[Skip to Navigation]
Sign In
Figure 1.  Survival to Hospital Discharge, Favorable Functional Outcome at Hospital Discharge, and Prehospital Return of Spontaneous Circulation (ROSC) Stratified by Timing of Epinephrine Administration in Patients With Out-of-Hospital Cardiac Arrest and Initial Shockable Cardiac Rhythms
Survival to Hospital Discharge, Favorable Functional Outcome at Hospital Discharge, and Prehospital Return of Spontaneous Circulation (ROSC) Stratified by Timing of Epinephrine Administration in Patients With Out-of-Hospital Cardiac Arrest and Initial Shockable Cardiac Rhythms

Figure shows the risk ratio (RR) point estimates (squares) with the 95% CIs (upper and lower bounds indicated by the blue dashed lines) for administration of epinephrine after arrival of emergency medical services personnel at the scene associated with survival to hospital discharge (A), favorable functional status at discharge (B), and ROSC (C). Timing of epinephrine administration was treated as a continuous variable. A, The RR per minute decreased 5.5% (95% CI, 3.4%-7.5%; P < .001 for the interaction). B, The RR per minute decreased 6.4% (95% CI, 3.8%-8.9%; P < .001 for the interaction). C, The RR per minute increased 1.4% (95% CI, 0.2%-2.7%; P = .02 for the interaction). The solid line represents the outcome. Risk ratios greater than 1.00 (horizontal line) favored receiving epinephrine; those less than 1.00, not receiving epinephrine. The error bars indicate 95% CIs.

Figure 2.  Survival to Hospital Discharge, Favorable Functional Outcome at Hospital Discharge, and Prehospital Return of Spontaneous Circulation (ROSC) Stratified by the Timing of Epinephrine Administration in Patients With Out-of-Hospital Cardiac Arrest and Initial Nonshockable Cardiac Rhythms
Survival to Hospital Discharge, Favorable Functional Outcome at Hospital Discharge, and Prehospital Return of Spontaneous Circulation (ROSC) Stratified by the Timing of Epinephrine Administration in Patients With Out-of-Hospital Cardiac Arrest and Initial Nonshockable Cardiac Rhythms

Figure shows the risk ratio point estimates (squares) with the 95% CIs (upper and lower bounds indicated by the blue dashed lines) associated with survival to hospital discharge (A), favorable functional status at discharge (B), and ROSC (C). Timing of epinephrine administration was treated as a continuous variable. A, The RR per minute decreased 4.4% (95% CI, 0.8%-7.9%; P = .02 for the interaction). B, The RR per minute decreased 7.1% (95% CI, 1.7%-12.3%; P = .01 for the interaction). C, The RR per minute increased 1.5% (95% CI, 0.6%-2.4%; P = .001 for the interaction). The solid line represents the outcome. Risk ratios greater than 1.00 (horizontal line) favored receiving epinephrine; those less than 1.00, not receiving epinephrine. The error bars indicate 95% CIs.

Table 1.  Characteristics and Covariates of Adults With Out-of-Hospital Cardiac Arrest With and Without Epinephrine in Original Cohorta
Characteristics and Covariates of Adults With Out-of-Hospital Cardiac Arrest With and Without Epinephrine in Original Cohorta
Table 2.  Characteristics and Covariates of Adults With Out-of-Hospital Cardiac Arrest With Epinephrine and at Risk of Receiving Epinephrine in the Time-Dependent Propensity Score–Matched Cohorta
Characteristics and Covariates of Adults With Out-of-Hospital Cardiac Arrest With Epinephrine and at Risk of Receiving Epinephrine in the Time-Dependent Propensity Score–Matched Cohorta
Table 3.  Outcomes in Time-Dependent Propensity Score–Matched Cohort
Outcomes in Time-Dependent Propensity Score–Matched Cohort
Supplement.

eMethods.

eReferences.

eFigure 1. Patient Flow

eFigure 2. Association of Epinephrine Administration With Survival to Hospital Discharge (A), Favorable Functional Outcome at Hospital Discharge (B), and Prehospital ROSC (C) by the Timing of the Administration for Patients With Shockable OHCA (Matching Without Replacement)

eFigure 3. Association of Epinephrine Administration With Survival to Hospital Discharge (A), Favorable Functional Outcome at Hospital Discharge (B), and Prehospital ROSC (C) by the Timing of the Administration for Patients With Nonshockable OHCA (Matching Without Replacement)

eFigure 4. Association of Epinephrine Administration With Survival to Hospital Discharge (A), Favorable Functional Outcome at Hospital Discharge (B), and Prehospital ROSC (C) by the Timing of the Administration for Patients With Shockable OHCA Excluding Those Who Had ROSC or TOR Within 5 Minutes After ALS EMS Arrival on the Scene

eFigure 5. Association of Epinephrine Administration With Survival to Hospital Discharge (A), Favorable Functional Outcome at Hospital Discharge (B), and Prehospital ROSC (C) by the Timing of the Administration for Patients With Nonshockable OHCA Excluding Those Who Had ROSC or TOR Within 5 Minutes After ALS EMS Arrival on the Scene

eFigure 6. Association of Epinephrine Administration With Survival to Hospital Discharge (A), Favorable Functional Outcome at Hospital Discharge (B), and Prehospital ROSC (C) by the Timing of the Administration for Patients With Bystander Witnessed Shockable OHCA

eFigure 7. Association of Epinephrine Administration With Survival to Hospital Discharge (A), Favorable Functional Outcome at Hospital Discharge (B), and Prehospital ROSC (C) for Patients With Bystander Witnessed Nonshockable OHCA

eTable 1. Characteristics of Adult Patients With Out-of-Hospital Cardiac Arrest With Epinephrine and at Risk of Receiving Epinephrine in Time-Dependent Propensity Score Matched Cohort (Matching Without Replacement)

eTable 2. Outcomes in Time-Dependent Propensity Score Matched Cohort (Matching Without Replacement)

eTable 3. Characteristics of Adult Patients With Out-of-Hospital Cardiac Arrest With and Without Epinephrine in Original Cohort, Excluding Those Who Had ROSC or TOR Within 5 Minutes After ALS EMS Arrival

eTable 4. Characteristics of Adult Patients With Out-of-Hospital Cardiac Arrest With Epinephrine and at Risk of Receiving Epinephrine in Time-Dependent Propensity Score Matched Cohort, Excluding Those Who Had ROSC or TOR Within 5 Minutes After ALS EMS Arrival

eTable 5. Outcomes in Time-Dependent Propensity Score Matched Cohort, Excluding Those Who Had ROSC or TOR Within 5 Minutes After ALS EMS Arrival

eTable 6. Characteristics of Adult Patients With Bystander Witnessed Out-of-Hospital Cardiac Arrest With and Without Epinephrine

eTable 7. Characteristics of Adult Patients With Bystander Witnessed Out-of-Hospital Cardiac Arrest With Epinephrine and at Risk of Receiving Epinephrine in Time-Dependent Propensity Score Matched Cohort

eTable 8. Outcomes in Time-Dependent Propensity Score Matched Cohort of Bystander Witnessed Out-of-Hospital Cardiac Arrest

