https://orcid.org/0000-0002-1179-3718
Abstract
Extracorporeal cardiopulmonary resuscitation (ECPR) is considered for potentially reversible out-of-hospital cardiac arrest (OHCA). However, the association between time to ECPR and outcome has not been well established.
Between June 2014 and December 2017, we enrolled 34 754 OHCA patients in a multicentre, prospective fashion [Japanese Association for Acute Medicine (JAAM)-OHCA registry]. After the application of exclusion criteria, 695 OHCA patients who underwent ECPR for cardiac causes were eligible for this study. We investigated the association between the call-to-ECPR interval and favourable neurological outcome (cerebral performance category 1 or 2) at 30 days. Seventy-seven patients (11%) had a favourable neurological outcome at 30 days. The call-to-ECPR intervals in these patients were significantly shorter than in those with an unfavourable neurological outcome [49 (41–58) vs. 58 (48–68) min, respectively, P < 0.001]. A longer call-to-ECPR interval was associated with a smaller proportion of patients undergoing percutaneous coronary intervention (PCI) (P = 0.034) or target temperature management (TTM) (P < 0.001). Stepwise multivariable logistic regression analysis revealed that the call-to-ECPR interval was an independent predictor of favourable neurological outcome [odds ratio (OR) 0.96, 95% confidence interval (CI) 0.94–0.99, P = 0.001], as were age, male gender, initial shockable rhythm, transient return of spontaneous circulation in the prehospital setting, arterial pH at hospital arrival, PCI (OR 2.30, 95% CI 1.14–4.66, P = 0.019), and TTM (OR 2.28, 95% CI 1.13–4.62, P = 0.019).
A shorter call-to-ECPR interval and implementation of PCI and TTM predicted a favourable neurological outcome at 30 days in OHCA patients who underwent ECPR for cardiac causes.
Introduction
Patients with out-of-hospital cardiac arrest (OHCA) still have a poor prognosis in industrialized countries.1,2 Early recognition of cardiac arrest and bystander-initiated cardiopulmonary resuscitation (CPR) are critical for improving survival in OHCA patients.3,4 However, prolonged conventional CPR leads to suboptimal circulation in end-organs if OHCA patients fail to obtain return of spontaneous circulation (ROSC).5 Extracorporeal cardiopulmonary resuscitation (ECPR) using veno-arterial extracorporeal membrane oxygenation (VA-ECMO) may achieve effective circulatory and respiratory support if the cause of cardiac arrest is reversible, thus improving survival and neurological outcome.6–9 Moreover, initiating ECPR early, rather than after complete failure of conventional CPR, correlated with superior clinical outcomes.10,11
For post-cardiac arrest patients, current guidelines recommend target temperature management (TTM) to improve neurological outcome.1,2,12 Moreover, emergent coronary angiography (CAG) and subsequent percutaneous coronary intervention (PCI) are recommended in patients with ST elevation after cardiac arrest and in those without post-cardiac arrest ST elevation who demonstrate haemodynamic or electrical instability.1,2 If OHCA patients fail to achieve sustained ROSC with conventional CPR, VA-ECMO initiation should sometimes precede emergent PCI and TTM. However, the appropriate timing of ECPR and the usefulness of TTM and PCI in OHCA patients who undergo ECPR is not well established. This study examined the association between the time to ECPR and clinical outcome.
The Japanese Association for Acute Medicine (JAAM)-OHCA Registry is a multicentre, prospective registry that enrolls OHCA patients transported to critical care medical centres or hospitals with an emergency care department in Japan.13 Using this nationwide registry, which aims to improve survival after OHCA by providing an evidence-based therapeutic strategy and an emergency medical system, we focused on the association between time to ECPR and favourable neurological outcome at 30 days.
Methods
Data collection and study design
The study methodology and details of the emergency medical services (EMS) system were previously described.13 In brief, between June 2014 and December 2017, the JAAM-OHCA registry enrolled OHCA patients for whom resuscitation was attempted and who were transported to 87 participating institutions. In-hospital data from the JAAM-OHCA registry were systemically merged with Utstein-style prehospital data from the All-Japan Utstein Registry of the Fire and Disaster Management Agency of Japan based on Ustein-style international guidelines for reporting OHCA.14 This registry was approved by the Ethics Committee of Kyoto University (R1045) and the National Cerebral and Cardiovascular Center (R20046), and each hospital was also approved by the JAAM-OHCA registry protocol as necessary.
The inclusion criteria were as follows: adult patients with OHCA due to cardiac causes who were transported to participating hospitals; cardiac arrest recognized on hospital arrival; ROSC not achieved after hospital arrival; and ECPR performed (Figure 1). The exclusion criteria were as follows: age under 18 years; OHCA of non-cardiac cause; ROSC present at hospital arrival; ROSC achieved between hospital arrival and ECPR initiation; and ECMO not performed.
Flow chart of study patients. ECMO, extracorporeal cardiopulmonary oxygenation; ECPR, extracorporeal cardiopulmonary resuscitation; ROSC, return of spontaneous circulation.
After ECPR initiation, decisions to perform CAG, PCI, and TTM were made by physicians. Cardiac causes were classified into acute coronary syndrome (ACS), other heart disease, and presumed cardiac conditions diagnosed as the cause of cardiac arrest. ACS was diagnosed as the cause of cardiac arrest based on previous cardiac arrest symptoms, electrocardiography, laboratory data, and CAG. A favourable neurological outcome was defined as cerebral performance category (CPC) 1 or 2.
Data analysis
We divided patients into those with or without a favourable neurological outcome at 30 days, then compared the implementation rates of PCI and TTM in terms of both the ECPR time frame and the proportion of individuals with a favourable neurological outcome. The call-to-ECPR interval was defined as the interval from emergency call receipt to initiation of circulatory support by VA-ECMO, and the call-to-PCI interval was defined as the interval from emergency call receipt to successful coronary reperfusion.
Statistical analysis
Statistical analyses were conducted using JMP16.0 (SAS Institute Japan, Tokyo, Japan). A P-value <0.05 indicated significance. Data are expressed as median [interquartile range (IQR)]. Intergroup comparisons of continuous variables were performed with the Wilcoxon rank sum test. Nominal variables were compared using the χ2 test or Fisher’s exact test. Differences between the proportions of patients who underwent PCI and TTM according to the call-to-ECPR interval were analysed using the Cochran–Armitage test for trend.
