Abstract
Background
Little evidence guides the appropriate duration of resuscitation in out-of-hospital cardiac arrest (OHCA), and case features justifying longer or shorter durations are ill-defined. We estimated the impact of resuscitation duration on the probability of favorable functional outcome in OHCA using a large, multi-center cohort.
Methods
Secondary analysis of a North American, single blind, multi-center, cluster-randomized clinical trial (ROC-PRIMED) of consecutive adults with non-traumatic, EMS-treated, OHCA. Primary exposure was duration of resuscitation in minutes (onset of professional resuscitation to return of spontaneous circulation [ROSC] or termination of resuscitation). Primary outcome was survival to hospital discharge with favorable outcome (modified Rankin scale [mRS] 0-3). Subjects were additionally classified as survival with unfavorable outcome (mRS 4-5), ROSC without survival (mRS 6), or without ROSC. Subject accrual was plotted as a function of resuscitation duration, and the dynamic probability of favorable outcome at discharge was estimated for the whole cohort and subgroups. Adjusted logistic regression models tested the association between resuscitation duration and survival with favorable outcome.
Results
The primary cohort included 11,368 subjects (median age 69 years [IQR: 56-81 years]; 7,121 men [62.6%]). Of these, 4,023 (35.4%) achieved ROSC, 1,232 (10.8%) survived to hospital discharge, and 905 (8.0%) had mRS 0-3 at discharge. Distribution of CPR duration differed by outcome (p<0.00001). For CPR duration up to 37.0 minutes (95%CI 34.9-40.9 minutes), 99% with eventual mRS 0-3 at discharge achieved ROSC. Dynamic probability of mRS 0-3 at discharge declined over elapsed resuscitation duration, but subjects with initial shockable cardiac rhythm, witnessed cardiac arrest, and bystander CPR were more likely to survive with favorable outcome after prolonged efforts (30-40 minutes). Adjusting for prehospital (OR 0.93; 95%CI 0.92-0.95) and inpatient (OR 0.97; 95%CI 0.95-0.99) covariates, resuscitation duration was associated with survival to discharge with mRS 0-3.
Conclusions
Shorter resuscitation duration was associated with likelihood of favorable outcome at hospital discharge. Subjects with favorable case features were more likely to survive prolonged resuscitation up to 47 minutes.
Keywords: cardiac arrest, cardiopulmonary resuscitation, termination of resuscitation
Journal Subject Terms: Cardiopulmonary Resuscitation and Emergency Cardiac Care, Cardiopulmonary Arrest
Introduction
More than 356,000 persons experience sudden out-of-hospital cardiac arrest (OHCA) annually in the United States, [1-3] and little evidence guides the appropriate duration of resuscitation in OHCA. When initial prehospital resuscitation measures fail to achieve return of spontaneous circulation (ROSC), some emergency medical services (EMS) systems transport patients with ongoing CPR to hospital, where the same therapies are typically repeated. In either setting, few additional resources are brought to bear, and most patients without rapid ROSC will not survive or experience good functional outcome. [4] Other EMS systems terminate the resuscitation based on case features, termination of resuscitation (TOR) guidelines, and/or practical consideration of resuscitation duration. The ideal incorporation of resuscitation duration into decisions to continue or terminate resuscitation is unclear. Some guidelines advocate ≥20 minutes of resuscitation [5-7], whereas others require structured assessment after three cycles of CPR and rhythm analysis (6 minutes) [8] or merely brief attempts (1-2 minutes) in subjects with unfavorable case features. [9] Still others do not incorporate resuscitation duration, [10] leaving clinicians to make ad hoc decisions at their discretion.
Multi-center, observational data on in-hospital cardiac arrest (IHCA) suggest that patients treated at hospitals with systematically longer durations of resuscitation attempts have higher likelihood of ROSC and survival to discharge. [11] Other than one observational study in Japanese bystander-witnessed OHCA, [12] similarly robust patient-level data are limited for OHCA. In a single site registry, we previously identified that after 20 minutes of traditional interventions, repeating these yielded minimal incremental survival with favorable neurologic outcome.[4]
Our objective was to estimate the impact of resuscitation duration (until ROSC or TOR) on the probability of favorable functional outcome in OHCA using a large, multi-center cohort. We analyzed clinical trial data to plot accrual of subjects with favorable and unfavorable outcomes over time, estimate dynamic probabilities of favorable functional outcome as a function of resuscitation duration, and test the association between resuscitation duration and favorable functional outcome. We also performed subgroup analyses to determine if subjects with particular phenotypes or case features justified prolonging or terminating resuscitation.
Methods
Data Source
We examined de-identified clinical trial data from the Resuscitation Outcomes Consortium (ROC) Prehospital Resuscitation using an IMpedance valve and Early versus Delayed (PRIMED) trial, [13,14] which was conducted using exception from informed consent under United States and Canadian regulations. Institutional Review Boards (IRB) at participating sites provided approval. Trial data was obtained from National Institutes of Health Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC). Michigan State University IRB granted waiver of consent for this secondary analysis of a de-identified dataset. The original trials were neutral for differences between interventions, [13,14] avoiding confounding from trial arms.
ROC comprises 10 regional centers across North America and their respective EMS systems. The geographic footprint of ROC spans 218 prehospital agencies covering 35,000 square miles and 24 million persons. From June 2007 to July 2010, 150 EMS agencies participated in PRIMED. Detailed description of methodology has been given previously. [15,16] PRIMED studied two different resuscitation strategies: thirty-seconds vs. three-minutes of CPR prior to initial rhythm analysis, and an impedance valve vs. sham device during OHCA. [13,14] Research personnel at participating sites prospectively collected patient-level data, including prehospital data streams and audio recordings, from consecutive OHCA subjects. They reviewed hospital records to determine in-hospital interventions, in-hospital mortality, and functional status at hospital discharge. Modified Rankin Scale (mRS) score was assigned at hospital discharge with a standardized chart-review instrument. All participating ROC sites provided data for this study.
Study Design and Population
This was a retrospective cohort study of EMS-treated adult subjects experiencing OHCA. Cardiac arrest was defined as receiving chest compressions or rescue shocks from a professional provider or automated external defibrillator. We excluded suspected and confirmed cases of traumatic arrest.