1.
Virani  SS, Alonso  A, Benjamin  EJ,  et al; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics—2020 update: a report from the American Heart Association.   Circulation. 2020;141(9):e139-e596. doi:10.1161/CIR.0000000000000757 PubMedGoogle ScholarCrossref
2.
Kiguchi  T, Okubo  M, Nishiyama  C,  et al.  Out-of-hospital cardiac arrest across the world: first report from the International Liaison Committee on Resuscitation (ILCOR).   Resuscitation. 2020;152:39-49. doi:10.1016/j.resuscitation.2020.02.044 PubMedGoogle ScholarCrossref
3.
Perkins  GD, Ji  C, Deakin  CD,  et al; PARAMEDIC2 Collaborators.  A randomized trial of epinephrine in out-of-hospital cardiac arrest.   N Engl J Med. 2018;379(8):711-721. doi:10.1056/NEJMoa1806842 PubMedGoogle ScholarCrossref
4.
Soar  J, Maconochie  I, Wyckoff  MH,  et al.  2019 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations: summary from the Basic Life Support; Advanced Life Support; Pediatric Life Support; Neonatal Life Support; Education, Implementation, and Teams; and First Aid Task Forces.   Circulation. 2019;140(24):e826-e880. doi:10.1161/CIR.0000000000000734 PubMedGoogle ScholarCrossref
5.
Berg  KM, Soar  J, Andersen  LW,  et al; Adult Advanced Life Support Collaborators.  Adult advanced life support: 2020 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations.   Circulation. 2020;142(16)(suppl 1):S92-S139. doi:10.1161/CIR.0000000000000893PubMedGoogle Scholar
6.
Panchal  AR, Bartos  JA, Cabañas  JG,  et al; Adult Basic and Advanced Life Support Writing Group.  Part 3: adult basic and advanced life support: 2020 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.   Circulation. 2020;142(16)(suppl 2):S366-S468. doi:10.1161/CIR.0000000000000916PubMedGoogle Scholar
7.
Holmberg  MJ, Issa  MS, Moskowitz  A,  et al; International Liaison Committee on Resuscitation Advanced Life Support Task Force Collaborators.  Vasopressors during adult cardiac arrest: a systematic review and meta-analysis.   Resuscitation. 2019;139:106-121. doi:10.1016/j.resuscitation.2019.04.008 PubMedGoogle ScholarCrossref
8.
Andersen  LW, Grossestreuer  AV, Donnino  MW.  “Resuscitation time bias”—a unique challenge for observational cardiac arrest research.   Resuscitation. 2018;125:79-82. doi:10.1016/j.resuscitation.2018.02.006 PubMedGoogle ScholarCrossref
9.
Reynolds  JC, Frisch  A, Rittenberger  JC, Callaway  CW.  Duration of resuscitation efforts and functional outcome after out-of-hospital cardiac arrest: when should we change to novel therapies?   Circulation. 2013;128(23):2488-2494. doi:10.1161/CIRCULATIONAHA.113.002408 PubMedGoogle ScholarCrossref
10.
Reynolds  JC, Grunau  BE, Rittenberger  JC, Sawyer  KN, Kurz  MC, Callaway  CW.  Association between duration of resuscitation and favorable outcome after out-of-hospital cardiac arrest: implications for prolonging or terminating resuscitation.   Circulation. 2016;134(25):2084-2094. doi:10.1161/CIRCULATIONAHA.116.023309 PubMedGoogle ScholarCrossref
11.
Andersen  LW, Granfeldt  A, Callaway  CW,  et al; American Heart Association’s Get With The Guidelines–Resuscitation Investigators.  Association between tracheal intubation during adult in-hospital cardiac arrest and survival.   JAMA. 2017;317(5):494-506. doi:10.1001/jama.2016.20165 PubMedGoogle ScholarCrossref
12.
Andersen  LW, Raymond  TT, Berg  RA,  et al; American Heart Association’s Get With The Guidelines–Resuscitation Investigators.  Association between tracheal intubation during pediatric in-hospital cardiac arrest and survival.   JAMA. 2016;316(17):1786-1797. doi:10.1001/jama.2016.14486 PubMedGoogle ScholarCrossref
13.
Izawa  J, Komukai  S, Gibo  K,  et al.  Pre-hospital advanced airway management for adults with out-of-hospital cardiac arrest: nationwide cohort study.   BMJ. 2019;364:l430. doi:10.1136/bmj.l430 PubMedGoogle Scholar
14.
Matsuyama  T, Komukai  S, Izawa  J,  et al.  Pre-hospital administration of epinephrine in pediatric patients with out-of-hospital cardiac arrest.   J Am Coll Cardiol. 2020;75(2):194-204. doi:10.1016/j.jacc.2019.10.052 PubMedGoogle ScholarCrossref
15.
Okubo  M, Komukai  S, Izawa  J,  et al.  Prehospital advanced airway management for paediatric patients with out-of-hospital cardiac arrest: a nationwide cohort study.   Resuscitation. 2019;145:175-184. doi:10.1016/j.resuscitation.2019.09.007 PubMedGoogle ScholarCrossref
16.
Davis  DP, Garberson  LA, Andrusiek  DL,  et al.  A descriptive analysis of emergency medical service systems participating in the Resuscitation Outcomes Consortium (ROC) network.   Prehosp Emerg Care. 2007;11(4):369-382. doi:10.1080/10903120701537147 PubMedGoogle ScholarCrossref
17.
Morrison  LJ, Nichol  G, Rea  TD,  et al; ROC Investigators.  Rationale, development and implementation of the Resuscitation Outcomes Consortium Epistry—Cardiac Arrest.   Resuscitation. 2008;78(2):161-169. doi:10.1016/j.resuscitation.2008.02.020 PubMedGoogle ScholarCrossref
18.
National Heart, Lung, and Blood Institute. Biologic Specimen and Data Repository Information Coordinating Center. Accessed May 1, 2019. https://biolincc.nhlbi.nih.gov/home/
19.
Link  MS, Berkow  LC, Kudenchuk  PJ,  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):S444-S464. doi:10.1161/CIR.0000000000000261 PubMedGoogle Scholar
20.
Newgard  CD, Haukoos  JS.  Advanced statistics: missing data in clinical research—part 2: multiple imputation.   Acad Emerg Med. 2007;14(7):669-678. doi:10.1197/j.aem.2006.11.038PubMedGoogle Scholar
21.
Li  Y, Propert  K, Rosenbaum  P.  Balanced risk set matching.   J Am Stat Assoc. 2001;96:870-882. doi:10.1198/016214501753208573 Google ScholarCrossref
22.
Lu  B.  Propensity score matching with time-dependent covariates.   Biometrics. 2005;61(3):721-728. doi:10.1111/j.1541-0420.2005.00356.x PubMedGoogle ScholarCrossref
23.
Nakahara  S, Tomio  J, Takahashi  H,  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:f6829. doi:10.1136/bmj.f6829 PubMedGoogle ScholarCrossref
24.
Beyersmann  J, Schumacher  M.  Time-dependent covariates in the proportional subdistribution hazards model for competing risks.   Biostatistics. 2008;9(4):765-776. doi:10.1093/biostatistics/kxn009 PubMedGoogle ScholarCrossref
25.
Stuart  EA.  Matching methods for causal inference: a review and a look forward.   Stat Sci. 2010;25(1):1-21. doi:10.1214/09-STS313 PubMedGoogle ScholarCrossref
26.
Austin  PC.  An introduction to propensity score methods for reducing the effects of confounding in observational studies.   Multivariate Behav Res. 2011;46(3):399-424. doi:10.1080/00273171.2011.568786 PubMedGoogle ScholarCrossref
27.
Zeger  SL, Liang  KY.  Longitudinal data analysis for discrete and continuous outcomes.   Biometrics. 1986;42(1):121-130. doi:10.2307/2531248 PubMedGoogle ScholarCrossref
28.
Koscik  C, Pinawin  A, McGovern  H,  et al.  Rapid epinephrine administration improves early outcomes in out-of-hospital cardiac arrest.   Resuscitation. 2013;84(7):915-920. doi:10.1016/j.resuscitation.2013.03.023 PubMedGoogle ScholarCrossref
29.
Perkins  GD, Kenna  C, Ji  C,  et al.  The influence of time to adrenaline administration in the Paramedic 2 randomised controlled trial.   Intensive Care Med. 2020;46(3):426-436. doi:10.1007/s00134-019-05836-2 PubMedGoogle ScholarCrossref
30.
Hansen  M, Schmicker  RH, Newgard  CD,  et al; Resuscitation Outcomes Consortium Investigators.  Time to epinephrine administration and survival from nonshockable out-of-hospital cardiac arrest among children and adults.   Circulation. 2018;137(19):2032-2040. doi:10.1161/CIRCULATIONAHA.117.033067 PubMedGoogle ScholarCrossref
31.
Elmer  J, Torres  C, Aufderheide  TP,  et al; Resuscitation Outcomes Consortium.  Association of early withdrawal of life-sustaining therapy for perceived neurological prognosis with mortality after cardiac arrest.   Resuscitation. 2016;102:127-135. doi:10.1016/j.resuscitation.2016.01.016 PubMedGoogle ScholarCrossref
32.
Starks  MA, Schmicker  RH, Peterson  ED,  et al; Resuscitation Outcomes Consortium (ROC).  Association of neighborhood demographics with out-of-hospital cardiac arrest treatment and outcomes: where you live may matter.   JAMA Cardiol. 2017;2(10):1110-1118. doi:10.1001/jamacardio.2017.2671 PubMedGoogle ScholarCrossref
33.
Kyriacou  DN, Lewis  RJ.  Confounding by indication in clinical research.   JAMA. 2016;316(17):1818-1819. doi:10.1001/jama.2016.16435 PubMedGoogle ScholarCrossref
34.
Angus  DC.  Whether to intubate during cardiopulmonary resuscitation: conventional wisdom vs big data.   JAMA. 2017;317(5):477-478. doi:10.1001/jama.2016.20626 PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    Emergency Medicine
    August 10, 2021

