-2, 44. 24. Neumar, R.W., et al., Part 8: adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2010. 122(18 Suppl 3): p. S729-67. 25. Sayre, M.R., et al., Part 2: evidence evaluation and management of potential or perceived conflicts of interest: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2010. 122(18 Suppl 3): p. S657-64. 26. Fouche, P.F., et al., Airways in out-of-hospital cardiac arrest: systematic review and meta-analysis. Prehosp Emerg Care, 2014. 18(2): p. 244-56. 27. Tiah, L., et al., Does pre-hospital endotracheal intubation improve survival in adults with non-traumatic out-of-hospital cardiac arrest? A systematic review. West J Emerg Med, 2014. 15(7): p. 749-57. 28. Benoit, J.L., et al., Endotracheal intubation versus supraglottic airway placement in out-of-hospital cardiac arrest: A meta-analysis. Resuscitation, 2015. 93: p. 20-6. 29. Jeong, S., K.O. Ahn, and S.D. Shin, The role of prehospital advanced airway management on outcomes for out-of-hospital cardiac arrest patients: a meta-analysis. Am J Emerg Med, 2016. 34(11): p. 2101-2106. 30. Jabre, P., et al., Effect of Bag-Mask Ventilation vs Endotracheal Intubation During Cardiopulmonary Resuscitation on Neurological Outcome After Out-of-Hospital Cardiorespiratory Arrest: A Randomized Clinical Trial. JAMA, 2018. 319(8): p. 779-787. 31. Tintinalli, J.E., et al., Tintinalli's emergency medicine : a comprehensive study guide. 2011, McGraw-Hill Education,: New York. p. 1 online resource. 32. Goto, Y., T. Maeda, and Y. Goto, Effects of prehospital epinephrine during out-of-hospital cardiac arrest with initial non-shockable rhythm: an observational cohort study. Crit Care, 2013. 17(5): p. R188. 153   33. Hagihara, A., et al., Prehospital epinephrine use and survival among patients with out-of-hospital cardiac arrest. JAMA, 2012. 307(11): p. 1161-8. 34. Jacobs, I.G., et al., Effect of adrenaline on survival in out-of-hospital cardiac arrest: A randomised double-blind placebo-controlled trial. Resuscitation, 2011. 82(9): p. 1138-43. 35. Nakahara, S., et al., Evaluation of pre-hospital administration of adrenaline (epinephrine) by emergency medical services for patients with out of hospital cardiac arrest in Japan: controlled propensity matched retrospective cohort study. Bmj, 2013. 347: p. f6829. 36. Atiksawedparit, P., et al., Effects of prehospital adrenaline administration on out-of-hospital cardiac arrest outcomes: a systematic review and meta-analysis. Crit Care, 2014. 18(4): p. 463. 37. Lin, S., et al., Adrenaline for out-of-hospital cardiac arrest resuscitation: a systematic review and meta-analysis of randomized controlled trials. Resuscitation, 2014. 85(6): p. 732-40. 38. Loomba, R.S., et al., Increased return of spontaneous circulation at the expense of neurologic outcomes: Is prehospital epinephrine for out-of-hospital cardiac arrest really worth it? J Crit Care, 2015. 30(6): p. 1376-81. 39. Fukuda, T., et al., Effect of prehospital epinephrine on out-of-hospital cardiac arrest: a report from the national out-of-hospital cardiac arrest data registry in Japan, 2011-2012. Eur J Clin Pharmacol, 2016. 72(10): p. 1255-1264. 40. Sagisaka, R., et al., Effects of repeated epinephrine administration and administer timing on witnessed out-of-hospital cardiac arrest patients. Am J Emerg Med, 2017. 35(10): p. 1462-1468. 41. Tanaka, H., et al., Favorable neurological outcomes by early epinephrine administration within 19 minutes after EMS call for out-of-hospital cardiac arrest patients. Am J Emerg Med, 2016. 34(12): p. 2284-2290. 