Study Definitions and Outcomes
The primary outcome was survival to hospital discharge with favorable functional status, (mRS 0-3). We classified subjects into three other groups based on resuscitation outcome: survival to hospital discharge with unfavorable functional status (mRS 4–5), ROSC without survival to hospital discharge (mRS 6), and no ROSC. Resuscitation duration (CPR duration in minutes) was the primary independent variable. CPR duration was defined as elapsed time from first professional chest compression to first ROSC or TOR. TOR guidelines were according to local protocol without input from the primary investigation or ROC. Some sites employ rigid guidelines [6], whereas others allow paramedics to terminate resuscitation at the scene after consultation with a physician. Among ROC sites, transport is typically initiated in absence of documented ROSC for 58% of cases (site range: 14-95%). [17]
Time-stamped data (hours:minutes:seconds) on initiation and conclusion of CPR by EMS providers were recorded by the monitor-defibrillator, which stayed with the patient throughout resuscitation. CPR was detected indirectly by changes in thoracic impedance recorded from external defibrillator electrodes, or directly by an accelerometer between the rescuer and patient's chest. CPR quality for the initial 10 minutes of resuscitation was assessed by chest compression fraction, the proportion of time in which chest compressions were performed during each minute of resuscitation.
Statistical Analyses
Analyses were performed with STATA 12.0 (StataCorp, College Station, TX). We stratified subjects by outcome and tabulated subject characteristics, cardiac arrest characteristics, EMS interventions, and inpatient interventions. We compared these variables across outcomes with one way ANOVA (Kruskal Wallis test for non-parametric variables) and chi-square or Fisher's exact test. In this de-identified dataset, age >89 years was not specified. To preserve age as a continuous variable, we coded any age >89 years as 90 years.
We constructed simple curves of the proportion of subjects achieving ROSC over time, stratified by outcome (mRS 0–3, mRS 4–5, mRS 6, no ROSC), and compared distributions with log-rank test. We then estimated 50th, 75th, and 99th percentiles of CPR duration for each stratum.
We also calculated the dynamic probability of survival to hospital discharge with mRS 0-3 among all attempted resuscitations and pre-defined subgroups to determine if subjects with particular case features associated with favorable outcome (witnessed arrest, shockable initial rhythm, bystander CPR, and quartiles of EMS dispatch interval [911 call to onset of professional resuscitation]) have incremental benefit in prolonging resuscitation. We plotted dynamic probabilities of mRS 0-3 for representative subject phenotypes with different combinations of these case features.
Finally, we created unadjusted and adjusted logistic regression models to test the association between CPR duration and mRS 0-3 at hospital discharge. Because factors that affect short-term survival may have blunted or opposite relationships with later survival [18], we used two separate models to test the association between CPR duration and survival to hospital discharge with mRS 0-3: a ‘prehospital’ model and an ‘inpatient’ model. The prehospital model containing all subjects was adjusted for a priori prehospital covariates associated with outcome (age, sex, witnessed arrest, bystander CPR, EMS dispatch interval, shockable initial rhythm, mean chest compression fraction, advanced airway attempts [endotracheal intubation or supraglottic airway], and epinephrine administration). [13,14] The inpatient model containing only subjects that survived to hospital admission was further adjusted for a priori inpatient interventions associated with outcome (induced hypothermia [19,20] and cardiac catheterization [21,22]). We tested model discrimination with the c-statistic and model fit with Hosmer-Lemeshow goodness of fit test. Unfortunately, the de-identified nature of this dataset precluded determination of the site at which subjects were enrolled, rendering adjustment for regional variability or clustering impossible.
Results
Among a 24 million total catchment population, the PRIMED dataset captured 17,445 subjects. Of these, 11,368 had complete time and outcome data. (Figure 1) Among 11,368 attempted resuscitations, 4,023 (35.4%) achieved ROSC, 1,232 (10.8%) survived to hospital discharge, and 905 (8.0%) had mRS 0-3 at hospital discharge. Median professional CPR duration was 20 minutes (IQR: 12-27.3 minutes) overall, 13.5 minutes (IQR: 8-20 minutes) for those with ROSC, and 23.4 minutes (IQR: 16.5-30 minutes) for those without ROSC. The longest observed duration of CPR was 202 minutes in any subject, and 47 minutes in subjects with eventual mRS 0-3.
Figure 1.
Study cohort and exclusions. CPR: cardiopulmonary resuscitation.
Table 1 contains demographic and clinical features stratified by outcome. CPR duration, and prevalence of case features, EMS interventions, and inpatient interventions differed by outcome. Figure 2 shows curves and estimated percentiles of CPR duration for subjects achieving ROSC, stratified by patient outcome. The distribution of CPR duration differed across strata (p<0.00001).
Table 1.
Clinical features stratified by functional outcome at hospital discharge.