    Association of Timing of Epinephrine Administration With Outcomes in Adults With Out-of-Hospital Cardiac Arrest

    Author Affiliations
    • 1Department of Emergency Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania
    • 2Division of Biomedical Statistics, Department of Integrated Medicine, Osaka University Graduate School of Medicine, Suita, Japan
    • 3Department of Internal Medicine, Okinawa Prefectural Yaeyama Hospital, Okinawa, Japan
    JAMA Netw Open. 2021;4(8):e2120176. doi:10.1001/jamanetworkopen.2021.20176
    Key Points

    Question  Is timing of epinephrine administration associated with outcomes in adults with out-of-hospital cardiac arrest?

    Findings  In this cohort study with time-dependent propensity score and risk-set matching analysis of 41 079 adult patients from a large out-of-hospital cardiac arrest registry in the United States and Canada, survival to hospital discharge and favorable functional status at hospital discharge were statistically significant and differed according to the timing of epinephrine administration, and the risk ratios for survival and favorable functional status decreased with delayed administration of epinephrine.

    Meaning  Findings of this study suggest that early epinephrine administration is associated with better survival outcomes in adult patients with shockable and nonshockable out-of-hospital cardiac arrest.

    Abstract

    Importance  Administration of epinephrine has been found to be associated with an increased chance of survival after out-of-hospital cardiac arrest (OHCA), but the optimal timing of administration has not been fully investigated.

    Objective  To ascertain whether there is an association between timing of epinephrine administration and patient outcomes after OHCA.

    Design, Setting, and Participants  This cohort study included adults 18 years or older with OHCA treated by emergency medical services (EMS) personnel from April 1, 2011, to June 30, 2015. Initial cardiac rhythm was stratified as either initially shockable (ventricular defibrillation or pulseless ventricular tachycardia) or nonshockable (pulseless electrical activity or asystole). Eligible individuals were identified from among publicly available, deidentified patient-level data from the Resuscitation Outcomes Consortium Cardiac Epidemiologic Registry, a prospective registry of adults with EMS-treated, nontraumatic OHCA with 10 sites in North America. Data analysis was conducted from May 2019 to April 2021.

    Exposures  Interval between advanced life support (ALS)–trained EMS personnel arrival at the scene and the first prehospital intravenous or intraosseous administration of epinephrine.

    Main Outcomes and Measures  The primary outcome was survival to hospital discharge. In each cohort of initial cardiac rhythms, patients who received epinephrine at any period (minutes) after EMS arrival at the scene were matched with patients who were at risk of receiving epinephrine within the same period using time-dependent propensity scores calculated from patient demographic characteristics, arrest characteristics, and EMS interventions.