42. Ueta, H., et al., Quick epinephrine administration induces favorable neurological outcomes in out-of-hospital cardiac arrest patients. Am J Emerg Med, 2017. 35(5): p. 676-680. 43. Hagihara, A., et al., Effects of advanced life support on patients who suffered cardiac arrest outside of hospital and were defibrillated. Am J Emerg Med, 2018. 36(1): p. 73-78. 44. Standards for cardiopulmonary resuscitation (CPR) and emergency cardiac care (ECC). 3. Advanced life support. JAMA, 1974. 227(7): p. Suppl:852-60. 45. Wang, H.E., et al., Design and implementation of the Resuscitation Outcomes Consortium Pragmatic Airway Resuscitation Trial (PART). Resuscitation, 2016. 101: p. 57-64. 46. Taylor, J., et al., Design and implementation of the AIRWAYS-2 trial: A multi-centre cluster randomised controlled trial of the clinical and cost effectiveness of the i-gel supraglottic airway device versus tracheal intubation in the initial airway management of out of hospital cardiac arrest. Resuscitation, 2016. 109: p. 25-32. 154   47. Kronick, S.L., et al., Part 4: Systems of Care and Continuous Quality Improvement: 2015 American Heart Association Guidelines Update for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation, 2015. 132(18 Suppl 2): p. S397-413. 48. Eisenberg, M.S., L. Bergner, and A.P. Hallstrom, Sudden cardiac death in the community. 1984, New York: Praeger. xxiii, 163 p. 49. Kay, C.F., Editorial: Cardiopulmonary resuscitation: a chain of many links. Ann Intern Med, 1974. 80(3): p. 411-3. 50. Peberdy, M.A. and J.P. Ornato, Post-resuscitation care: is it the missing link in the Chain of Survival? Resuscitation, 2005. 64(2): p. 135-137. 51. Hanada, H. and K. Okumura, From 4-Links to 5-Links of "Chain of Survival" - Post-Resuscitation Care is Critical for Good Neurological Recovery. Circulation Journal, 2009. 73(10): p. 1797-1798. 52. Sasson, C., et al., Predictors of survival from out-of-hospital cardiac arrest: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes, 2010. 3(1): p. 63-81. 53. Patanwala, A.E., et al., Effect of epinephrine on survival after cardiac arrest: a systematic review and meta-analysis. Minerva Anestesiol, 2014. 80(7): p. 831-43. 54. Huang, Y., et al., Antiarrhythmia drugs for cardiac arrest: a systemic review and meta-analysis. Crit Care, 2013. 17(4): p. R173. 55. Kudenchuk, P.J., et al., Amiodarone, Lidocaine, or Placebo in Out-of-Hospital Cardiac Arrest. N Engl J Med, 2016. 374(18): p. 1711-22. 56. Sanghavi, P., et al., Outcomes after out-of-hospital cardiac arrest treated by basic vs advanced life support. JAMA Intern Med, 2015. 175(2): p. 196-204. 57. Huang, Y., et al., Cardiopulmonary resuscitation (CPR) plus delayed defibrillation versus immediate defibrillation for out-of-hospital cardiac arrest. Cochrane Database Syst Rev, 2014(9): p. CD009803. 58. Biondi-Zoccai, G., et al., An updated systematic review and meta-analysis on impedance threshold devices in patients undergoing cardiopulmonary resuscitation. Heart Lung Vessel, 2014. 6(2): p. 105-13. 59. Nichol, G., et al., Trial of Continuous or Interrupted Chest Compressions during CPR. N Engl J Med, 2015. 373(23): p. 2203-14. 60. Spaite, D.W., et al., Statewide Regionalization of Postarrest Care for Out-of-Hospital Cardiac Arrest: Association With Survival and Neurologic Outcome. Ann Emerg Med, 2014. 64(5): p. 496-506.e1. 61. Davis, D.P., et al., A descriptive analysis of Emergency Medical Service Systems participating in the Resuscitation Outcomes Consortium (ROC) network. Prehosp Emerg Care, 2007. 11(4): p. 369-82. 62. Morrison, L.J., et al., Rationale, development and implementation of the Resuscitation Outcomes Consortium Epistry-Cardiac Arrest. Resuscitation, 2008. 