| Missing Data (n=11,368) | mRS 0-3 (n=905) | mRS 4-5 (n=327) | mRS 6 (n=2,791) | ROSC=0 (n=7,345) | p-value | |
|---|---|---|---|---|---|---|
| Baseline characteristics | ||||||
| Age (years) | 38 (0.3%) | 59 (IQR: 49, 68) | 66 (IQR: 57, 76) | 72 (IQR: 59, 82) | 70 (IQR: 56, 82) | 0.0001 |
| Male sex | 1 (0.01%) | 666 (73.6%) | 205 (62.7%) | 1,650 (59.1%) | 4,600 (62.6%) | < 0.001 |
| Residence prior to event | < 0.001 | |||||
| Home | 880 (97.2%) | 276 (84.4%) | 1,616 (60.0%) | 220 (3.0%) | ||
| Rehabilitation center | 1 (0.1%) | 4 (1.2%) | 72 (0.3%) | 4 (0.1%) | ||
| Assisted living facility | 6 (0.7%) | 9 (2.8%) | 53 (1.9%) | 11 (0.2%) | ||
| Nursing home | 3 (0.3%) | 31 (9.5%) | 165 (5.9%) | 7 (0.1%) | ||
| Unknown | 15 (1.7%) | 7 (2.1%) | 890 (31.9%) | 7,110 (96.8%) | ||
| Event characteristics | ||||||
| Witnessed status | 0 (0%) | 739 (81.7%) | 254 (77.7%) | 1,718 (61.6%) | 2,606 (35.5%) | < 0.001 |
| EMS-witnessed | 0 (0%) | 189 (20.9%) | 52 (15.9%) | 353 (12.6%) | 408 (5.6%) | < 0.001 |
| Bystander-witnessed | 1,260 (11%) | 550 (60.8%) | 202 (61.8%) | 1,365 (48.9%) | 2,198 (29.9%) | < 0.001 |
| Bystander CPR | 0 (0%) | 415 (45.9%) | 138 (42.2%) | 1,102 (39.5%) | 2,673 (36.4%) | < 0.001 |
| AED shock delivered | 11,196 (98%) | 25 (2.8%) | 6 (1.8%) | 29 (1.0%) | 28 (0.4%) | < 0.001 |
| Initial ECG Rhythm | < 0.001 | |||||
| VF/VT | 673 (74.4%) | 166 (50.8%) | 706 (25.3%) | 977 (13.3%) | ||
| PEA | 129 (14.3%) | 97 (29.6%) | 928 (33.3%) | 1,421 (19.4%) | ||
| Asystole | 45 (5.0%) | 45 (13.8%) | 900 (32.3%) | 4,190 (57.1%) | ||
| AED – no shock advised | 34 (3.8%) | 11 (3.4%) | 232 (8.3%) | 690 (9.4%) | ||
| Unknown | 11 (1.2%) | 3 (0.9%) | 18 (0.6%) | 64 (0.9%) | ||
| EMS interventions | ||||||
| 911 call – EMS arrival (min) | 0 (0%) | 7.8 (IQR: 6.1, 10.5) | 8.1 (IQR: 6.0, 10.7) | 8.5 (IQR: 6.7, 11.1) | 8.5 (IQR: 6.8, 10.7) | 0.0001 |
| CPR duration (min) | 0 (0%) | 7.7 (IQR: 3.9, 12.7) | 10.0 (IQR: 5.7, 15.3) | 15.7 (IQR: 10.5, 21.7) | 23.4 (IQR: 16.5, 30.0) | 0.0001 |
| Mean CCF first 10 minutes | 3,843 (33%) | 0.66 ± 0.19 | 0.67 ± 0.18 | 0.72 ± 0.15 | 0.72 ± 0.14 | < 0.001 |
| Advanced airway attempted | 0 (0%) | 680 (75.1%) | 280 (85.6%) | 2,633 (94.3%) | 6,120 (83.3%) | < 0.001 |
| Endotracheal intubation | 0 (0%) | 639 (70.6%) | 263 (80.4%) | 2,455 (88.0%) | 5,518 (75.1%) | < 0.001 |
| Supraglottic airway | 0 (0%) | 60 (6.6%) | 27 (8.3%) | 272 (9.8%) | 922 (12.6%) | < 0.001 |
| Epinephrine administered | 33 (0.3%) | 376 (41.6%) | 211 (64.5%) | 2,451 (87.8%) | 6,160 (83.9%) | < 0.001 |
| Dose epinephrine (mg) | 2,180 (19%) | 2.2 ± 1.7 | 2.4 ± 1.7 | 2.8 ± 1.8 | 3.8 ± 1.9 | < 0.001 |
| Any shock delivered | 4 (0.04%) | 711 (78.6%) | 194 (59.3%) | 1,113 (39.9%) | 2,081 (28.3%) | < 0.001 |
| Number of shocks delivered | 7,273 (64%) | 2 (IQR: 1, 4) | 2 (IQR: 1, 4) | 2 (IQR: 1, 4) | 2 (IQR: 1, 4) | 0.33 |
| Inpatient interventions | ||||||
| Therapeutic hypothermia | 0 (0%) | 453 (50.0%) | 177 (54.1%) | 856 (30.7%) | < 0.001 | |
| Cardiac catheterization | 1,783 (16%) | 457 (50.5%) | 84 (25.7%) | 184 (6.6%) | < 0.001 | |
| PCI | 1,783 (16%) | 334 (36.9%) | 59 (18.0%) | 152 (5.5%) | < 0.001 | |
| CABG | 1,783 (16%) | 67 (7.4%) | 11 (3.4%) | 6 (0.2%) | n/a | < 0.001 |
| Pacemaker/ICD implant | 0 (0%) | 308 (34.0%) | 37 (11.3%) | 21 (0.8%) | < 0.001 | |
| ICU length of stay (days) | 859 (8%) | 6 (IQR: 4, 11) | 10 (IQR: 6, 18) | 3 (IQR: 1, 5) | 0.0001 | |
| DNR order | 836 (7%) | 26 (2.9%) | 99 (30.3%) | 1,732 (62.1%) | < 0.001 | |
Continuous variables are given either as mean ± standard deviation, or median (interquartile range). Categorical variables are given as count (%). EMS: emergency medical services. CCF: chest compression fraction. ECG: electrocardiogram. VF: ventricular fibrillation. VT: ventricular tachycardia. PEA: pulseless electrical activity. AED: automated external defibrillator. min: minutes. CPR: cardiopulmonary resuscitation. SBP: systolic blood pressure. PCI: percutaneous coronary intervention. CABG: coronary artery bypass graft. ICD: implantable cardiac defibrillator. ICU: intensive care unit. DNR: do not resuscitate. N/A: not applicable.
Figure 2.
Distribution of CPR duration for all patients with attempted resuscitation, stratified by outcome (p<0.00001). Point estimates and 95% confidence interval for percentiles of CPR duration are provided in the accompanying Table. The upper bound of the 95% confidence interval for the 99th percentile of CPR duration could not be estimated for mRS 4-5, since it exceeded the values of this dataset. ROSC: return of spontaneous circulation. mRS: modified Rankin scale.
Figures 3-4 present dynamic probabilities and 95% confidence intervals of survival with mRS 0-3 for case features and patient phenotypes. Each point on the curve represents favorable functional outcome among subjects with CPR durations greater than or equal to the respective interval. We did not observe systematic bias for or against prolonged resuscitation among patient phenotypes. Even though median CPR duration differed by case features (Supplemental Table), the likelihood of still receiving CPR at a given elapsed interval was consistent across phenotypes (50% of subjects were still receiving CPR at 20 minutes, 20% at 30 minutes, and 5% at 40 minutes) (Figure 4). Stratification by initial cardiac rhythm had the greatest discrepancy in probability of mRS 0-3, followed by witnessed cardiac arrest, bystander CPR, and EMS dispatch interval. (Figure 3) Subjects with shockable initial cardiac rhythm, witnessed arrest, and bystander CPR had the highest probability of mRS 0-3 at hospital discharge: twofold to threefold higher than other phenotypes (Figure 4).
Figure 3.