    Results  Of 41 079 eligible individuals (median [interquartile range] age, 67 [55-79] years), 26 579 (64.7%) were men. A total of 10 088 individuals (24.6%) initially had shockable cardiac rhythms, and 30 991 (75.4%) had nonshockable rhythms. Those who received epinephrine included 8223 patients (81.5%) with shockable cardiac rhythms and 27 901 (90.0%) with nonshockable rhythms. In the shockable cardiac rhythm cohort, the risk ratio (RR) for receipt of epinephrine with survival to hospital discharge was highest between 0 and 5 minutes after EMS arrival (1.12; 95% CI, 0.99-1.26) across the categorized timing of the administration of epinephrine by 5-minute intervals after EMS arrival; however, that finding was not statistically significant. Treating the timing of epinephrine administration as a continuous variable, the RR for survival to hospital discharge decreased 5.5% (95% CI, 3.4%-7.5%; P < .001 for the interaction between epinephrine administration and time to matching) per minute after EMS arrival. In the nonshockable cardiac rhythm cohort, the RR for the association of receipt of epinephrine with survival to hospital discharge was the highest between 0 and 5 minutes (1.28; 95% CI, 0.95-1.72), although not statistically significant, and decreased 4.4% (95% CI, 0.8%-7.9%; P for interaction = .02) per minute after EMS arrival.

    Conclusions and Relevance  Among adults with OHCA, survival to hospital discharge differed across the timing of epinephrine administration and decreased with delayed administration for both shockable and nonshockable rhythms.

    Introduction

    Out-of-hospital cardiac arrest (OHCA) is a major public health problem with high mortality, affecting more than 350 000 individuals annually in the United States.1 Intravenous and intraosseous administration of epinephrine has been widely used for OHCA in the prehospital setting.2 In a recent randomized clinical trial, the use of epinephrine in adults with OHCA increased survival.3 However, evidence about the optimal timing of epinephrine administration is insufficient.4 The 2020 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations5 and 2020 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care6 recommend administration of epinephrine as soon as feasible for individuals with nonshockable cardiac rhythms (strong recommendation with a low certainty of evidence) and suggest administration of epinephrine after initial defibrillation attempts are unsuccessful for shockable cardiac rhythms (weak recommendation with low certainty of evidence). The low certainty of evidence for these recommendations suggests that the optimal timing of epinephrine administration is an existing knowledge gap.4

    A 2019 systematic review indicated that previous studies evaluating the timing of epinephrine had inconsistent findings and a critical risk of bias.7 Notably, none of the included studies addressed an essential factor: resuscitation time bias.8 When timing of an intra-arrest intervention (eg, epinephrine) is assessed, it is crucial to account for this bias.8 Patients cannot achieve return of spontaneous circulation (ROSC) before the intra-arrest intervention.8 Therefore, the late intervention group tends to have longer resuscitation duration and is biased toward harm compared with the early intervention group because longer resuscitation duration is associated with worse outcomes.9,10

    One approach to address resuscitation time bias and time-varying confounders is a time-dependent propensity score and risk-set matching analysis,11-15 which, to our knowledge, has not been used to assess the timing of epinephrine administration for OHCA. The aim of the present study was to use this approach to ascertain whether the timing of epinephrine administration was associated with survival and functional outcomes in patients with OHCA.

    Methods
    Study Design and Setting

    The Resuscitation Outcomes Consortium (ROC) was a clinical research network that conducted trials in OHCA at 10 regional coordinating sites across North America.16,17 In this cohort study, we performed a secondary analysis of patients included in the ROC Cardiac Epidemiologic Registry, a database of prospectively identified, consecutive patients with EMS-treated OHCA (April 2011 through June 2015).16,17 We obtained the publicly available, deidentified patient-level data from the National Heart, Lung, and Blood Institute.18 The institutional review boards at the University of Pittsburgh and Osaka University approved this study and waived the requirement for informed consent because publicly available deidentified data were used. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

    Study Participants

    Included in the study were adults 18 years or older with EMS-treated, non-traumatic OHCA defined as initiation of resuscitation attempts with shock delivery by an external defibrillator (by layperson or EMS personnel) or chest compression by EMS personnel.17 We excluded patients with EMS-witnessed OHCA; those without advanced life support (ALS) involvement; those in whom resuscitations were terminated in the prehospital setting because of confirmation of preexisting written do-not-resuscitate orders; those with missing data on age, initial cardiac rhythm, epinephrine administration status, and primary outcome; those who received vasopressin or endotracheal epinephrine administration; and those with missing or negative values in resuscitation time variables. Resuscitation time variables included intervals between the 9-1-1 call and the first EMS vehicle arrival, between ALS arrival and shock delivery by ALS-trained EMS personnel (if an ALS-trained EMS personnel delivered the shock), between ALS arrival and the first epinephrine administration (if a patient received epinephrine), between ALS arrival and advanced airway management (if a patient received advanced airway management), between ALS arrival and departure from the scene (if a patient was transported), between ALS arrival and prehospital ROSC (if a patient had ROSC), between ALS arrival and prehospital termination of resuscitation (TOR) (if a patient had TOR), and between ALS arrival and hospital arrival (if a patient was transported).

    Exposure and Outcomes

    The main exposure was the interval between ALS-trained EMS personnel arrival at the scene and the first prehospital intravenous or intraosseous administration of epinephrine. The interval was defined in whole minutes; therefore, epinephrine administration at 0 minute indicates that the patient received epinephrine within the same whole minute that EMS arrived.

    The primary outcome was survival to hospital discharge. Secondary outcomes were favorable functional status at hospital discharge, which was defined as a modified Rankin Scale score of 3 or lower, and prehospital ROSC.

    Statistical Analysis

    We stratified patients into 2 cohorts based on their initial cardiac rhythms—shockable (ventricular defibrillation or pulseless ventricular tachycardia) or nonshockable (pulseless electrical activity or asystole) rhythms—because current resuscitation guidelines recommend 2 treatment algorithms according to the initial cardiac rhythms.6,19 In this deidentified data set, age 89 years or older was not specified; therefore, we coded any age 89 years or older as 89 years. We performed multiple imputations to address missing data for functional outcomes, assuming missing at random20; 20 imputed data sets were created through this process, which was conducted after risk-set matching. The regression coefficients for the separately analyzed imputed data sets were averaged and the variances were estimated using mathematical rules as described by Newgard and Haukoos.20 We rounded decimal places to use whole numbers when imputing the number of patients with favorable functional status.

    To assess for an association between the timing of epinephrine administration and outcomes, we performed a time-dependent propensity score and risk-set matching analysis in each cohort of initial cardiac rhythms.11-15,21-23 We calculated the propensity score as the time-varying probability of receiving epinephrine using a Fine-Gray regression model.13-15,24 In the survival analysis model, time to receipt of the first epinephrine administration was the dependent variable, and EMS arrival was time 0 because patients were at risk of receiving epinephrine only after this period. We included the covariates (eg, location, witnessed collapse, and cardiopulmonary resuscitation performed by a bystander) shown in Table 1. Additional methodological details are provided in the eMethods in the Supplement.