To examine whether each variable predicted a favourable neurological outcome at 30 days, univariable and multivariable logistic regression models were constructed using age, gender, witnessed, bystander CPR, and ventricular fibrillation or pulseless ventricular tachycardia (VF/pVT) as the first documented rhythm at EMS arrival, transient ROSC in prehospital setting, arterial pH at hospital arrival, arterial lactate at hospital arrival, call-to-ECPR interval, ACS diagnosed as a cause of cardiac arrest, PCI, and TTM. Stepwise multivariable logistic regression analysis with a P-value of 0.1 for backward elimination was performed to select the best predictive model.
Results
Study participants
During the study period, 34 754 OHCA patients were registered. Of the 33 921 patients who underwent resuscitation by a physician after hospital arrival, 30 856 (91.0%) were matched with prehospital data. After applying exclusion criteria, 695 OHCA patients who underwent ECPR for cardiac causes were eligible (Figure 1). Seventy-seven (11%) patients had a favourable neurological outcome and 618 (89%) did not. The number of patients with each CPC score at 30 days was as follows: CPC 1, 54 (8%); CPC 2, 23 (3%); CPC 3, 29 (4%); CPC 4, 45 (6%); and CPC 5, 544 (78%).
Clinical characteristics and time intervals in patients with and without a favourable neurological outcome
Table 1 shows comparisons of clinical characteristics and time management in patients with and without a favourable neurological outcome. The proportions of patients with VF/pVT as the first documented rhythm at EMS arrival and those who obtained transient ROSC in the prehospital setting were significantly higher among patients with a favourable neurological outcome, while these patients were significantly younger and significantly less likely to receive intravenous fluid and adrenaline in the prehospital setting. In the in-hospital setting, the proportions of patients with VF/pVT as the first documented rhythm at hospital arrival, those with ACS diagnosed as a cause of cardiac arrest, and those who underwent CAG, PCI, intra-aortic balloon pump (IABP) implantation, TTM, and 12-lead electrocardiogram after ROSC were significantly higher in patients with a favourable neurological outcome, while these patients had a significantly lower arterial PaCO2 at hospital arrival. In patients with a favourable neurological outcome, the median call-to-field arrival interval (6 min, IQR 5–7 min), call-to-hospital arrival interval (28 min, IQR 23–35 min), call-to-ECPR interval (49 min, IQR 41–58 min), and call-to-PCI interval (108 min, IQR 83–141 min) were all significantly shorter than the corresponding values in patients with an unfavourable neurological outcome (7 min, IQR 5–9 min; 32 min, IQR 26–40 min; 58 min, IQR 48–68 min; 131 min, IQR 96–179 min, respectively, P < 0.05 for all). The median hospital arrival-to-ECPR interval (23 min, IQR 17–32 min) and hospital arrival-to-PCI interval (84 min; IQR 60–114 min) were significantly shorter in patients with a favourable neurological outcome than in those without (29 min, IQR 21–39 min; 106 min; IQR 65–150 min, respectively, P < 0.05). In contrast, there were no significant differences between the two groups regarding the intervals of call-to-CPR initiation by EMS personnel or call-to-defibrillation by EMS personnel. Figure 2 shows the distribution of patients according to the call-to-ECPR interval between the two groups. Comparisons of clinical characteristics and time intervals in patients with/without survival at 30 days, those with ACS/non-ACS, and those who did or did not undergo PCI are shown in the Supplementary material online, Tables S1–S3.
Distribution of call-to-extracorporeal cardiopulmonary resuscitation intervals in patients with (left) and without (right) a favourable neurological outcome. Bar graphs show the distribution of the call-to-extracorporeal cardiopulmonary resuscitation interval in the overall sample. Numbers indicate time intervals as median (interquartile range). CPC, cerebral performance category.
Baseline characteristics, prehospital and in-hospital data, and time intervals in patients with and without favourable neurological outcome at 30 days
| Overall | Favourable neurological outcome | Unfavourable neurological outcome | Missing | P-value | |
|---|---|---|---|---|---|
| Number | 695 | 77 | 618 | ||
| Basic information | |||||
| Age, years | 59 (47–68) | 51 (45–66) | 60 (48–68) | 0 | <0.001 |
| Male | 585 (84) | 61 (79) | 524 (85) | 0 | 0.207 |
| Prehospital data | |||||
| Witnessed | 542 (78) | 65 (84) | 477 (77) | 0 | 0.149 |
| Bystander-initiated CPR | 344 (49) | 42 (55) | 302 (49) | 0 | 0.347 |
| Defibrillation with a public-access AED | 53 (8) | 10 (13) | 43 (7) | 0 | 0.060 |
| Dispatcher instructions | 299 (43) | 32 (42) | 267 (43) | 15 | 0.932 |
| VF/pulseless VT as the first documented rhythm at EMS arrival | 503 (72) | 63 (82) | 440 (71) | 39 | 0.004 |