Dynamic probability (black lines) with 95% confidence intervals (gray lines) of survival to hospital discharge with mRS 0-3 as a function of CPR duration, stratified by initial shockable rhythm (A), witnessed cardiac arrest (B), bystander CPR (C), and quartile of elapsed interval from 911 dispatch to onset of professional resuscitation (D). mRS: modified Rankin scale. CPR: cardiopulmonary resuscitation.
Figure 4.
Dynamic probability (black lines) with 95% confidence intervals (gray lines) of survival to hospital discharge with mRS 0-3 as a function of CPR duration, for those with (A) and without (B) shockable initial cardiac rhythms, stratified by patient phenotype with different combinations of case features. mRS: modified Rankin scale. CPR: cardiopulmonary resuscitation.
Table 2 contains logistic regression models. In unadjusted analysis, CPR duration (minutes) was associated with survival to hospital discharge with mRS 0-3 (OR 0.87; 95% CI 0.86-0.87; p<0.0001). Adjusting for prehospital covariates, CPR duration was independently associated with survival to hospital discharge with mRS 0-3 (OR 0.93; 95%CI 0.92-0.95; p<0.0001) (c-statistic 0.93). Further adjusting for inpatient covariates among subjects surviving to hospital admission, the independent association persisted (OR 0.97; 95%CI 0.95-0.99; p=0.04) (c-statistic 0.91). Final adjusted models had acceptable fit (p=0.97 and p=0.98). Given 33% missing data for mean chest compression fraction, we performed post-hoc sensitivity analyses with multiple imputations for mean chest compression fraction in the logistic regression models. Neither the magnitude or significance of the odds ratios for both the predictor variable of interest (CPR duration) and the imputed variable (mean chest compression fraction) differed between models with complete cases and imputed data.
Table 2.
Unadjusted and adjusted logistic regression models for mRS 0-3 on hospital discharge.
| Crude Odds Ratio (95% CI) | p-value | Adjusted Odds Ratio (95% CI) | p-value | |
|---|---|---|---|---|
| Prehospital Model (c-statistic 0.93) | ||||
| CPR duration (minutes) | 0.87 (0.86, 0.87) | < 0.0001 | 0.93 (0.92, 0.95) | < 0.0001 |
| Age (years) | 0.97 (0.96, 0.97) | < 0.0001 | 0.95 (0.94, 0.96) | < 0.0001 |
| Male sex | 1.73 (1.48, 2.02) | < 0.0001 | 1.23 (0.88, 1.71) | 0.22 |
| Witnessed arrest (EMS or bystander) | 5.72 (4.82, 6.80) | < 0.0001 | 2.26 (1.62, 3.15) | < 0.001 |
| Bystander CPR | 1.42 (1.24, 1.63) | < 0.0001 | 0.89 (0.66, 1.18) | 0.41 |
| Dispatch to first professional CPR (minutes) | 1.01 (1.01, 1.02) | 0.002 | 0.94 (0.91, 0.98) | 0.002 |
| Shockable initial rhythm | 13.51 (11.55, 15.82) | < 0.0001 | 13.66 (9.67, 19.29) | < 0.0001 |
| Mean CCF first 10 minutes | 0.09 (0.06, 0.15) | < 0.0001 | 1.38 (0.53, 3.63) | 0.51 |
| Advanced airway attempted | 0.66 (0.57, 0.77) | < 0.0001 | 0.64 (0.35, 1.16) | 0.14 |
| Epinephrine administered (mg) | 0.13 (0.11, 0.15) | < 0.0001 | 0.68 (0.61, 0.76) | < 0.0001 |
| Inpatient Model (c-statistic 0.91) | ||||
| CPR duration (minutes) | 0.90 (0.89, 0.91) | < 0.0001 | 0.97 (0.95, 0.99) | 0.04 |
| Age (years) | 0.97 (0.96, 0.97) | < 0.0001 | 0.96 (0.95, 0.97) | < 0.0001 |
| Male sex | 1.77 (1.50, 2.10) | < 0.0001 | 1.22 (0.81, 1.84) | 0.35 |
| Witnessed arrest (EMS or bystander) | 2.51 (2.08, 3.03) | < 0.0001 | 1.61 (1.07, 2.42) | 0.02 |
| Bystander CPR | 1.19 (1.02, 1.39) | 0.03 | 0.77 (0.54, 1.11) | 0.17 |
| Dispatch to first professional CPR (minutes) | 1.01 (1.01, 1.02) | 0.04 | 0.93 (0.88, 0.97) | 0.003 |
| Shockable initial rhythm | 6.27 (5.28, 7.45) | < 0.0001 | 3.69 (2.38, 5.72) | < 0.001 |
| Mean CCF first 10 minutes | 0.15 (0.09, 0.26) | < 0.0001 | 0.83 (0.26, 2.69) | 0.76 |
| Advanced airway attempted | 0.38 (0.32, 0.46) | < 0.0001 | 0.36 (0.13, 0.99) | 0.04 |
| Epinephrine administered (mg) | 1.13 (0.11, 0.16) | < 0.0001 | 0.85 (0.75, 0.96) | 0.01 |
| Therapeutic hypothermia | 1.34 (1.15, 1.57) | < 0.0001 | 1.51 (0.99, 2.29) | 0.06 |
| Cardiac catheterization | 12.41 (10.12, 15.22) | < 0.0001 | 7.63 (5.14, 11.35) | < 0.001 |
CI: confidence interval. CPR: cardiopulmonary resuscitation. EMS: emergency medical services. CCF: chest compression fraction.
Discussion
In a multi-center cohort of >11,000 subjects, we observed rapidly diminishing probability of favorable functional status at hospital discharge with increasing durations of CPR. After 37.0 minutes of CPR (95%CI 34.9-40.9 minutes), 99% of subjects with eventual mRS 0-3 at hospital discharge had achieved ROSC (Figure 2). Moreover, we calculated dynamic, time-based probability estimates of favorable outcome for specific key patient features (Figures 3 and 4). CPR duration was independently associated with favorable functional status, adjusting for both prehospital and inpatient covariates (Table 2).
The appropriate duration of resuscitation in cardiac arrest is fraught with clinical and ethical implications. Our findings suggest that using conventional resuscitation, 90% of subjects with good outcome have ROSC within 20 minutes, and 99% within 37 minutes (Figure 2). Our data can address three clinical questions: 1) In which phenotypes/subgroups are prolonged resuscitation efforts justified? 2) In which subgroups may shorter attempts prior to TOR be appropriate? 3) When should novel resuscitation strategies be initiated in appropriate candidates?