    Using the time-dependent propensity scores, we performed 1:1 matching with replacement. Each patient receiving epinephrine at any given minute after EMS arrival was sequentially matched to a patient who was at risk of receiving epinephrine within the same minute to estimate the mean treatment effect (risk-set matching). At-risk patients included those who received epinephrine after the matching and those who never received epinephrine, because matching should be independent of future events.11-15,21,22 At-risk patients could have been subsequently matched multiple times as at-risk patients or as patients receiving epinephrine (only if the patients received epinephrine) until receiving epinephrine (matching with replacement).12,13,15 Matching with replacement was used to decrease bias by reducing the number of unmatched, exposed patients.25 Without replacement, the number of at-risk patients would have decreased as the matching progressed from time 0, and the number of unmatched patients who received epinephrine would have increased.12,15,25 We set the caliper width for the nearest-neighbor matching at 0.2 SD of the propensity scores in the logit scale.25,26 To assess the performance of the risk-set matching, we calculated the standardized difference for each covariate. We considered a standardized difference of less than 0.25 to be a well-matched balance.25

    To ascertain whether there was an association between epinephrine administration and each outcome, we fitted a log link function in generalized estimating equations (GEEs) to estimate risk ratios (RRs) with 95% CIs compared with being at risk of receiving epinephrine (analyses without timing variable).27 We used GEEs to address potential within-pair correlation of risk-set matching. We used frequency weighting adjustment because some patients in the at-risk group could not be independent because of the matching with replacement.25

    To evaluate the timing of epinephrine administration, we fitted 2 models with log link function in GEEs (analyses with timing variables) with frequency weighting adjustment. One model treated the timing of epinephrine as a categorical variable by 5-minute intervals. The other model treated timing of epinephrine administration as a continuous variable. In the model with the continuous variable, we included an interaction term between epinephrine administration and time to matching (ie, time from EMS arrival to the time of matching) and estimated the RRs of epinephrine at each minute, assuming a linear association between each outcome and the timing of epinephrine administration. Additional details are provided in the eMethods in the Supplement. When the P value for the interaction term was significant (P < .05), we considered the timing of epinephrine administration to be associated with the outcome. We calculated the change in RRs with 95% CIs per minute.

    In addition, we conducted 3 sensitivity analyses. First, we performed the risk-set matching without replacement and repeated the same analysis except that we did not use frequency weighting adjustment because at-risk patients were independent. Second, we excluded those who had ROSC or TOR within 5 minutes after EMS arrival because these patients were successfully resuscitated or died before epinephrine could have been feasibly administered. We repeated the same time-dependent propensity score and risk-set matching analysis (matching with replacement). Third, we included only patients with bystander-witnessed arrest and repeated the time-dependent propensity score and risk-set matching analysis (matching with replacement). All tests were 2-sided; we regarded P < .05 as statistically significant. Data analysis was conducted from May 2019 to April 2021. All statistical analyses were performed with R software, version 3.5.1 (R Foundation for Statistical Computing). We reported statistical codes of time-dependent propensity score and risk-set matching in the eMethods in the Supplement.

    Results

    A total of 41 079 adults with a median (interquartile range [IQR]) age of 67 (55-79) years (26 579 men [64.7%] and 14 481 women [35.3%]; 19 patients were missing information on sex), including 10 088 (24.6%) with shockable and 30 991 (75.4%) with nonshockable initial cardiac rhythms, were eligible for inclusion in the analyses (eFigure 1 in the Supplement). Functional outcome data were missing in 573 individuals (5.7%) with shockable and 417 (1.3%) with nonshockable rhythms.

    Table 1 describes participants’ characteristics. Individuals with OHCA who received epinephrine included 8223 (81.5%) with a shockable cardiac rhythm and 27 901 (90.0%) with a nonshockable rhythm. The median (IQR) intervals between EMS arrival and epinephrine administration were 7.3 (5.3-10.0) minutes in those with shockable rhythms and 8.1 (6.0-11.0) minutes in those with nonshockable rhythms.

    Using risk-set matching, 8213 patients with shockable and 27 882 with nonshockable initial cardiac rhythms who received epinephrine were matched with patients at risk of receiving epinephrine (Table 2). Among those matched as at-risk patients, 6626 (80.7%) in the shockable and 23 729 (85.1%) in the nonshockable rhythm cohorts received epinephrine after the matching. In both cohorts, standardized differences were within 0.25 for all variables, indicating a good postmatching balance. For shockable rhythms, median (IQR) intervals between arrival of ALS-trained EMS personnel and epinephrine administration were 7.0 (5.0-9.0) minutes for the epinephrine group and 10.0 (8.0-14.0) minutes for the at-risk group. For nonshockable rhythms, median (IQR) intervals between ALS arrival and epinephrine administration were 8.0 (6.0-11.0) minutes for the epinephrine group and 12.0 (9.0-15.0) minutes for the at-risk group.

    Analyses Without Timing Variable

    For shockable cardiac rhythms, receipt of epinephrine was not associated with survival to hospital discharge compared with being at risk of receiving epinephrine (RR, 0.96; 95% CI, 0.89-1.03) (Table 3). Although receipt of epinephrine was not associated with favorable functional outcome (RR, 0.95; 95% CI, 0.87-1.04), epinephrine was associated with prehospital ROSC (RR, 1.16; 95% CI, 1.12-1.21).

    In the nonshockable cardiac rhythm cohort, survival to hospital discharge (RR, 1.01; 95% CI, 0.88-1.15) and favorable functional outcome (RR, 0.84; 95% CI, 0.68-1.02) did not differ between the epinephrine and at-risk groups. However, receipt of epinephrine was associated with prehospital ROSC (RR, 1.35; 95% CI, 1.31-1.40).

    Analyses With Timing Variables

    Figure 1 and Table 3 show the RRs of epinephrine administration associated with outcomes stratified according to the timing of epinephrine administration for shockable cardiac rhythms. The RR point estimates in the time-dependent propensity score–matched cohorts for the analysis of epinephrine administration and survival to hospital discharge were 1.12 (95% CI, 0.99-1.26) for 0-5 minutes, 1.07 (95% CI, 0.97-1.17) for 5 to 10 minutes, 0.80 (95% CI, 0.66-0.98) for 10 to 15 minutes, 0.55 (95% CI, 0.33-0.89) for 15 to 20 minutes, and 0.13 (95% CI, 0.05-0.37) for more than 20 minutes after ALS arrival (Figure 1A and Table 3). Treating the timing of epinephrine as a continuous variable, RRs decreased 5.5% (95% CI, 3.4%-7.5%; P < .001 for the interaction) for survival to hospital discharge (Figure 1A) and 6.4% (95% CI, 3.8%-8.9%; P < .001 for the interaction) for functional outcome (Figure 1B) per minute after EMS arrival. In contrast, the RR for prehospital ROSC increased 1.4% per minute after EMS arrival (95% CI, 0.2%-2.7%, P = .02 for the interaction) (Figure 1C).