| Defibrillation by EMS personnel | 582 (84) | 70 (91) | 512 (83) | 24 | 0.289 |
| Departure of physician-staffed ambulance or helicopter | 151 (22) | 22 (29) | 129 (21) | 0 | 0.123 |
| Advanced airway management | 377 (54) | 33 (43) | 344 (56) | 121 | 0.263 |
| Intravenous fluid | 322 (46) | 27 (35) | 295 (48) | 1 | 0.034 |
| Adrenaline administration | 287 (41) | 20 (26) | 267 (43) | 0 | 0.004 |
| Prehospital transient ROSC | 51 (7) | 12 (16) | 39 (6) | 0 | 0.003 |
| In-hospital data | |||||
| VF/pulseless VT as the first documented rhythm at hospital arrival | 394 (57) | 62 (81) | 332 (54) | 4 | <0.001 |
| Body temperature at hospital arrival, °C | 35.3 (34.3–36.0) | 35.1 (34.4–35.8) | 35.4 (34.3–36.0) | 231 | 0.152 |
| Defibrillation after hospital arrival | 515 (74) | 63 (82) | 452 (73) | 0 | 0.101 |
| Tracheal intubation after hospital arrival | 606 (87) | 69 (90) | 537 (87) | 0 | 0.466 |
| Implementation of 12-lead ECG after ROSC | 471 (68) | 71 (92) | 400 (65) | 0 | <0.001 |
| ST elevation on 12-lead ECG after ROSC (n = 471) | 231 (49) | 41 (58) | 190 (48) | 0 | 0.112 |
| Arterial blood gases at hospital arrival | |||||
| pH | 6.92 (6.82–7.02) | 6.96 (6.82–7.08) | 6.92 (6.81–7.01) | 122 | 0.112 |
| PaCO2, mmHg | 71.2 (53.4–88.0) | 57.2 (42.4–76.1) | 72.4 (54.8–90.2) | 121 | <0.001 |
| PaO2, mmHg | 64.3 (26.1–143.0) | 89.6 (48.5–176.5) | 61.8 (25.4–139.1) | 128 | 0.084 |
| , mmol/L | 13.8 (10.4–17.5) | 11.9 (9.8–16.3) | 14.0 (10.5–17.6) | 127 | 0.050 |
| Lactate, mg/dL | 118 (88–145) | 118 (79–159) | 119 (89–145) | 131 | 0.980 |
| Coronary angiography | 500 (72) | 76 (99) | 424 (69) | 0 | <0.001 |
| Percutaneous coronary intervention | 285 (41) | 47 (61) | 238 (39) | 0 | <0.001 |
| Successful coronary reperfusion (n = 285) | 262 (38) | 45 (96) | 217 (91) | 9 | 0.136 |
| Intra-aortic balloon pump implantation | 413 (59) | 67 (87) | 346 (56) | 0 | <0.001 |
| Target temperature management | 307 (44) | 51 (66) | 256 (41) | 6 | <0.001 |
| Targeted hypothermia | 246 | 44 | 202 | ||
| Targeted normothermia | 61 | 7 | 54 | ||
| Drug administration during cardiac arrest | |||||
| Adrenaline | 655 (94) | 71 (92) | 584 (94) | 4 | 0.275 |
| Amiodarone | 386 (56) | 48 (62) | 338 (55) | 5 | 0.230 |
| Nifekalant | 42 (6) | 6 (8) | 36 (6) | 10 | 0.456 |
| Lidocaine | 25 (4) | 4 (5) | 21 (3) | 6 | 0.511 |
| Atropine | 16 (2) | 1 (1) | 15 (2) | 10 | 1.000 |
| Magnesium | 89 (13) | 17 (22) | 72 (12) | 10 | 0.012 |
| Vasopressin | 7 (1) | 1 (1) | 6 (1) | 8 | 0.567 |
| Aetiology of cardiac cause | 0 | <0.001 | |||
| Acute coronary syndrome | 360 (52) | 48 (62) | 312 (50) | ||
| Other heart disease | 179 (26) | 25 (32) | 154 (25) | ||
| Presumed cardiac cause | 156 (22) | 4 (5) | 152 (25) | ||
| Time course | |||||
| Call-to-field arrival interval, min | 7 (5–9) | 6 (5–7) | 7 (5–9) | 1 | <0.001 |
| Call-to-CPR initiation by EMS personnel interval, min | 9 (7–11) | 8 (6–10) | 9 (7–11) | 4 | 0.077 |
| Call-to-defibrillation by EMS personnel interval, min (n = 582) | 10 (9–13) | 10 (8–12) | 10 (9–13) | 1 | 0.167 |
| Call-to-hospital arrival interval, min | 31 (26–40) | 28 (23–35) | 32 (26–40) | 14 | 0.004 |
| Call-to-ECPR interval, min | 56 (47–68) | 49 (41–58) | 58 (48–68) | 2 | <0.001 |
| Call-to-PCI interval, min (n = 262) | 126 (94–171) | 108 (83–141) | 131 (96–179) | 12 | 0.008 |
| Hospital arrival-to-ECPR interval, min | 29 (20–38) | 23 (17–32) | 29 (21–39) | 2 | <0.001 |
| Hospital arrival-to-PCI interval, min (n = 262) | 98 (64–143) | 84 (60–114) | 106 (65–150) | 12 | 0.017 |
| Overall | Favourable neurological outcome | Unfavourable neurological outcome | Missing | P-value | |
|---|---|---|---|---|---|
| Number | 695 | 77 | 618 | ||
| Basic information | |||||
| Age, years | 59 (47–68) | 51 (45–66) | 60 (48–68) | 0 | <0.001 |
| Male | 585 (84) | 61 (79) | 524 (85) | 0 | 0.207 |
| Prehospital data | |||||
| Witnessed | 542 (78) | 65 (84) | 477 (77) | 0 | 0.149 |
| Bystander-initiated CPR | 344 (49) | 42 (55) | 302 (49) | 0 | 0.347 |
| Defibrillation with a public-access AED | 53 (8) | 10 (13) | 43 (7) | 0 | 0.060 |
| Dispatcher instructions | 299 (43) | 32 (42) | 267 (43) | 15 | 0.932 |
| VF/pulseless VT as the first documented rhythm at EMS arrival | 503 (72) | 63 (82) | 440 (71) | 39 | 0.004 |
| Defibrillation by EMS personnel | 582 (84) | 70 (91) | 512 (83) | 24 | 0.289 |
| Departure of physician-staffed ambulance or helicopter | 151 (22) | 22 (29) | 129 (21) | 0 | 0.123 |
| Advanced airway management | 377 (54) | 33 (43) | 344 (56) | 121 | 0.263 |
| Intravenous fluid | 322 (46) | 27 (35) | 295 (48) | 1 | 0.034 |
| Adrenaline administration | 287 (41) | 20 (26) | 267 (43) | 0 | 0.004 |
| Prehospital transient ROSC | 51 (7) | 12 (16) | 39 (6) | 0 | 0.003 |
| In-hospital data | |||||
| VF/pulseless VT as the first documented rhythm at hospital arrival | 394 (57) | 62 (81) | 332 (54) | 4 | <0.001 |