78(2): p. 161-9. 63. Shah, M.N., The formation of the emergency medical services system. Am J Public Health, 2006. 96(3): p. 414-23. 64. Jacobs, I., et al., Cardiac arrest and cardiopulmonary resuscitation outcome reports: update and simplification of the Utstein templates for resuscitation registries: a statement for healthcare professionals from a task force of the International Liaison Committee on 155   Resuscitation (American Heart Association, European Resuscitation Council, Australian Resuscitation Council, New Zealand Resuscitation Council, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Councils of Southern Africa). Circulation, 2004. 110(21): p. 3385-97. 65. Vittinghoff, E., Regression methods in biostatistics : linear, logistic, survival, and repeated measures models. Statistics for biology and health. 2005, New York: Springer. xv, 340 p. 66. Rea, T.D., et al., Predicting survival after out-of-hospital cardiac arrest: role of the Utstein data elements. Ann Emerg Med, 2010. 55(3): p. 249-57. 67. Andrusiek, D.L., et al., Effects of Chain of Survival Interventions and Other Factors on Survival Following Non-traumatic Out-of-Hospital Cardiac Arrest, in Unpublished. 2018. 68. Andrusiek, D., et al., Effect of an intact "chain of survival" sequence on survival to discharge from out-of-hospital cardiac arrest. Canadian Journal of Emergency Medicine, 2017. 19(S1): p. s30. 69. Joseph, K.S., A. Mehrabadi, and S. Lisonkova, Confounding by indication and related concepts. Current Epidemiology Reports, 2014. 1(1): p. 1-8. 70. Ewy, G.A., et al., The time dependent association of adrenaline administration and survival from out-of-hospital cardiac arrest. Resuscitation, 2015. 96: p. 180-5. 71. Machida, M., et al., Effect of intravenous adrenaline before arrival at the hospital in out-of-hospital cardiac arrest. J Cardiol, 2012. 60(6): p. 503-7. 72. Berg, R.A. and B.J. Bobrow, Cardiac resuscitation: Is an advanced airway harmful during out-of-hospital CPR? Nat Rev Cardiol, 2013. 10(4): p. 188-9. 73. Hasegawa, K., et al., Association of prehospital advanced airway management with neurologic outcome and survival in patients with out-of-hospital cardiac arrest. JAMA, 2013. 309(3): p. 257-66. 74. Wang, H.E., et al., Out-of-hospital endotracheal intubation experience and patient outcomes. Ann Emerg Med, 2010. 55(6): p. 527-537.e6. 75. Perkins, G.D., et al., Pre-hospital Assessment of the Role of Adrenaline: Measuring the Effectiveness of Drug administration In Cardiac arrest (PARAMEDIC-2): Trial protocol. Resuscitation, 2016. 108: p. 75-81. 76. Aufderheide, T.P., et al., A trial of an impedance threshold device in out-of-hospital cardiac arrest. N Engl J Med, 2011. 365(9): p. 798-806. 77. Stiell, I.G., et al., Early versus later rhythm analysis in patients with out-of-hospital cardiac arrest. N Engl J Med, 2011. 365(9): p. 787-97. 78. Jaccard, J. and SAGE Research Methods Core, Interaction Effects in Logistic Regression. Vol. 135, in Quantitative Applications in the Social Sciences Ser. 2001, SAGE Publications, Incorporated: Thousand Oaks. p. 1 online resource. 79. Weisfeldt, M.L. and L.B. Becker, Resuscitation after cardiac arrest: a 3-phase time-sensitive model. Jama, 2002. 288(23): p. 3035-8. 80. Dumas, F., et al., Is epinephrine during cardiac arrest associated with worse outcomes in resuscitated patients? J Am Coll Cardiol, 2014. 64(22): p. 2360-7. 156   81. Voss, S., et al., How do paramedics manage the airway during out of hospital cardiac arrest? Resuscitation, 2014. 