Prolonged resuscitation efforts appear worthwhile in subjects with shockable initial cardiac rhythm, witnessed cardiac arrest, bystander CPR, or brief interval between collapse and professional resuscitation (Figure 3). Shockable initial rhythm appears to be the best prognostic indicator (Figures 3A and 4A). Among those with shockable initial cardiac rhythms, witnessed cardiac arrest appears to be a better prognostic indicator than bystander CPR (Figure 4A). This may reflect a greater ischemic insult in unwitnessed cardiac arrest that is not reversible even with earlier reperfusion provided by bystander CPR. The longest duration of professional CPR observed in any subject with eventual mRS 0-3 was 47 minutes. We observed no additional accrual of subjects with eventual mRS 0-3 beyond 47 minutes among those with shockable initial cardiac rhythm, 41 minutes among those with witnessed cardiac arrest, and 47 minutes among those with bystander CPR. Because resuscitation efforts longer than this were so rare, these data cannot determine futility to continue efforts longer. Likewise, the median duration of CPR was 23.4 minutes in subjects without ROSC – it is unknown how those subjects would have fare had resuscitation continued longer. These findings are consistent with Grunau, et al. and Nehme, et al. who studied the impact of shockable initial cardiac rhythm and EMS-witnessed cardiac arrest on resuscitation duration and survival in a single EMS system. [23,24] They found higher ‘resilience’ to resuscitation lasting 30-40 minutes in subjects with shockable initial cardiac rhythms and EMS-witnessed cardiac arrest. Nonetheless, we expect that beyond some ultimate duration of resuscitation, incremental survival becomes asymptotic.
Goldberger, et al. compared median duration of resuscitation attempts in non-survivors of IHCA to assess hospitals' tendency for duration of resuscitation efforts, and related this tendency to survival. [11] Subjects at hospitals with longer resuscitation attempts were more likely to achieve ROSC and survive to hospital discharge. Importantly, there was no difference in proportion of good functional outcome at hospital discharge between groups. This was compelling evidence that prolonging resuscitation efforts can increase survival without substantial increase in severe neurology injury among survivors. Our data identify patient phenotypes in OHCA most likely to benefit from this approach.
Conversely, clinicians may use resuscitation duration in the absence of ROSC to justify TOR after some elapsed interval. At face value, the probability curves in Figure 3 appear to support this in subjects with unfavorable case features. One commonly accepted definition of medical futility is <1% probability of success. [25,26] The upper bound of the 95% confidence interval fell below 1% after 12 minutes of CPR in the subgroup with non-shockable initial cardiac rhythm, and 17 minutes of CPR in the subgroup with unwitnessed cardiac arrest. However, hypothetical termination at these points would have missed 53 (23%) and 26 (16%) subsequent subjects with favorable outcome, respectively. Drennan, et al. tested the combination of CPR duration and absence of ROSC as a hypothetical TOR rule. [27] In that cohort, hypothetical TOR based solely on absence of ROSC after 20 minutes of resuscitation would have missed 10% of all survivors and 10% of survivors with favorable functional outcome. Taken together, these data argue against using resuscitation duration in isolation or with ad hoc case features to justify TOR. Instead, we turn attention toward validated TOR decision rules. [28,29]
Finally, our findings support consideration of novel resuscitation strategies in appropriate candidates who do not immediately respond to conventional resuscitation. The current strategies have been optimized for 60 years, but the essence of resuscitation has not fundamentally changed. A new paradigm may be needed to achieve more than modest improvements in patient outcome. One such intervention is extra-corporeal CPR (E-CPR), the incorporation of extra-corporeal life support into cardiac arrest resuscitation. [30] This resource-intensive therapy is associated with improved functionally favorable survival in selected candidates with favorable case features. [31-33] However, the cost and resource-intensity of E-CPR mandate that it be applied in a rational manner with optimal chance to benefit patients. Our data demonstrate declining proportions of subjects who have favorable recovery with each minute that traditional CPR fails to achieve ROSC. Furthermore traditional resuscitation usually fails, making it reasonable to mobilize a novel therapy like E-CPR early after recognition of cardiac arrest with a favorable phenotype that can withstand prolonged efforts, concurrently with traditional CPR. Considering the time demands of transporting to hospital and initiating E-CPR, early mobilization is also logistically necessary to implement E-CPR within the therapeutic window. However, the challenges of providing quality CPR during transport may reduce survival for those who would have achieved ROSC with further on-scene resuscitation. Considering the trajectory of the curves in Figures 2-4, 50% of subjects with eventual mRS 0-3 at hospital discharge had achieved ROSC by ∼8 minutes, and 90% of subjects by ∼20 minutes. The likelihood of accruing additional cases with eventual mRS 0-3 beyond 20 minutes fell to ∼1-15% depending on subject phenotype. Taken together, these data suggest that 8-20 minutes of professional resuscitation is a reasonable window to mobilize toward E-CPR. This timeframe may shift forward or backward depending patient phenotype (i.e. shockable initial cardiac rhythm, witnessed cardiac arrest, bystander CPR). In those who achieve ROSC rapidly with traditional CPR, mobilization of novel therapy can be discontinued. Grunau, et al. found similar results in a regional cohort of 1,206 hypothetical E-CPR eligible subjects: 16 minutes of professional on-scene resuscitation best balanced the risks and benefits of early vs. later transport. [34]
Nagao, et al. calculated the minimum duration of prehospital resuscitation efforts among bystander-witnessed OHCA to achieve ≥99% sensitivity for favorable 30-day neurologic outcome in a nation-wide, population-based Japanese registry.[12] Depending on the phenotype (shockable initial cardiac rhythm and bystander resuscitation), they concluded that prehospital resuscitation efforts should be continued for at least 40-45 minutes in all adults with bystander-witnessed OHCA. They also found a steady decline in the likelihood of favorable outcomes with increasing duration of resuscitation (adjusted OR 0.84; 95%CI 0.83-0.84). Our findings from a North American clinical trial dataset are remarkably consistent despite important differences in study design and setting. Whereas Nagao, et al. included only subjects with bystander-witnessed OHCA, we included all subjects with attempted professional resuscitation, irrespective of witnessed status. We were also able to account for some degree of CPR quality (CPR fraction) and post-cardiac arrest care (therapeutic hypothermia and cardiac catheterization). Furthermore, prehospital providers in Japan must continue resuscitation efforts until ROSC or hospital arrival; they are not legally permitted to terminate resuscitation efforts in the field. [35] On one hand, this setting is ideal to observe the relationship between resuscitation duration and outcomes without confounding elements of TOR practices. However, unlike Japan, most of the world incorporates some form of TOR to reduce futile transport to hospital. Therefore, our data are useful for generalizability to other countries. Similar to Nagao, et al. we found that 42.0 minutes (95%CI 40.5-45.0 minutes) of professional resuscitation accrued 99% of all survivors, and 37.0 minutes (95%CI 34.9-40.9 minutes) accrued 99% of those with favorable functional outcome (Figure 2). We also identified a similar decline in probability of favorable functional outcome over elapsed duration of professional resuscitation (adjusted OR 0.93; 95%CI 0.92-0.95).