| Body temperature at hospital arrival, °C | 35.3 (34.3–36.0) | 35.1 (34.4–35.8) | 35.4 (34.3–36.0) | 231 | 0.152 |
| Defibrillation after hospital arrival | 515 (74) | 63 (82) | 452 (73) | 0 | 0.101 |
| Tracheal intubation after hospital arrival | 606 (87) | 69 (90) | 537 (87) | 0 | 0.466 |
| Implementation of 12-lead ECG after ROSC | 471 (68) | 71 (92) | 400 (65) | 0 | <0.001 |
| ST elevation on 12-lead ECG after ROSC (n = 471) | 231 (49) | 41 (58) | 190 (48) | 0 | 0.112 |
| Arterial blood gases at hospital arrival | |||||
| pH | 6.92 (6.82–7.02) | 6.96 (6.82–7.08) | 6.92 (6.81–7.01) | 122 | 0.112 |
| PaCO2, mmHg | 71.2 (53.4–88.0) | 57.2 (42.4–76.1) | 72.4 (54.8–90.2) | 121 | <0.001 |
| PaO2, mmHg | 64.3 (26.1–143.0) | 89.6 (48.5–176.5) | 61.8 (25.4–139.1) | 128 | 0.084 |
| , mmol/L | 13.8 (10.4–17.5) | 11.9 (9.8–16.3) | 14.0 (10.5–17.6) | 127 | 0.050 |
| Lactate, mg/dL | 118 (88–145) | 118 (79–159) | 119 (89–145) | 131 | 0.980 |
| Coronary angiography | 500 (72) | 76 (99) | 424 (69) | 0 | <0.001 |
| Percutaneous coronary intervention | 285 (41) | 47 (61) | 238 (39) | 0 | <0.001 |
| Successful coronary reperfusion (n = 285) | 262 (38) | 45 (96) | 217 (91) | 9 | 0.136 |
| Intra-aortic balloon pump implantation | 413 (59) | 67 (87) | 346 (56) | 0 | <0.001 |
| Target temperature management | 307 (44) | 51 (66) | 256 (41) | 6 | <0.001 |
| Targeted hypothermia | 246 | 44 | 202 | ||
| Targeted normothermia | 61 | 7 | 54 | ||
| Drug administration during cardiac arrest | |||||
| Adrenaline | 655 (94) | 71 (92) | 584 (94) | 4 | 0.275 |
| Amiodarone | 386 (56) | 48 (62) | 338 (55) | 5 | 0.230 |
| Nifekalant | 42 (6) | 6 (8) | 36 (6) | 10 | 0.456 |
| Lidocaine | 25 (4) | 4 (5) | 21 (3) | 6 | 0.511 |
| Atropine | 16 (2) | 1 (1) | 15 (2) | 10 | 1.000 |
| Magnesium | 89 (13) | 17 (22) | 72 (12) | 10 | 0.012 |
| Vasopressin | 7 (1) | 1 (1) | 6 (1) | 8 | 0.567 |
| Aetiology of cardiac cause | 0 | <0.001 | |||
| Acute coronary syndrome | 360 (52) | 48 (62) | 312 (50) | ||
| Other heart disease | 179 (26) | 25 (32) | 154 (25) | ||
| Presumed cardiac cause | 156 (22) | 4 (5) | 152 (25) | ||
| Time course | |||||
| Call-to-field arrival interval, min | 7 (5–9) | 6 (5–7) | 7 (5–9) | 1 | <0.001 |
| Call-to-CPR initiation by EMS personnel interval, min | 9 (7–11) | 8 (6–10) | 9 (7–11) | 4 | 0.077 |
| Call-to-defibrillation by EMS personnel interval, min (n = 582) | 10 (9–13) | 10 (8–12) | 10 (9–13) | 1 | 0.167 |
| Call-to-hospital arrival interval, min | 31 (26–40) | 28 (23–35) | 32 (26–40) | 14 | 0.004 |
| Call-to-ECPR interval, min | 56 (47–68) | 49 (41–58) | 58 (48–68) | 2 | <0.001 |
| Call-to-PCI interval, min (n = 262) | 126 (94–171) | 108 (83–141) | 131 (96–179) | 12 | 0.008 |
| Hospital arrival-to-ECPR interval, min | 29 (20–38) | 23 (17–32) | 29 (21–39) | 2 | <0.001 |
| Hospital arrival-to-PCI interval, min (n = 262) | 98 (64–143) | 84 (60–114) | 106 (65–150) | 12 | 0.017 |
Results are shown as median (IQR) or number (%).
AED, automated external defibrillator; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; EMS, emergency medical service; PCI, percutaneous coronary intervention; ROSC, return of spontaneous circulation; VF, ventricular fibrillation; VT, ventricular tachycardia.
Baseline characteristics, prehospital and in-hospital data, and time intervals in patients with and without favourable neurological outcome at 30 days
| Overall | Favourable neurological outcome | Unfavourable neurological outcome | Missing | P-value | |
|---|---|---|---|---|---|
| Number | 695 | 77 | 618 | ||
| Basic information | |||||
| Age, years | 59 (47–68) | 51 (45–66) | 60 (48–68) | 0 | <0.001 |
| Male | 585 (84) | 61 (79) | 524 (85) | 0 | 0.207 |
| Prehospital data | |||||
| Witnessed | 542 (78) | 65 (84) | 477 (77) | 0 | 0.149 |
| Bystander-initiated CPR | 344 (49) | 42 (55) | 302 (49) | 0 | 0.347 |
| Defibrillation with a public-access AED | 53 (8) | 10 (13) | 43 (7) | 0 | 0.060 |
| Dispatcher instructions | 299 (43) | 32 (42) | 267 (43) | 15 | 0.932 |
| VF/pulseless VT as the first documented rhythm at EMS arrival | 503 (72) | 63 (82) | 440 (71) | 39 | 0.004 |
| Defibrillation by EMS personnel | 582 (84) | 70 (91) | 512 (83) | 24 | 0.289 |
| Departure of physician-staffed ambulance or helicopter | 151 (22) | 22 (29) | 129 (21) | 0 | 0.123 |
| Advanced airway management | 377 (54) | 33 (43) | 344 (56) | 121 | 0.263 |
| Intravenous fluid | 322 (46) | 27 (35) | 295 (48) | 1 | 0.034 |
| Adrenaline administration | 287 (41) | 20 (26) | 267 (43) | 0 | 0.004 |
| Prehospital transient ROSC | 51 (7) | 12 (16) | 39 (6) | 0 | 0.003 |
| In-hospital data | |||||
| VF/pulseless VT as the first documented rhythm at hospital arrival | 394 (57) | 62 (81) | 332 (54) | 4 | <0.001 |
| Body temperature at hospital arrival, °C | 35.3 (34.3–36.0) | 35.1 (34.4–35.8) | 35.4 (34.3–36.0) | 231 | 0.152 |
| Defibrillation after hospital arrival | 515 (74) | 63 (82) | 452 (73) | 0 | 0.101 |
| Tracheal intubation after hospital arrival | 606 (87) | 69 (90) | 537 (87) | 0 | 0.466 |