85(12): p. 1662-6. 82. Pearl, J., Causality : models, reasoning, and inference. 2nd ed. 2009, Cambridge ; New York: Cambridge University Press. xix, 464 p. 83. Newgard, C.D., et al., The Resuscitation Outcomes Consortium Epistry-Trauma: design, development, and implementation of a North American epidemiologic prehospital trauma registry. Resuscitation, 2008. 78(2): p. 170-8. 84. Austin, P.C., A Tutorial and Case Study in Propensity Score Analysis: An Application to Estimating the Effect of In-Hospital Smoking Cessation Counseling on Mortality. Multivariate Behav Res, 2011. 46(1): p. 119-151. 85. Austin, P.C., An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies. Multivariate Behav Res, 2011. 46(3): p. 399-424. 86. Fox, K., Discovering the aetiology of heart failure, one link in the chain to improved survival. International Journal of Cardiology, 2000. 74(1): p. 75-76. 87. McMullan, J., et al., Airway management and out-of-hospital cardiac arrest outcome in the CARES registry. Resuscitation, 2014. 85(5): p. 617-22. 88. Andersen, L.W., et al., Early administration of epinephrine (adrenaline) in patients with cardiac arrest with initial shockable rhythm in hospital: propensity score matched analysis. BMJ, 2016. 353: p. i1577. 89. Newgard, C.D., et al., Advanced statistics: the propensity score--a method for estimating treatment effect in observational research. Acad Emerg Med, 2004. 11(9): p. 953-61. 90. Weisfeldt, M.L., A three phase temporal model for cardiopulmonary resuscitation following cardiac arrest. Trans Am Clin Climatol Assoc, 2004. 115: p. 115-22. 91. Henlin, T., et al., Oxygenation, ventilation, and airway management in out-of-hospital cardiac arrest: a review. Biomed Res Int, 2014. 2014: p. 376871. 92. Aufderheide, T.P., et al., Hyperventilation-induced hypotension during cardiopulmonary resuscitation. Circulation, 2004. 109(16): p. 1960-5. 93. Llitjos, J.F., et al., Hyperoxia toxicity after cardiac arrest: What is the evidence? Ann Intensive Care, 2016. 6(1): p. 23. 94. Grunau, B., et al., A comprehensive regional clinical and educational ECPR protocol decreases time to ECMO in patients with refractory out-of-hospital cardiac arrest. CJEM, 2017. 19(6): p. 424-433. 95. Sakamoto, T., et al., Extracorporeal cardiopulmonary resuscitation versus conventional cardiopulmonary resuscitation in adults with out-of-hospital cardiac arrest: a prospective observational study. Resuscitation, 2014. 85(6): p. 762-8. 96. Zhang, Q., et al., Efficacy of vasopressin-epinephrine compared to epinephrine alone for out of hospital cardiac arrest patients: A systematic review and meta-analysis. Am J Emerg Med, 2017. 35(10): p. 1555-1560. 97. Fisk, C.A., et al., Lower-dose epinephrine administration and out-of-hospital cardiac arrest outcomes. Resuscitation, 2018. 124: p. 43-48. 98. Neset, A., et al., Effects of adrenaline on rhythm transitions in out-of-hospital cardiac arrest. Acta Anaesthesiol Scand, 2013. 57(10): p. 1260-7. 157   99. Paradis, N.A. and E.M. Koscove, Epinephrine in cardiac arrest: a critical review. Ann Emerg Med, 1990. 19(11): p. 1288-301. 100. Li, H., et al., Mechanical versus manual chest compressions for cardiac arrest: a systematic review and meta-analysis. Scand J Trauma Resusc Emerg Med, 2016. 24: p. 10. 101. Perkins, G.D., et al., A Randomized Trial of Epinephrine in Out-of-Hospital Cardiac Arrest. N Engl J Med, 2018. 379(8): p. 711-721.    158   Appendix Appendix Table 1.2 Systematic reviews and meta-analyses of airway interventions research in OHCA  Study Design, Cohort, Sample Size,  Comparator Adjustment Outcomes Notes Fouche, 2014 [26] Systematic review and meta-analysis of advanced airway management vs. basic airway management.  