Limitations
The de-identified nature of this dataset rendered it impossible to make site-specific comparisons or adjust for regional variability or clustering. We do know there is variation in patient features and survival between ROC sites. [36] Some subjects were transported to hospital during ongoing resuscitation; it is unknown if this affected the likelihood of favorable outcome. Furthermore, no data describe subject comorbidities, degree of post-cardiac arrest illness severity, inpatient prognostication, and other care processes associated with clinical outcomes (e.g. oxygenation, ventilation, hemodynamic management, glycemic control, etc.). However, the advantage of these trial data (as opposed to self-reported registry data) is the high degree of accuracy, especially in prehospital time-stamped data [37] and structured, prospective outcome assessment. We do note that many subjects did not have complete time-stamped data (Figure 1). Summary features of excluded subjects were comparable to included subjects. Finally, the elapsed interval from 911 call to onset of professional resuscitation is a best estimate of ischemic insult, but does not capture the full magnitude of “no-flow time” for subjects with unwitnessed cardiac arrest or delays in activation of the emergency response system. Likewise, it does not capture presence or quality of bystander resuscitation attempts.
Conclusions
Each elapsed minute of resuscitation is independently associated with lower odds of favorable functional outcome at hospital discharge. However, those with favorable case features (shockable initial cardiac rhythm, bystander CPR, witnessed cardiac arrest) were more likely to survive with favorable functional outcome after longer resuscitations, and warrant early consideration of novel therapies such as E-CPR. Resuscitation duration in the absence of ROSC should not be used as an ad hoc criterion for TOR. In the present model of OHCA resuscitation in North America, few are likely to have favorable outcome after 47 minutes of CPR.
Supplementary Material
Clinical Perspective.
What is New?
In this multi-center, North American study of > 11,000 patients with out-of-hospital cardiac arrest, we describe the relationship between duration of cardiopulmonary resuscitation and the likelihood of survival with favorable neurological outcome.
There is a steady and predictable decline in the likelihood of survival with favorable neurological outcome as the duration of cardiopulmonary resuscitation accumulates. This decline is qualitatively different for subject with different features of out-of-hospital cardiac arrest (shockable initial cardiac rhythm, witnessed cardiac arrest, bystander CPR).
The duration of cardiopulmonary resuscitation independently predicts survival with favorable neurologic outcome adjusting for other factors.
What are the Clinical Implications?
Conventional resuscitation is most effective within the first 20 minutes, by which time 90% of patients with favorable neurological recovery had achieved return of spontaneous circulation.
Patients with shockable initial cardiac rhythms, witnessed cardiac arrest, and bystander CPR were more likely to survive with favorable neurologic outcome after resuscitation efforts > 20 minutes.
Hypothetical termination of resuscitation based solely on duration of cardiopulmonary resuscitation would have resulted in unacceptable losses of subjects with ultimate favorable neurologic outcome.
Acknowledgments
None
Dr. Reynolds had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Sources of Funding: There were no sources of direct funding for this secondary analysis of de-identified clinical trial data.
Rittenberger: Dr. Rittenberger is funded through the AHA Grant-in-Aid program. He has received travel reimbursement and honoraria from C. R. Bard.
Kurz: Dr. Kurz is funded by NIH 5U01DK096037 (Site PI); K23AG038548 (Site PI); R01GM101197 (Site PI); R01GM103799 (Site PI); American Heart Association; Society for Critical Care Medicine, Rapid Pathogen Screening Inc; Boehringer-Ingelheim; Abbott; to the University of Alabama at Birmingham. Dr. Kurz has also personally received honoraria from Zoll Medical Corporation.
Callaway: Dr. Callaway is funded by NHLBI HL077871 (PI); K12 HL109068 (Program Director) to University of Pittsburgh.