| Implementation of 12-lead ECG after ROSC | 471 (68) | 71 (92) | 400 (65) | 0 | <0.001 |
| ST elevation on 12-lead ECG after ROSC (n = 471) | 231 (49) | 41 (58) | 190 (48) | 0 | 0.112 |
| Arterial blood gases at hospital arrival | |||||
| pH | 6.92 (6.82–7.02) | 6.96 (6.82–7.08) | 6.92 (6.81–7.01) | 122 | 0.112 |
| PaCO2, mmHg | 71.2 (53.4–88.0) | 57.2 (42.4–76.1) | 72.4 (54.8–90.2) | 121 | <0.001 |
| PaO2, mmHg | 64.3 (26.1–143.0) | 89.6 (48.5–176.5) | 61.8 (25.4–139.1) | 128 | 0.084 |
| , mmol/L | 13.8 (10.4–17.5) | 11.9 (9.8–16.3) | 14.0 (10.5–17.6) | 127 | 0.050 |
| Lactate, mg/dL | 118 (88–145) | 118 (79–159) | 119 (89–145) | 131 | 0.980 |
| Coronary angiography | 500 (72) | 76 (99) | 424 (69) | 0 | <0.001 |
| Percutaneous coronary intervention | 285 (41) | 47 (61) | 238 (39) | 0 | <0.001 |
| Successful coronary reperfusion (n = 285) | 262 (38) | 45 (96) | 217 (91) | 9 | 0.136 |
| Intra-aortic balloon pump implantation | 413 (59) | 67 (87) | 346 (56) | 0 | <0.001 |
| Target temperature management | 307 (44) | 51 (66) | 256 (41) | 6 | <0.001 |
| Targeted hypothermia | 246 | 44 | 202 | ||
| Targeted normothermia | 61 | 7 | 54 | ||
| Drug administration during cardiac arrest | |||||
| Adrenaline | 655 (94) | 71 (92) | 584 (94) | 4 | 0.275 |
| Amiodarone | 386 (56) | 48 (62) | 338 (55) | 5 | 0.230 |
| Nifekalant | 42 (6) | 6 (8) | 36 (6) | 10 | 0.456 |
| Lidocaine | 25 (4) | 4 (5) | 21 (3) | 6 | 0.511 |
| Atropine | 16 (2) | 1 (1) | 15 (2) | 10 | 1.000 |
| Magnesium | 89 (13) | 17 (22) | 72 (12) | 10 | 0.012 |
| Vasopressin | 7 (1) | 1 (1) | 6 (1) | 8 | 0.567 |
| Aetiology of cardiac cause | 0 | <0.001 | |||
| Acute coronary syndrome | 360 (52) | 48 (62) | 312 (50) | ||
| Other heart disease | 179 (26) | 25 (32) | 154 (25) | ||
| Presumed cardiac cause | 156 (22) | 4 (5) | 152 (25) | ||
| Time course | |||||
| Call-to-field arrival interval, min | 7 (5–9) | 6 (5–7) | 7 (5–9) | 1 | <0.001 |
| Call-to-CPR initiation by EMS personnel interval, min | 9 (7–11) | 8 (6–10) | 9 (7–11) | 4 | 0.077 |
| Call-to-defibrillation by EMS personnel interval, min (n = 582) | 10 (9–13) | 10 (8–12) | 10 (9–13) | 1 | 0.167 |
| Call-to-hospital arrival interval, min | 31 (26–40) | 28 (23–35) | 32 (26–40) | 14 | 0.004 |
| Call-to-ECPR interval, min | 56 (47–68) | 49 (41–58) | 58 (48–68) | 2 | <0.001 |
| Call-to-PCI interval, min (n = 262) | 126 (94–171) | 108 (83–141) | 131 (96–179) | 12 | 0.008 |
| Hospital arrival-to-ECPR interval, min | 29 (20–38) | 23 (17–32) | 29 (21–39) | 2 | <0.001 |
| Hospital arrival-to-PCI interval, min (n = 262) | 98 (64–143) | 84 (60–114) | 106 (65–150) | 12 | 0.017 |
| Overall | Favourable neurological outcome | Unfavourable neurological outcome | Missing | P-value | |
|---|---|---|---|---|---|
| Number | 695 | 77 | 618 | ||
| Basic information | |||||
| Age, years | 59 (47–68) | 51 (45–66) | 60 (48–68) | 0 | <0.001 |
| Male | 585 (84) | 61 (79) | 524 (85) | 0 | 0.207 |
| Prehospital data | |||||
| Witnessed | 542 (78) | 65 (84) | 477 (77) | 0 | 0.149 |
| Bystander-initiated CPR | 344 (49) | 42 (55) | 302 (49) | 0 | 0.347 |
| Defibrillation with a public-access AED | 53 (8) | 10 (13) | 43 (7) | 0 | 0.060 |
| Dispatcher instructions | 299 (43) | 32 (42) | 267 (43) | 15 | 0.932 |
| VF/pulseless VT as the first documented rhythm at EMS arrival | 503 (72) | 63 (82) | 440 (71) | 39 | 0.004 |
| Defibrillation by EMS personnel | 582 (84) | 70 (91) | 512 (83) | 24 | 0.289 |
| Departure of physician-staffed ambulance or helicopter | 151 (22) | 22 (29) | 129 (21) | 0 | 0.123 |
| Advanced airway management | 377 (54) | 33 (43) | 344 (56) | 121 | 0.263 |
| Intravenous fluid | 322 (46) | 27 (35) | 295 (48) | 1 | 0.034 |
| Adrenaline administration | 287 (41) | 20 (26) | 267 (43) | 0 | 0.004 |
| Prehospital transient ROSC | 51 (7) | 12 (16) | 39 (6) | 0 | 0.003 |
| In-hospital data | |||||
| VF/pulseless VT as the first documented rhythm at hospital arrival | 394 (57) | 62 (81) | 332 (54) | 4 | <0.001 |
| Body temperature at hospital arrival, °C | 35.3 (34.3–36.0) | 35.1 (34.4–35.8) | 35.4 (34.3–36.0) | 231 | 0.152 |
| Defibrillation after hospital arrival | 515 (74) | 63 (82) | 452 (73) | 0 | 0.101 |
| Tracheal intubation after hospital arrival | 606 (87) | 69 (90) | 537 (87) | 0 | 0.466 |
| Implementation of 12-lead ECG after ROSC | 471 (68) | 71 (92) | 400 (65) | 0 | <0.001 |
| ST elevation on 12-lead ECG after ROSC (n = 471) | 231 (49) | 41 (58) | 190 (48) | 0 | 0.112 |
| Arterial blood gases at hospital arrival | |||||
| pH | 6.92 (6.82–7.02) | 6.96 (6.82–7.08) | 6.92 (6.81–7.01) | 122 | 0.112 |
| PaCO2, mmHg | 71.2 (53.4–88.0) | 57.2 (42.4–76.1) | 72.4 (54.8–90.2) | 121 | <0.001 |
| PaO2, mmHg | 64.3 (26.1–143.0) | 89.6 (48.5–176.5) | 61.8 (25.4–139.1) | 128 | 0.084 |
| , mmol/L | 13.8 (10.4–17.5) | 11.9 (9.8–16.3) | 14.0 (10.5–17.6) | 127 | 0.050 |
| Lactate, mg/dL | 118 (88–145) | 118 (79–159) | 119 (89–145) | 131 | 0.980 |