17 studies included with a pooled n=649,359. Mixed adult and pediatric population. Included traumatic arrest.  Advanced airway management included: ETI, laryngeal mask airways, supraglottic airways, double-lumen airways, and trans-tracheal or trans-cricothyroid airways compared with bag-mask ventilation, mouth-to-mouth ventilation with or without an oropharyngeal or nasopharyngeal airway. Bias adjusted pooled analysis. Any advanced airway compared to basic airway: ROSC (OR 0.84, 0.62–1.13); survival (OR 0.49, 0.37–0.65). ETI compared to basic airway: ROSC (OR 0.56, 0.40–0.78); survival (OR 0.35, 0.28–0.44). Significant heterogeneity in pooled results.  Note that many of the devices included in the advanced airway category are no longer in use or advocated, with the exception of supraglottic airways and ETI. 159   Tiah, 2014 [27] Systematic review of ETI vs. SGA. Five studies with a pooled n=303,348.  Included only studies comparing ETI to SGA. Adult, non-traumatic cases. ETI vs. SGA.   Included studies graded as good to excellent, based on the Newcastle-Ottawa Scale. The investigators conclude there is no trend towards superiority of ETI over SGA based on the included studies. No meta-analysis performed. Benoit, 2015 [28] Systematic review and meta-analysis of observational studies of tracheal intubation vs. supraglottic airway.  Ten studies with a pooled n=75,649. Sensitivity analysis performed by removing low GRADE score studies. Excluded pediatric, and traumatic arrest, as well as any instances where nurses or physicians intubated, and rapid sequence intubation (RSI), video assist, and any older airway devices.  ETI compared to SGA.  Age, sex, first rhythm, witness status, bystander CPR, bystander automatic external defibrillator (AED). ROSC: OR 1.28, 1.05–1.55: Hospital Admission: OR 1.34, 1.03–1.75; hospital discharge: OR 1.15, 0.97–1.37.  No RCTs available for this analysis. Jeong, 2016 [29] Systematic review and meta-analysis with Advanced airway Random effects modeling used.  Survival: Advanced airway management vs. basic No experimental studies included. 160   advanced airway management compared with basic airway management in non-traumatic cases. Ten studies included, pooled n=84,905. management (ETI and SGA) vs. basic airway management (bag-mask ventilation and oral airway). airway management: (pooled OR 0.51, 0.29–0.90); ETI vs. basic airway management (pooled OR 0.44, 0.16–1.23).   161   Appendix Table 1.3 Systematic reviews and meta-analyses of epinephrine vs. no epinephrine Study Design, Cohort, Sample Size,  Comparator Adjustment Outcomes Notes Atiksawedparit, 2014 [36] Meta-analysis of observational studies, 15 studies, pooled cohort for prehospital ROSC n=421,459; pooled cohort for overall ROSC n=2,381; pooled cohort for hospital admission n=8,470; pooled cohort for discharge alive n=4,743.  Standard dose epinephrine vs. no epinephrine. Random effects models applied when heterogeneity was detected via Cochrane’s Q or I2 Prehospital ROSC RR=2.89 (2.36-3.54); Overall ROSC RR=0.93 (0.5-1.75); Hospital Admission RR=1.05 (0.8–1.38); Discharged Alive RR=0.69 (0.48-1.00) Studies included had a mix of initial rhythm and traumatic and non-traumatic arrest; one in-hospital arrest study (n=492) contained only 34.4% presumed cardiac cause.  Lin, 2014 [37] Systematic review and meta-analysis of RCT and quasi-RCT. Pooled cohort=12,246.  RCT Epinephrine vs. placebo n=534.  