Footnotes
Disclosures: Reynolds: Nothing to disclose
Grunau: Nothing to disclose
Sawyer: Nothing to disclose
References
- 1.Writing Group Members; Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, Das SR, de Ferranti S, Després JP, Fullerton HJ, Howard VJ, Huffman MD, Isasi CR, Jiménez MC, Judd SE, Kissela BM, Lichtman JH, Lisabeth LD, Liu S, Mackey RH, Magid DJ, McGuire DK, Mohler ER, 3rd, Moy CS, Muntner P, Mussolino ME, Nasir K, Neumar RW, Nichol G, Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Rosamond W, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani SS, Woo D, Yeh RW, Turner MB American Heart Association Statistics Committee; Stroke Statistics Subcommittee. Heart Disease and Stroke Statistics-2016 Update: A Report From the American Heart Association. Circulation. 2016;133:e38–360. doi: 10.1161/CIR.0000000000000350. [DOI] [PubMed] [Google Scholar]
- 2.Daya MR, Schmicker RH, Zive DM, Rea TD, Nichol G, Buick JE, Brooks S, Christenson J, MacPhee R, Craig A, Rittenberger JC, Davis DP, May S, Wigginton J, Wang H Resuscitation Outcomes Consortium Investigators. Out-of-hospital cardiac arrest survival improving over time: Results from the Resuscitation Outcomes Consortium (ROC) Resuscitation. 2015;91:108–15. doi: 10.1016/j.resuscitation.2015.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Chan PS, McNally B, Tang F, Kellermann A CARES Surveillance Group. Recent trends in survival from out-of-hospital cardiac arrest in the United States. Circulation. 2014;130:1876–82. doi: 10.1161/CIRCULATIONAHA.114.009711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.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;128:2488–94. doi: 10.1161/CIRCULATIONAHA.113.002408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bailey ED, Wydro GC, Cone DC. Termination of resuscitation in the prehospital setting for adult patients suffering nontraumatic cardiac arrest. National Association of EMS Physicians Standards and Clinical Practice Committee. Prehosp Emerg Care. 2000;4:190–5. doi: 10.1080/10903120090941498. [DOI] [PubMed] [Google Scholar]
- 6.Millin MG, Khandker SR, Malki A. Termination of resuscitation of nontraumatic cardiopulmonary arrest: resource document for the National Association of EMS Physicians position statement. Prehosp Emerg Care. 2011;15:547–54. doi: 10.3109/10903127.2011.608872. Epub 2011 Aug 15. [DOI] [PubMed] [Google Scholar]
- 7.National Association of EMS Physicians. Termination of resuscitation in nontraumatic cardiopulmonary arrest. Prehosp Emerg Care. 2011;15:542. doi: 10.3109/10903127.2011.598621. [DOI] [PubMed] [Google Scholar]
- 8.Morrison LJ, Visentin LM, Kiss A, Theriault R, Eby D, Vermeulen M, Sherbino J, Verbeek PR TOR Investigators. Validation of a rule for termination of resuscitation in out-of-hospital cardiac arrest. N Engl J Med. 2006;355:478–87. doi: 10.1056/NEJMoa052620. [DOI] [PubMed] [Google Scholar]
- 9.Marsden AK, Ng GA, Dalziel K, Cobbe SM. When is it futile for ambulance personnel to initiate cardiopulmonary resuscitation? BMJ. 1995;311:49–51. doi: 10.1136/bmj.311.6996.49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Verbeek PR, Vermeulen MJ, Ali FH, Messenger DW, Summers J, Morrison LJ. Derivation of a termination-of-resuscitation guideline for emergency medical technicians using automated external defibrillators. Acad Emerg Med. 2002;9:671–8. doi: 10.1111/j.1553-2712.2002.tb02144.x. [DOI] [PubMed] [Google Scholar]
- 11.Goldberger ZD, Chan PS, Berg RA, Kronick SL, Cooke CR, Lu M, Banerjee M, Hayward RA, Krumholz HM, Nallamothu BK American Heart Association Get With The Guidelines—Resuscitation (formerly National Registry of Cardiopulmonary Resuscitation) Investigators. Duration of resuscitation efforts and survival after in-hospital cardiac arrest: an observational study. Lancet. 2012;380:1473–81. doi: 10.1016/S0140-6736(12)60862-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Nagao K, Nonogi H, Yonemoto N, Gaieski DF, Ito N, Takayama M, Shirai S, Furuya S, Tani S, Kimura T, Saku K Japanese Circulation Society with Resuscitation Science Study (JCS-ReSS) Group. Duration of Prehospital Resuscitation Efforts After Out-of-Hospital Cardiac Arrest. Circulation. 2016;133:1386–96. doi: 10.1161/CIRCULATIONAHA.115.018788. [DOI] [PubMed] [Google Scholar]
- 13.Aufderheide TP, Nichol G, Rea TD, Brown SP, Leroux BG, Pepe PE, Kudenchuk PJ, Christenson J, Daya MR, Dorian P, Callaway CW, Idris AH, Andrusiek D, Stephens SW, Hostler D, Davis DP, Dunford JV, Pirrallo RG, Stiell IG, Clement CM, Craig A, Van Ottingham L, Schmidt TA, Wang HE, Weisfeldt ML, Ornato JP, Sopko G Resuscitation Outcomes Consortium (ROC) Investigators. A trial of an impedance threshold device in out-of-hospital cardiac arrest. N Engl J Med. 2011;365:798–806. doi: 10.1056/NEJMoa1010821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Stiell IG, Nichol G, Leroux BG, Rea TD, Ornato JP, Powell J, Christenson J, Callaway CW, Kudenchuk PJ, Aufderheide TP, Idris AH, Daya MR, Wang HE, Morrison LJ, Davis D, Andrusiek D, Stephens S, Cheskes S, Schmicker RH, Fowler R, Vaillancourt C, Hostler D, Zive D, Pirrallo RG, Vilke GM, Sopko G, Weisfeldt M ROC Investigators. Early versus later rhythm analysis in patients with out-of-hospital cardiac arrest. N Engl J Med. 2011;365:787–97. doi: 10.1056/NEJMoa1010076. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Aufderheide TP, Kudenchuk PJ, Hedges JR, Nichol G, Kerber RE, Dorian P, Davis DP, Idris AH, Callaway CW, Emerson S, Stiell IG, Terndrup TE ROC Investigators. Resuscitation Outcomes Consortium (ROC) PRIMED cardiac arrest trialmethods part 1: rationale and methodology for the impedance threshold device (ITD) protocol. Resuscitation. 2008;78:179–85. doi: 10.1016/j.resuscitation.2008.01.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Stiell IG, Callaway C, Davis D, Terndrup T, Powell J, Cook A, Kudenchuk PJ, Daya M, Kerber R, Idris A, Morrison LJ, Aufderheide T ROC Investigators. Resuscitation Outcomes Consortium (ROC) PRIMED cardiac arrest trialmethods part 2: rationale and methodology for “Analyze Later vs. Analyze Early” protocol. Resuscitation. 2008;78:186–95. doi: 10.1016/j.resuscitation.2008.01.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zive D, Koprowicz K, Schmidt T, Stiell I, Sears G, Van Ottingham L, Idris A, Stephens S, Daya M Resuscitation Outcomes Consortium Investigators. Variation in out-of-hospital cardiac arrest resuscitation and transport practices in the Resuscitation Outcomes Consortium: ROC Epistry-Cardiac Arrest. Resuscitation. 2011;82:277–84. doi: 10.1016/j.resuscitation.2010.10.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wang HE, Min A, Hostler D, Chang CC, Callaway CW. Differential effects of out-of-hospital interventions on short- and long-term survival after cardiopulmonary arrest. Resuscitation. 2005;67:69–74. doi: 10.1016/j.resuscitation.2005.05.008. [DOI] [PubMed] [Google Scholar]