| Coronary angiography | 500 (72) | 76 (99) | 424 (69) | 0 | <0.001 |
| Percutaneous coronary intervention | 285 (41) | 47 (61) | 238 (39) | 0 | <0.001 |
| Successful coronary reperfusion (n = 285) | 262 (38) | 45 (96) | 217 (91) | 9 | 0.136 |
| Intra-aortic balloon pump implantation | 413 (59) | 67 (87) | 346 (56) | 0 | <0.001 |
| Target temperature management | 307 (44) | 51 (66) | 256 (41) | 6 | <0.001 |
| Targeted hypothermia | 246 | 44 | 202 | ||
| Targeted normothermia | 61 | 7 | 54 | ||
| Drug administration during cardiac arrest | |||||
| Adrenaline | 655 (94) | 71 (92) | 584 (94) | 4 | 0.275 |
| Amiodarone | 386 (56) | 48 (62) | 338 (55) | 5 | 0.230 |
| Nifekalant | 42 (6) | 6 (8) | 36 (6) | 10 | 0.456 |
| Lidocaine | 25 (4) | 4 (5) | 21 (3) | 6 | 0.511 |
| Atropine | 16 (2) | 1 (1) | 15 (2) | 10 | 1.000 |
| Magnesium | 89 (13) | 17 (22) | 72 (12) | 10 | 0.012 |
| Vasopressin | 7 (1) | 1 (1) | 6 (1) | 8 | 0.567 |
| Aetiology of cardiac cause | 0 | <0.001 | |||
| Acute coronary syndrome | 360 (52) | 48 (62) | 312 (50) | ||
| Other heart disease | 179 (26) | 25 (32) | 154 (25) | ||
| Presumed cardiac cause | 156 (22) | 4 (5) | 152 (25) | ||
| Time course | |||||
| Call-to-field arrival interval, min | 7 (5–9) | 6 (5–7) | 7 (5–9) | 1 | <0.001 |
| Call-to-CPR initiation by EMS personnel interval, min | 9 (7–11) | 8 (6–10) | 9 (7–11) | 4 | 0.077 |
| Call-to-defibrillation by EMS personnel interval, min (n = 582) | 10 (9–13) | 10 (8–12) | 10 (9–13) | 1 | 0.167 |
| Call-to-hospital arrival interval, min | 31 (26–40) | 28 (23–35) | 32 (26–40) | 14 | 0.004 |
| Call-to-ECPR interval, min | 56 (47–68) | 49 (41–58) | 58 (48–68) | 2 | <0.001 |
| Call-to-PCI interval, min (n = 262) | 126 (94–171) | 108 (83–141) | 131 (96–179) | 12 | 0.008 |
| Hospital arrival-to-ECPR interval, min | 29 (20–38) | 23 (17–32) | 29 (21–39) | 2 | <0.001 |
| Hospital arrival-to-PCI interval, min (n = 262) | 98 (64–143) | 84 (60–114) | 106 (65–150) | 12 | 0.017 |
Results are shown as median (IQR) or number (%).
AED, automated external defibrillator; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; EMS, emergency medical service; PCI, percutaneous coronary intervention; ROSC, return of spontaneous circulation; VF, ventricular fibrillation; VT, ventricular tachycardia.
Association of percutaneous coronary intervention and target temperature management utilization with call-to-extracorporeal cardiopulmonary resuscitation interval
The association between call-to-ECPR interval and CPC value is shown in Figure 3A. While 25% of patients with a call-to-ECPR interval of 0–30 min had a favourable neurological outcome, this percentage decreased to 6% for an interval >100 min (P for trend <0.001). The association between call-to-ECPR interval and PCI utilization is shown in Figure 3B. The proportion of patients who underwent PCI decreased from 63% in those with a call-to-ECPR interval of 0–30 min to 41% in those with an interval >100 min (P for trend =0.034). The association between call-to-ECPR interval and TTM implementation is shown in Figure 3C. The proportion of patients who underwent TTM, including targeted hypothermia or normothermia, decreased from 88% in those with a call-to-ECPR interval of 0–30 min to 24% in those with an interval >100 min (P for trend <0.001).
Associations between call-to-extracorporeal cardiopulmonary resuscitation intervals and both a favourable neurological outcome and implementation of percutaneous coronary intervention and target temperature management. (A) Association between call-to-extracorporeal cardiopulmonary resuscitation interval and cerebral performance category value. (B) Association between call-to-extracorporeal cardiopulmonary resuscitation interval and percutaneous coronary intervention utilization. (C) Association between call-to-extracorporeal cardiopulmonary resuscitation interval and target temperature management implementation. Data are shown as the percent of the overall patient cohort. The number in each bar indicates the number of patients. CPC, cerebral performance category; ECPR, extracorporeal cardiopulmonary resuscitation; PCI, percutaneous coronary intervention; TTM, target temperature management.
Multivariable analysis for predicting a favourable neurological outcome at 30 days
Table 2 shows predictors of a favourable neurological outcome at 30 days as determined by logistic regression analysis. In univariable analysis, significant predictors (P < 0.05) were age, VF/pVT as the first documented rhythm at EMS arrival, prehospital transient ROSC, arterial pH at hospital arrival, call-to-ECPR interval, ACS, PCI, and TTM. In stepwise multivariable logistic regression analysis (n = 532), age, male gender, VF/pVT as the first documented rhythm at EMS arrival, prehospital transient ROSC, arterial pH at hospital arrival, call-to-ECPR interval, PCI, and TTM were independent predictors of a favourable neurological outcome at 30 days.