    Figure 2 and Table 3 show the RRs of epinephrine administration associated with outcomes stratified according to the timing of epinephrine administration for nonshockable cardiac rhythms. The point estimates for administration of epinephrine and survival to hospital discharge were 1.28 (95% CI, 0.95-1.72) for 0 to 5 minutes, 1.14 (95% CI, 0.96-1.34) for 5 to 10 minutes, 1.01 (95% CI, 0.75-1.35) for 10 to 15 minutes, 0.60 (95% CI, 0.31-1.15) for 15 to 20 minutes, and 0.36 (95% CI, 0.11-1.23) for more than 20 minutes (Figure 2A and Table 3). Treating the timing of epinephrine as a continuous variable, RRs for survival to hospital discharge decreased 4.4% per minute (95% CI, 0.8%-7.9%; P = .02 for the interaction) (Figure 2A) and 7.1% per minute (95% CI, 1.7%-12.3%; P = .01 for the interaction) for functional outcome (Figure 2B). The RR for prehospital ROSC increased 1.5% per minute after ALS-trained EMS personnel arrival (95% CI, 0.6%-2.4%, P = .001 for the interaction).

    Sensitivity Analyses

    Characteristics of the baseline (eTables 3 and 6 in the Supplement) and matched cohorts (eTables 1, 4, and 7 in the Supplement) had good postmatching balance. Risk ratios of epinephrine administration associated with outcomes in the matched cohorts were similar to the results of the primary analysis (eTables 2, 5, and 8 in the Supplement). eFigures 2-7 in the Supplement show the RR point estimates for receipt of epinephrine associated with outcomes stratified according to the timing of epinephrine administration. For shockable cardiac rhythms, the timing of epinephrine administration (P values for the interaction term) was associated with survival to hospital discharge and favorable functional outcome in all sensitivity analyses. In contrast, for nonshockable rhythms, the timing of epinephrine administration was not associated with survival to hospital discharge or favorable functional outcome in the analyses with risk-set matching without replacement (eFigure 3A and B in the Supplement) and with bystander-witnessed OHCA (eFigure 7A and B in the Supplement). The sensitivity analysis excluding those who had ROSC or TOR within 5 minutes after ALS-trained EMS personnel arrival showed findings similar to the primary analysis in shockable and nonshockable rhythms (eFigures 4 and 5 in the Supplement).

    Discussion

    In this cohort study with a time-dependent propensity score and risk-set matching analysis performed using data from a large OHCA registry with 10 sites in North America, the association of receipt of epinephrine with survival to hospital discharge and favorable functional outcome differed based on the timing of the administration in adults with OHCA and initial cardiac rhythms that were shockable or nonshockable. The findings were consistent among 3 sensitivity analyses for shockable rhythms; however, for nonshockable rhythms, the timing of epinephrine was not associated with survival to hospital discharge and favorable functional outcome in analyses of matching without replacement and with bystander-witnessed OHCAs, which may be explained by the smaller sample size (29 506 patients in the analysis without replacement and 18 836 patients in the analysis of bystander-witnessed OHCAs) and limited outcome events in these analyses.

    Comparison With Previous Studies

    Previous studies have reported inconsistent findings about the timing of epinephrine administration in patients with OHCA. A retrospective observational study in Michigan found that early epinephrine administration (time from 9-1-1 call to epinephrine administration, ≤10 minutes) was not associated with survival to hospital discharge (odds ratio [OR], 0.91; 95% CI, 0.35-2.37) for adults with OHCA compared with late epinephrine administration (time from 9-1-1 call to epinephrine administration, >10 minutes).28 A secondary analysis of a clinical trial in the UK that included 4810 patients with OHCA found that 30-day survival (interaction OR, 0.98; 95% CI, 0.94-1.03) and favorable functional outcomes at hospital discharge (interaction OR, 0.98; 95% CI, 0.93-1.03) were not substantively different over time between the epinephrine and placebo groups, suggesting that timing of epinephrine administration is not an effect modifier on survival and functional outcome.29 In contrast, a secondary analysis of the ROC registry reported that each additional minute of time from EMS arrival to epinephrine administration was associated with a 4% decrease in the odds of survival to hospital discharge (OR, 0.96; 95% CI, 0.95-0.98) for patients with nonshockable cardiac rhythms.30 In a 2019 systematic review, the authors recognized that all included studies investigating the timing of epinephrine administration had a critical risk of bias attributable to confounding and/or selection bias.7 The high degree of heterogeneity across the studies and the serious critical risk of bias precluded any meaningful assessment of the optimal timing of epinephrine administration.7

    Implications

    The latest resuscitation guidelines and international recommendations emphasize early epinephrine administration for OHCA.5,6 Findings of the present study support earlier epinephrine administration for OHCA with shockable and nonshockable cardiac rhythms and provide further evidence to complement these guidelines and recommendations. Another implication is that later epinephrine was found to be associated with ROSC but inversely associated with survival to hospital discharge and favorable functional outcomes. This discordance may suggest that later epinephrine administration might not be beneficial for survival to hospital discharge and functional recovery. The reasons for this discordance are unclear, but it is possible that longer resuscitation time may be associated with poor outcome, and epinephrine is the only intervention associated with increased odds of ROSC in patients with OHCA.

    Strengths and Limitations

    This study has strengths. First, we addressed resuscitation time bias and time-varying confounders. Although the secondary analysis of a clinical trial in the UK also addressed resuscitation time bias given its randomized double-blind design,29 the difference in the results of that study and the present study may be explained by the differences in time to epinephrine administration (median intervals between EMS arrival and epinephrine administration were 14.8 minutes in the epinephrine group and 14.5 minutes in the placebo group in the UK study) and sample size. Second, RRs in the present study should be interpreted as the ratio of the risk of outcomes with epinephrine at any given minute vs the risk of outcomes without epinephrine at the same minute. This interpretation is clinically relevant for deciding whether a patient should receive epinephrine now.