Standard dose epinephrine vs. placebo; high dose epinephrine, or vasopressin. Report results of Jacobs, 2011[34]. ROSC in Epinephrine vs. placebo (RR 2.8, 1.78–4.41); survival to admission to hospital (RR 1.95, 1.34–2.84). Jacobs, 2011 [34] was stopped early due to resistance from paramedic staff. Loomba, 2015[38] Systematic review and meta-analysis of observational and experimental studies. Prehospital administration of epinephrine vs. no Random effects model used, with heterogeneity assessed chi-square and I2 ROSC (pooled OR 2.84, 2.28–3.54); 1-month survival (pooled OR 1.03, 0.79–1.34); survival to discharge (pooled OR 0.82, Results from multiple studies make interpretation of significant ROSC vs. non-162   Pooled n=655,853, from 14 studies. administration of epinephrine.  0.46–1.48); neurologic status at discharge (pooled OR 0.51, 0.31–0.84). significant survival outcomes difficult to interpret.    163   Appendix Table 1.4 Studies published subsequent to the 2015 systematic reviews and meta-analyses  Fukuda, 2016 [39] Japanese adult population, propensity score matched analysis. n=237,068; propensity score matched analysis n=66,800  Standard dose epinephrine vs. no epinephrine. Age, sex, witnessed arrest, bystander resuscitation, first rhythm, arrest etiology, response time, time to epinephrine administration, dose of epinephrine, advanced airway. Shockable rhythm:  ROSC (OR 0.91, 95% CI 0.82–1.00); 1-month survival (OR 0.53, 95% CI 0.47–0.59).  Neurological outcome at 1 month (OR 0.38, 95% CI 0.33–0.43). Non–shockable rhythm: ROSC (OR 4.08, 95% CI 3.85–4.32); 1–month survival (OR 1.00, 95% CI 0.92–1.09), neurological outcome at 1 month (OR 0.47, 95% CI 0.39–0.56). Includes EMS witnessed arrests.  Excludes all cases of excessive (< 60 min) response time, or lengthy resuscitations (< 120 min to hospital arrival). Tanaka, 2016 [41] Japanese population based, retrospective cohort study.  n=119,639.  Cases were older teens and adults, with presumed cardiac cause of arrest.  Cohort divided into 1/4th based on response time: early (5–18 min), intermediate (19–23 min), late (24–29 min), very late (30–62 min). Late epinephrine response time Year, age, sex, bystander characteristics, initial rhythm, bystander resuscitation, airway management, number of defibrillations. ROSC: no epinephrine (OR 0.62, 95% CI 0.57–0.68); early vs. late (OR 1.66, 95% CI 1.49–1.85); intermediate (OR 1.25, 95% CI 1.12–1.40), very late (OR 0.76, 95% CI 0.67–0.87). Neurological status at 1 month: no epinephrine (OR 3.33, 95% CI 2.60–4.26); early vs. late (OR 2.49, 95% CI 1.90–3.27); Intermediate (OR 1.53, 95% CI 1.14–Administration of epinephrine dependent on local medical directives that are not described.  Correlation between response time and epinephrine administration time. No comment on why the no 164   was the reference. 2.05); very late (OR 0.71, 95% CI 0.47–1.08). epinephrine group has the highest odds of survival. Sagisaka, 2017 [40] Japanese population based, retrospective cohort study.  n=11,876. Cases were bystander witnessed (not EMS), cardiac cause, who had epinephrine administered. Cohort divided into thirds based on time elapsed to first administration of epinephrine (late administration as reference), and number of doses epinephrine administered. Age, sex, bystander CPR, public defibrillation, initial rhythm, type of advanced airway, response time, time elapsed to first epinephrine administration. Tested interaction between timing of administration and number of doses. Survival: Early administration (OR 4.18, 3.30–5.38), intermediate (OR 1.84, 1.41–2.43).   