Univariable and multivariable analyses of factors predicting favourable neurological outcome at 30 days
| Univariable | Multivariable (n = 522) | Stepwise (n = 532) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | |
| Age, per year increment | 0.97 | 0.96–0.99 | <0.001 | 0.98 | 0.95–1.00 | 0.077 | 0.97 | 0.95–1.00 | 0.031 |
| Gender, male | 0.68 | 0.38–1.24 | 0.222 | 0.40 | 0.17–0.93 | 0.041 | 0.40 | 0.17–0.92 | 0.039 |
| Witnessed | 1.60 | 0.84–3.05 | 0.135 | 1.43 | 0.61–3.36 | 0.400 | |||
| Bystander-initiated CPR | 1.26 | 0.78–2.02 | 0.347 | 1.70 | 0.88–3.30 | 0.111 | |||
| VF/pulseless VT as the first documented rhythm at EMS arrival | 2.99 | 1.34–6.67 | 0.003 | 3.23 | 1.06–9.80 | 0.020 | 3.59 | 1.19–10.83 | 0.010 |
| Prehospital transient ROSC | 2.74 | 1.37–5.50 | 0.008 | 3.45 | 1.23–9.69 | 0.029 | 3.47 | 1.24–9.68 | 0.028 |
| pH, increase per 0.1 increment | 1.20 | 1.01–1.43 | 0.042 | 1.27 | 1.00–1.61 | 0.050 | 1.25 | 1.00–1.56 | 0.047 |
| Lactate, increase per mg/dL increment | 1.00 | 0.99–1.00 | 0.695 | 1.00 | 0.99–1.01 | 0.733 | |||
| Call-to-ECPR interval, increase per min increment | 0.96 | 0.94–0.98 | <0.001 | 0.96 | 0.94–0.99 | 0.002 | 0.96 | 0.94–0.99 | 0.001 |
| Acute coronary syndrome | 1.62 | 1.00–2.64 | 0.048 | 0.94 | 0.29–3.00 | 0.913 | |||
| Percutaneous coronary intervention | 2.50 | 1.54–4.07 | <0.001 | 2.53 | 0.80–8.02 | 0.099 | 2.30 | 1.14–4.66 | 0.019 |
| Target temperature management | 2.97 | 1.78–495 | <0.001 | 2.38 | 1.17–4.84 | 0.014 | 2.28 | 1.13–4.62 | 0.019 |
| Univariable | Multivariable (n = 522) | Stepwise (n = 532) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | |
| Age, per year increment | 0.97 | 0.96–0.99 | <0.001 | 0.98 | 0.95–1.00 | 0.077 | 0.97 | 0.95–1.00 | 0.031 |
| Gender, male | 0.68 | 0.38–1.24 | 0.222 | 0.40 | 0.17–0.93 | 0.041 | 0.40 | 0.17–0.92 | 0.039 |
| Witnessed | 1.60 | 0.84–3.05 | 0.135 | 1.43 | 0.61–3.36 | 0.400 | |||
| Bystander-initiated CPR | 1.26 | 0.78–2.02 | 0.347 | 1.70 | 0.88–3.30 | 0.111 | |||
| VF/pulseless VT as the first documented rhythm at EMS arrival | 2.99 | 1.34–6.67 | 0.003 | 3.23 | 1.06–9.80 | 0.020 | 3.59 | 1.19–10.83 | 0.010 |
| Prehospital transient ROSC | 2.74 | 1.37–5.50 | 0.008 | 3.45 | 1.23–9.69 | 0.029 | 3.47 | 1.24–9.68 | 0.028 |
| pH, increase per 0.1 increment | 1.20 | 1.01–1.43 | 0.042 | 1.27 | 1.00–1.61 | 0.050 | 1.25 | 1.00–1.56 | 0.047 |
| Lactate, increase per mg/dL increment | 1.00 | 0.99–1.00 | 0.695 | 1.00 | 0.99–1.01 | 0.733 | |||
| Call-to-ECPR interval, increase per min increment | 0.96 | 0.94–0.98 | <0.001 | 0.96 | 0.94–0.99 | 0.002 | 0.96 | 0.94–0.99 | 0.001 |
| Acute coronary syndrome | 1.62 | 1.00–2.64 | 0.048 | 0.94 | 0.29–3.00 | 0.913 | |||
| Percutaneous coronary intervention | 2.50 | 1.54–4.07 | <0.001 | 2.53 | 0.80–8.02 | 0.099 | 2.30 | 1.14–4.66 | 0.019 |
| Target temperature management | 2.97 | 1.78–495 | <0.001 | 2.38 | 1.17–4.84 | 0.014 | 2.28 | 1.13–4.62 | 0.019 |
CI, confidence interval; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; EMS, emergency medical service; OR, odds ratio; PCI, percutaneous coronary intervention; ROSC, return of spontaneous circulation; VF, ventricular fibrillation; VT, ventricular tachycardia.
Univariable and multivariable analyses of factors predicting favourable neurological outcome at 30 days
| Univariable | Multivariable (n = 522) | Stepwise (n = 532) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | |
| Age, per year increment | 0.97 | 0.96–0.99 | <0.001 | 0.98 | 0.95–1.00 | 0.077 | 0.97 | 0.95–1.00 | 0.031 |
| Gender, male | 0.68 | 0.38–1.24 | 0.222 | 0.40 | 0.17–0.93 | 0.041 | 0.40 | 0.17–0.92 | 0.039 |
| Witnessed | 1.60 | 0.84–3.05 | 0.135 | 1.43 | 0.61–3.36 | 0.400 | |||
| Bystander-initiated CPR | 1.26 | 0.78–2.02 | 0.347 | 1.70 | 0.88–3.30 | 0.111 | |||
| VF/pulseless VT as the first documented rhythm at EMS arrival | 2.99 | 1.34–6.67 | 0.003 | 3.23 | 1.06–9.80 | 0.020 | 3.59 | 1.19–10.83 | 0.010 |
| Prehospital transient ROSC | 2.74 | 1.37–5.50 | 0.008 | 3.45 | 1.23–9.69 | 0.029 | 3.47 | 1.24–9.68 | 0.028 |
| pH, increase per 0.1 increment | 1.20 | 1.01–1.43 | 0.042 | 1.27 | 1.00–1.61 | 0.050 | 1.25 | 1.00–1.56 | 0.047 |
| Lactate, increase per mg/dL increment | 1.00 | 0.99–1.00 | 0.695 | 1.00 | 0.99–1.01 | 0.733 | |||
| Call-to-ECPR interval, increase per min increment | 0.96 | 0.94–0.98 | <0.001 | 0.96 | 0.94–0.99 | 0.002 | 0.96 | 0.94–0.99 | 0.001 |
| Acute coronary syndrome | 1.62 | 1.00–2.64 | 0.048 | 0.94 | 0.29–3.00 | 0.913 | |||
| Percutaneous coronary intervention | 2.50 | 1.54–4.07 | <0.001 | 2.53 | 0.80–8.02 | 0.099 | 2.30 | 1.14–4.66 | 0.019 |
| Target temperature management | 2.97 | 1.78–495 | <0.001 | 2.38 | 1.17–4.84 | 0.014 | 2.28 | 1.13–4.62 | 0.019 |
| Univariable | Multivariable (n = 522) | Stepwise (n = 532) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | P-value | OR | 95% CI | P-value | OR | 95% CI | P-value | |
| Age, per year increment | 0.97 | 0.96–0.99 | <0.001 | 0.98 | 0.95–1.00 | 0.077 | 0.97 | 0.95–1.00 | 0.031 |
| Gender, male | 0.68 | 0.38–1.24 | 0.222 | 0.40 | 0.17–0.93 | 0.041 | 0.40 | 0.17–0.92 | 0.039 |
| Witnessed | 1.60 | 0.84–3.05 | 0.135 | 1.43 | 0.61–3.36 | 0.400 | |||
| Bystander-initiated CPR | 1.26 | 0.78–2.02 | 0.347 | 1.70 | 0.88–3.30 | 0.111 | |||
| VF/pulseless VT as the first documented rhythm at EMS arrival | 2.99 | 1.34–6.67 | 0.003 | 3.23 | 1.06–9.80 | 0.020 | 3.59 | 1.19–10.83 | 0.010 |
| Prehospital transient ROSC | 2.74 | 1.37–5.50 | 0.008 | 3.45 | 1.23–9.69 | 0.029 | 3.47 | 1.24–9.68 | 0.028 |