    This study also has limitations. First, the timing of epinephrine administration may be a surrogate of EMS performance (ie, high-performing EMS personnel may administer epinephrine early). Because information on EMS systems was unavailable, we were unable to adjust for the clustering of patients within EMS systems. Similarly, we were unable to adjust for unmeasured confounders, such as patient comorbidity, postresuscitation practice,31 and neighborhood factors.32 Second, we cannot eliminate confounding by indication (ie, EMS personnel may not have administered epinephrine in patients who were expected to have early ROSC without epinephrine or in patients who had early TOR due to futility).33 To account for this, we conducted a sensitivity analysis excluding those who had ROSC or TOR within 5 minutes after ALS arrival and observed consistent results. However, residual confounding by indication may still exist. For example, we defined cases as patients who were successfully administered epinephrine, whereas patients who had delayed epinephrine administration because of difficulty in establishing vascular access could have been matched as a control.34 Third, given the observational design of the present study, we could not demonstrate causation. A clinical trial comparing early vs late epinephrine could assess for a causal association between early epinephrine administration and patient outcomes. However, given the current evidence about the survival benefit of epinephrine,3 such a trial would not be ethically feasible. Fourth, the findings of the present study may not be generalizable to other EMS systems.

    Conclusions

    In this cohort study of more than 40 000 adults with OHCAs in North America, for both initial shockable and nonshockable cardiac rhythms, we found that the associations of epinephrine administration with survival to hospital discharge and favorable functional status at hospital discharge differed on the basis of the timing of administration, and risk ratios for the association between receipt of epinephrine and patient outcomes decreased as administration of epinephrine was delayed.

    Back to top
    Article Information

    Accepted for Publication: May 23, 2021.

    Published: August 10, 2021. doi:10.1001/jamanetworkopen.2021.20176

    Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Okubo M et al. JAMA Network Open.

    Corresponding Author: Masashi Okubo, MD, MS, Department of Emergency Medicine, University of Pittsburgh School of Medicine, 3600 Forbes Ave, Iroquois Building 400A, Pittsburgh, PA 15260 (okubom@upmc.edu).

    Author Contributions: Drs Okubo and Komukai had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Okubo, Komukai, Izawa.

    Acquisition, analysis, or interpretation of data: Okubo, Callaway, Izawa.

    Drafting of the manuscript: Okubo, Komukai, Callaway.

    Critical revision of the manuscript for important intellectual content: Okubo, Callaway, Izawa.

    Statistical analysis: Komukai, Callaway, Izawa.

    Administrative, technical, or material support: Callaway.

    Supervision: Callaway, Izawa.

    Conflict of Interest Disclosures: Dr Callaway reported receiving grants from the National Institutes of Health to study emergency care and cardiac arrest outside the submitted work and previous work in the development of resuscitation guidelines. No other disclosures were reported.

    Additional Contributions: We wish to acknowledge and thank all of the participating emergency medical services personnel, agencies, and medical directors as well as the hospitals that collected and contributed data for the Resuscitation Outcomes Consortium. We thank our colleagues from Osaka University Center of Medical Data Science and the Advanced Clinical Epidemiology Investigator’s Research Project for providing insight and expertise for our research; they were not compensated for their contributions.