- 19.Bernard SA, Gray TW, Buist MD, Jones BM, Silvester W, Gutteridge G, Smith K. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med. 2002;346:557–63. doi: 10.1056/NEJMoa003289. [DOI] [PubMed] [Google Scholar]
- 20.Hypothermia after Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med. 2002;346:549–56. doi: 10.1056/NEJMoa012689. Erratum in: N Engl J Med 2002; 346:1756. [DOI] [PubMed] [Google Scholar]
- 21.Reynolds JC, Rittenberger JC, Toma C, Callaway CW Post Cardiac Arrest Service. Risk-adjusted outcome prediction with initial post-cardiac arrest illness severity: implications for cardiac arrest survivors being considered for early invasive strategy. Resuscitation. 2014;85:1232–9. doi: 10.1016/j.resuscitation.2014.05.037. [DOI] [PubMed] [Google Scholar]
- 22.Reynolds JC, Callaway CW, El Khoudary SR, Moore CG, Alvarez RJ, Rittenberger JC. Coronary angiography predicts improved outcome following cardiac arrest: propensity-adjusted analysis. J Intensive Care Med. 2009;24:179–86. doi: 10.1177/0885066609332725. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Grunau B, Reynolds JC, Scheuermeyer FX, Stenstrom R, Pennington S, Cheung C, Li J, Habibi M, Ramanathan K, Barbic D, Christenson J. Comparing the prognosis of those with initial shockable and non-shockable rhythms with increasing durations of CPR: Informing minimum durations of resuscitation. Resuscitation. 2016;101:50–6. doi: 10.1016/j.resuscitation.2016.01.021. [DOI] [PubMed] [Google Scholar]
- 24.Nehme Z, Andrew E, Bernard S, Smith K. Impact of cardiopulmonary resuscitation duration on survival from paramedic witnessed out-of-hospital cardiac arrests: An observational study. Resuscitation. 2016;100:25–31. doi: 10.1016/j.resuscitation.2015.12.011. [DOI] [PubMed] [Google Scholar]
- 25.Helft PR, Siegler M, Lantos J. The rise and fall of the futility movement. N Engl J Med. 2000;343:293–296. doi: 10.1056/NEJM200007273430411. [DOI] [PubMed] [Google Scholar]
- 26.Schneiderman LJ, Jecker NS, Jonsen AR. Medical futility: Response to critiques. Ann Intern Med. 1996;125:669–674. doi: 10.7326/0003-4819-125-8-199610150-00007. [DOI] [PubMed] [Google Scholar]
- 27.Drennan IR, Case E, Reynolds JC, Goldberger ZD, Herren H, Austin MA, Jasti J, Schmicker R, Idris A, Chareston M, Leslie PR, Verbeek PR, Morrison LJ. An observational analysis of termination of resuscitation in the out-of-hospital setting. Circulation. 2014;130:A54. [Google Scholar]
- 28.Morrison LJ, Verbeek PR, Zhan C, Kiss A, Allan KS. Validation of a universal prehospital termination of resuscitation clinical prediction rule for advanced and basic life support providers. Resuscitation. 2009;80:324–8. doi: 10.1016/j.resuscitation.2008.11.014. [DOI] [PubMed] [Google Scholar]
- 29.Sasson C, Hegg AJ, Macy M, Park A, Kellermann A, McNally B CARES Surveillance Group. Prehospital termination of resuscitation in cases of refractory out-of-hospital cardiac arrest. JAMA. 2008;300:1432–8. doi: 10.1001/jama.300.12.1432. [DOI] [PubMed] [Google Scholar]
- 30.Morimura N, Sakamoto T, Nagao K, Asai Y, Yokota H, Tahara Y, Atsumi T, Nara S, Hase M. Extracorporeal cardiopulmonary resuscitation for out-of-hospital cardiac arrest: A review of the Japanese literature. Resuscitation. 2011;82:10–4. doi: 10.1016/j.resuscitation.2010.08.032. [DOI] [PubMed] [Google Scholar]
- 31.Stub D, Bernard S, Pellegrino V, Smith K, Walker T, Sheldrake J, Hockings L, Shaw J, Duffy SJ, Burrell A, Cameron P, Smit de V, Kaye DM. Refractory cardiac arrest treated with mechanical CPR, hypothermia, ECMO and early reperfusion (the CHEER trial) Resuscitation. 2015;86:88–94. doi: 10.1016/j.resuscitation.2014.09.010. [DOI] [PubMed] [Google Scholar]
- 32.Sakamoto T, Morimura N, Nagao K, Asai Y, Yokota H, Nara S, Hase M, Tahara Y, Atsumi T SAVE-J Study Group. Extracorporeal cardiopulmonary resuscitation versus conventional cardiopulmonary resuscitation in adults with out-of-hospital cardiac arrest: a prospective observational study. Resuscitation. 2014;85:762–8. doi: 10.1016/j.resuscitation.2014.01.031. [DOI] [PubMed] [Google Scholar]
- 33.Kim SJ, Kim HJ, Lee HY, Ahn HS, Lee SW. Comparing extracorporeal cardiopulmonary resuscitation with conventional cardiopulmonary resuscitation: A meta-analysis. Resuscitation. 2016;103:106–16. doi: 10.1016/j.resuscitation.2016.01.019. [DOI] [PubMed] [Google Scholar]
- 34.Grunau B, Reynolds J, Scheuermeyer F, Stenstom R, Stub D, Pennington S, Cheskes S, Ramanathan K, Christenson J. Relationship between Time-to-ROSC and Survival in Out-of-hospital Cardiac Arrest ECPR Candidates: When is the Best Time to Consider Transport to Hospital? Prehosp Emerg Care. 2016;20:615–22. doi: 10.3109/10903127.2016.1149652. [DOI] [PubMed] [Google Scholar]
- 35.Japan Resuscitation Council and Japanese Foundation for Emergency Medicine, editor. Japan Resuscitation Council. 2010 Japan Resuscitation Council guidelines for resuscitation. 1st. Tokyo: Health Shuppansha; 2010. pp. 1–446. In Japanese. [Google Scholar]
- 36.Nichol G, Thomas E, Callaway CW, Hedges J, Powell JL, Aufderheide TP, Rea T, Lowe R, Brown T, Dreyer J, Davis D, Idris A, Stiell I Resuscitation Outcomes Consortium Investigators. Regional variation in out-of-hospital cardiac arrest incidence and outcome. JAMA. 2008;300:1423–31. doi: 10.1001/jama.300.12.1423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Frisch A, Reynolds JC, Condle J, Gruen D, Callaway CW. Documentation discrepancies of time-dependent critical events in out of hospital cardiac arrest. Resuscitation. 2014;85:1111–4. doi: 10.1016/j.resuscitation.2014.05.002. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.