Categories assigned to timing of administration of epinephrine were structurally related to response.  Additionally, repeated doses of epinephrine were administered only when the initial epinephrine was ineffective.  Excluded cases that were not treated with epinephrine.   Ueta, 2017 [42] Japanese population based retrospective cohort study. n=13,326.  Cases were bystander witnessed, cardiac cause of arrest who received epinephrine.   Population divided into groups based on response time (< 8 min vs. > 8 – 16 min), and time of administration of epinephrine (within 10 min Age, sex, bystander resuscitation, initial rhythm, number of defibrillations, type of airway,  ROSC: early administration (compared with delayed administration) within the early response group (OR 2.00, 95% CI 1.79–2.25), in the delayed response group (OR 2.00, 95% CI 1.79–2.25).  Good neurological outcome at 1 month in early administration group within This stratified analysis mimics the effect of an interaction. The effect of epinephrine on ROSC is inconsistent with other evidence of that relationship. 165   of first patient contact vs. longer than 10 min after first contact). early response group (OR 2.12, 95% CI 1.54–2.93), delayed response time (OR 2.66, 95% CI 1.97-2.39). Hagihara, 2018[43] Japanese population based retrospective cohort study, with propensity score methods. n=37,873 Cases were adults, with cardiac cause, who were bystander witnessed, with bystander defibrillation.    Advanced life support vs. no advanced life support.  Epinephrine and advanced airway vs. no epinephrine and no advanced airway. Year, age, sex, bystander resuscitation, ROSC, epinephrine, advanced airway, interaction between epinephrine and airway, physician on ambulance, qualification of ambulance crew.  Advanced life support: 1-month survival (propensity score): (OR 0.88, 0.80–0.97). Interaction of advanced airway and epinephrine: Epinephrine used without advanced airway (1-month survival): (OR 1.54, 0.90–2.64); Epinephrine used with an advanced airway (OR 0.79, 0.42–1.48).   Very few details provided of the overlap in intervention vs. control groups in the propensity-matched analyses. Tables of standardized differences show substantial differences between the groups for some variables.    166   Appendix Figure 1. Truncated conditional probability patient flowchart — proportions of patients alive at each node when patient achieved ROSC or had ongoing resuscitation at transport from scene.     Study Cohort n=12821 972/12821Prehospital ROSC or ongoing ResuscitaitonYes 972/7612Bystander ResusciationYes 464/4913No 508/2699EMS CPR preceded EMS rhythm analysisYes 383/1910No 125/789Yes 398/4133No 66/780Yes 83/627No 42/162Yes 311/1598No 72/312Yes 40/633No 26/147Yes 322/3440No 76/693Advanced AirwayEpineiphrineYes 178/2967No 144/473Yes 42/526No 41/60Yes 4/65No 68/247Yes 168/1349No 143/249Yes 25/580No 15/73Yes 4/76No 22/71Yes 9/147No 67/546Yes 3/53No 39/109Text167   Appendix Figure 2. Truncated conditional probability patient flowchart — proportions of patients alive at each node when resuscitation was terminated in the field.   Study Cohort n=12821 972/12821No 0/5029Yes 0/3396No 0/1813Yes 0/1434No 0/379Yes 0/3078 No 0/318Prehospital ROSC or ongoing ResuscitaitonBystander ResusciationEMS CPR preceded EMS rhythm analysisAdvanced AirwayEpineiphrineYes 0/2324No 508/2699Yes 0/207No 0/111Yes 0/2216No 0/108Yes 0/149No 0/605Yes 0/22No 0/89Yes 0/200No 0/7Yes 0/1166No 0/268Yes 0/290No 0/89Yes 0/1112No 0/54Yes 0/69No 0/199Yes 0/20No 0/69Yes 0/273No 0/17

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0372327/manifest

Comment

Related Items