    References
    1.
    Virani  SS, Alonso  A, Benjamin  EJ,  et al; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee.  Heart disease and stroke statistics—2020 update: a report from the American Heart Association.   Circulation. 2020;141(9):e139-e596. doi:10.1161/CIR.0000000000000757 PubMedGoogle ScholarCrossref
    2.
    Kiguchi  T, Okubo  M, Nishiyama  C,  et al.  Out-of-hospital cardiac arrest across the world: first report from the International Liaison Committee on Resuscitation (ILCOR).   Resuscitation. 2020;152:39-49. doi:10.1016/j.resuscitation.2020.02.044 PubMedGoogle ScholarCrossref
    3.
    Perkins  GD, Ji  C, Deakin  CD,  et al; PARAMEDIC2 Collaborators.  A randomized trial of epinephrine in out-of-hospital cardiac arrest.   N Engl J Med. 2018;379(8):711-721. doi:10.1056/NEJMoa1806842 PubMedGoogle ScholarCrossref
    4.
    Soar  J, Maconochie  I, Wyckoff  MH,  et al.  2019 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations: summary from the Basic Life Support; Advanced Life Support; Pediatric Life Support; Neonatal Life Support; Education, Implementation, and Teams; and First Aid Task Forces.   Circulation. 2019;140(24):e826-e880. doi:10.1161/CIR.0000000000000734 PubMedGoogle ScholarCrossref
    5.
    Berg  KM, Soar  J, Andersen  LW,  et al; Adult Advanced Life Support Collaborators.  Adult advanced life support: 2020 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations.   Circulation. 2020;142(16)(suppl 1):S92-S139. doi:10.1161/CIR.0000000000000893PubMedGoogle Scholar
    6.
    Panchal  AR, Bartos  JA, Cabañas  JG,  et al; Adult Basic and Advanced Life Support Writing Group.  Part 3: adult basic and advanced life support: 2020 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care.   Circulation. 2020;142(16)(suppl 2):S366-S468. doi:10.1161/CIR.0000000000000916PubMedGoogle Scholar
    7.
    Holmberg  MJ, Issa  MS, Moskowitz  A,  et al; International Liaison Committee on Resuscitation Advanced Life Support Task Force Collaborators.  Vasopressors during adult cardiac arrest: a systematic review and meta-analysis.   Resuscitation. 2019;139:106-121. doi:10.1016/j.resuscitation.2019.04.008 PubMedGoogle ScholarCrossref
    8.
    Andersen  LW, Grossestreuer  AV, Donnino  MW.  “Resuscitation time bias”—a unique challenge for observational cardiac arrest research.   Resuscitation. 2018;125:79-82. doi:10.1016/j.resuscitation.2018.02.006 PubMedGoogle ScholarCrossref
    9.
    Reynolds  JC, Frisch  A, Rittenberger  JC, Callaway  CW.  Duration of resuscitation efforts and functional outcome after out-of-hospital cardiac arrest: when should we change to novel therapies?   Circulation. 2013;128(23):2488-2494. doi:10.1161/CIRCULATIONAHA.113.002408 PubMedGoogle ScholarCrossref
    10.
    Reynolds  JC, Grunau  BE, Rittenberger  JC, Sawyer  KN, Kurz  MC, Callaway  CW.  Association between duration of resuscitation and favorable outcome after out-of-hospital cardiac arrest: implications for prolonging or terminating resuscitation.   Circulation. 2016;134(25):2084-2094. doi:10.1161/CIRCULATIONAHA.116.023309 PubMedGoogle ScholarCrossref
    11.
    Andersen  LW, Granfeldt  A, Callaway  CW,  et al; American Heart Association’s Get With The Guidelines–Resuscitation Investigators.  Association between tracheal intubation during adult in-hospital cardiac arrest and survival.   JAMA. 2017;317(5):494-506. doi:10.1001/jama.2016.20165 PubMedGoogle ScholarCrossref
    12.
    Andersen  LW, Raymond  TT, Berg  RA,  et al; American Heart Association’s Get With The Guidelines–Resuscitation Investigators.  Association between tracheal intubation during pediatric in-hospital cardiac arrest and survival.   JAMA. 2016;316(17):1786-1797. doi:10.1001/jama.2016.14486 PubMedGoogle ScholarCrossref
    13.
    Izawa  J, Komukai  S, Gibo  K,  et al.  Pre-hospital advanced airway management for adults with out-of-hospital cardiac arrest: nationwide cohort study.   BMJ. 2019;364:l430. doi:10.1136/bmj.l430 PubMedGoogle Scholar
    14.
    Matsuyama  T, Komukai  S, Izawa  J,  et al.  Pre-hospital administration of epinephrine in pediatric patients with out-of-hospital cardiac arrest.   J Am Coll Cardiol. 2020;75(2):194-204. doi:10.1016/j.jacc.2019.10.052 PubMedGoogle ScholarCrossref
    15.
    Okubo  M, Komukai  S, Izawa  J,  et al.  Prehospital advanced airway management for paediatric patients with out-of-hospital cardiac arrest: a nationwide cohort study.   Resuscitation. 2019;145:175-184. doi:10.1016/j.resuscitation.2019.09.007 PubMedGoogle ScholarCrossref
    16.
    Davis  DP, Garberson  LA, Andrusiek  DL,  et al.  A descriptive analysis of emergency medical service systems participating in the Resuscitation Outcomes Consortium (ROC) network.   Prehosp Emerg Care. 2007;11(4):369-382. doi:10.1080/10903120701537147 PubMedGoogle ScholarCrossref
    17.
    Morrison  LJ, Nichol  G, Rea  TD,  et al; ROC Investigators.  Rationale, development and implementation of the Resuscitation Outcomes Consortium Epistry—Cardiac Arrest.   Resuscitation. 2008;78(2):161-169. doi:10.1016/j.resuscitation.2008.02.020 PubMedGoogle ScholarCrossref
    18.
    National Heart, Lung, and Blood Institute. Biologic Specimen and Data Repository Information Coordinating Center. Accessed May 1, 2019. https://biolincc.nhlbi.nih.gov/home/
    19.
    Link  MS, Berkow  LC, Kudenchuk  PJ,  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):S444-S464. doi:10.1161/CIR.0000000000000261 PubMedGoogle Scholar
    20.
    Newgard  CD, Haukoos  JS.  Advanced statistics: missing data in clinical research—part 2: multiple imputation.   Acad Emerg Med. 2007;14(7):669-678. doi:10.1197/j.aem.2006.11.038PubMedGoogle Scholar
    21.
    Li  Y, Propert  K, Rosenbaum  P.  Balanced risk set matching.   J Am Stat Assoc. 2001;96:870-882. doi:10.1198/016214501753208573 Google ScholarCrossref
    22.
    Lu  B.  Propensity score matching with time-dependent covariates.   Biometrics. 2005;61(3):721-728. doi:10.1111/j.1541-0420.2005.00356.x PubMedGoogle ScholarCrossref
    23.
    Nakahara  S, Tomio  J, Takahashi  H,  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:f6829. doi:10.1136/bmj.f6829 PubMedGoogle ScholarCrossref
    24.
    Beyersmann  J, Schumacher  M.  Time-dependent covariates in the proportional subdistribution hazards model for competing risks.   Biostatistics. 2008;9(4):765-776. doi:10.1093/biostatistics/kxn009 PubMedGoogle ScholarCrossref
    25.
    Stuart  EA.  Matching methods for causal inference: a review and a look forward.   Stat Sci. 2010;25(1):1-21. doi:10.1214/09-STS313 PubMedGoogle ScholarCrossref
    26.
    Austin  PC.  An introduction to propensity score methods for reducing the effects of confounding in observational studies.   Multivariate Behav Res. 2011;46(3):399-424. doi:10.1080/00273171.2011.568786 PubMedGoogle ScholarCrossref
    27.
    Zeger  SL, Liang  KY.  Longitudinal data analysis for discrete and continuous outcomes.   Biometrics. 1986;42(1):121-130. doi:10.2307/2531248 PubMedGoogle ScholarCrossref
    28.
    Koscik  C, Pinawin  A, McGovern  H,  et al.  Rapid epinephrine administration improves early outcomes in out-of-hospital cardiac arrest.   Resuscitation. 2013;84(7):915-920. doi:10.1016/j.resuscitation.2013.03.023 PubMedGoogle ScholarCrossref
    29.
    Perkins  GD, Kenna  C, Ji  C,  et al.  The influence of time to adrenaline administration in the Paramedic 2 randomised controlled trial.   Intensive Care Med. 2020;46(3):426-436. doi:10.1007/s00134-019-05836-2 PubMedGoogle ScholarCrossref
    30.
    Hansen  M, Schmicker  RH, Newgard  CD,  et al; Resuscitation Outcomes Consortium Investigators.  Time to epinephrine administration and survival from nonshockable out-of-hospital cardiac arrest among children and adults.   Circulation. 2018;137(19):2032-2040. doi:10.1161/CIRCULATIONAHA.117.033067 PubMedGoogle ScholarCrossref
    31.
    Elmer  J, Torres  C, Aufderheide  TP,  et al; Resuscitation Outcomes Consortium.  Association of early withdrawal of life-sustaining therapy for perceived neurological prognosis with mortality after cardiac arrest.   Resuscitation. 2016;102:127-135. doi:10.1016/j.resuscitation.2016.01.016 PubMedGoogle ScholarCrossref
    32.
    Starks  MA, Schmicker  RH, Peterson  ED,  et al; Resuscitation Outcomes Consortium (ROC).  Association of neighborhood demographics with out-of-hospital cardiac arrest treatment and outcomes: where you live may matter.   JAMA Cardiol. 2017;2(10):1110-1118. doi:10.1001/jamacardio.2017.2671 PubMedGoogle ScholarCrossref
    33.
    Kyriacou  DN, Lewis  RJ.  Confounding by indication in clinical research.   JAMA. 2016;316(17):1818-1819. doi:10.1001/jama.2016.16435 PubMedGoogle ScholarCrossref
    34.
    Angus  DC.  Whether to intubate during cardiopulmonary resuscitation: conventional wisdom vs big data.   JAMA. 2017;317(5):477-478. doi:10.1001/jama.2016.20626 PubMedGoogle ScholarCrossref
    ×