| pH, increase per 0.1 increment | 1.20 | 1.01–1.43 | 0.042 | 1.27 | 1.00–1.61 | 0.050 | 1.25 | 1.00–1.56 | 0.047 |
| Lactate, increase per mg/dL increment | 1.00 | 0.99–1.00 | 0.695 | 1.00 | 0.99–1.01 | 0.733 | |||
| Call-to-ECPR interval, increase per min increment | 0.96 | 0.94–0.98 | <0.001 | 0.96 | 0.94–0.99 | 0.002 | 0.96 | 0.94–0.99 | 0.001 |
| Acute coronary syndrome | 1.62 | 1.00–2.64 | 0.048 | 0.94 | 0.29–3.00 | 0.913 | |||
| Percutaneous coronary intervention | 2.50 | 1.54–4.07 | <0.001 | 2.53 | 0.80–8.02 | 0.099 | 2.30 | 1.14–4.66 | 0.019 |
| Target temperature management | 2.97 | 1.78–495 | <0.001 | 2.38 | 1.17–4.84 | 0.014 | 2.28 | 1.13–4.62 | 0.019 |
CI, confidence interval; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; EMS, emergency medical service; OR, odds ratio; PCI, percutaneous coronary intervention; ROSC, return of spontaneous circulation; VF, ventricular fibrillation; VT, ventricular tachycardia.
Discussion
This study compared OHCA patients who underwent ECPR for cardiac causes in terms of those who did or did not achieve a favourable neurological outcome at 30 days. The main findings were as follows: (i) patients with a favourable neurological outcome had a significantly shorter call-to-ECPR; (ii) a longer call-to-ECPR interval was associated with a significantly lower proportion of patients undergoing PCI or TTM; and (iii) in multivariable analysis, the call-to-ECPR interval, PCI, and TTM independently predicted a favourable neurological outcome at 30 days.
Study strengths
This study analysed a nationwide, multicentre, prospective registry of patients who underwent ECPR because of OHCA due to cardiac causes. ECPR was stringently defined as VA-ECMO use in OHCA patients who demonstrated cardiac arrest on hospital arrival and who did not achieve ROSC between hospital arrival and VA-ECMO initiation. Veno-arterial extracorporeal membrane oxygenation initiated for haemodynamic decompensation following sustained ROSC should not be considered ECPR, but this important distinction is often neglected in published reports, complicating clinical outcome analysis. Moreover, the timing of collapse is uncertain in the absence of witnesses, and initial bystander resuscitation attempts are also difficult to record accurately. We adopted emergency call receipt as the initial time used to evaluate the duration to ECPR.
Comparison of previous studies in patients with cardiac arrest
Table 3 lists comparative data from key studies on time management of ECPR and clinical outcomes of cardiac arrest patients. This study enrolled a larger number of OHCA patients than previous studies of patients who underwent ECPR as a result of cardiac arrest due to cardiac causes. In a retrospective, single-centre, observational registry analysis, Kuroki et al.15 reported that in ACS patients undergoing ECPR and PCI, shorter intervals from collapse to ECPR and from collapse to reperfusion predicted survival with favourable neurological outcomes at 30 days. In a multicentre cohort study in ACS patients undergoing ECPR and PCI, Kagawa et al.16 found that the median interval from collapse to ECPR was 40 min in patients still alive at 30 days, and early ECPR plus intra-arrest PCI was associated with improved outcomes. In a retrospective single-centre study in ischaemic refractory cardiac arrest patients undergoing ECPR, Cesana et al.11 reported that shorter interval from collapse to ECPR predicted survival. However, these studies included both OHCA and in-hospital cardiac arrest (IHCA) patients. Avalli et al.17 reported that IHCA patients derived more benefits from ECPR than OHCA patients, possibly due to shorter low-flow times and higher-quality CPR in IHCA patients. Future analyses evaluating clinical outcomes and the timing of ECPR and intra-arrest PCI should avoid this confusion.
Key studies on time management of ECPR and clinical outcome
| Author/journal | Study design | Study period/study country | Sample size, n | OHCA, n | Age/male | Initial shockable rhythm, n | Witnessed, n | Bystander- initiated CPR, n | Time to CCPR initiation, min | Time to ECPR initiation, min | PCI, n | Survival, n | Favourable neurological outcome, n |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Kawakami et al. | Prospective, observational, multicentre | June 2014 to December 2017/Japan | 695 | 695 (100%) | 59 (median) /585 (84%) | 503 (72%) | 542 (78%) | 344 (49%) | 9 (median) from call receipt | 56 (median) from call receipt | 285 (41%) | 151 (28%) at 30 days | 77 (11%) at 30 days |
| Yannopoulos et al. Lancet 20207 | Randomized, ECPR vs. CCPR, single-centre | August 2019 to June 2020/USA | 36 | 36 (100%) | 59 (mean) /25 (83%) | 36 (100%) | 24 (80%) | 25 (83%) | 6 (mean) from call receipt to EMS arrival | 59 (mean) from call receipt | CAG: 13 (87%) PCI: NA | 6 (43%) at hospital discharge | 3 (21%) at hospital discharge |
| Bartos et al. Circulation 202018 | Retrospective, observational, single-centre | December 2015 to February 2019/USA | 160 | 160 (100%) | 57 (mean) /126 (79%) | 160 (100%) | 121 (76%) | 105 (66%) | NA | 60 (mean) from professional CCPR initiation | NA | NA | 52 (33%) at hospital discharge and 6 months |
| Bougouin et al. Eur Heart J 202019 | Prospective, observational, multicentre | May 2011 to January 2018/France | 525 | 525 (100%) | 50 (mean) /442 (84%) | 358 (68%) | 508 (97%) | 424 (81%) | >5 from collapse: 66 (18%) | 90 (median) from CCPR initiation | 159 (54%) | 44 (8%) at hospital discharge | 32 at hospital discharge |
| Cesana et al. Eur Heart J Acute Cardiovasc Care 201811 | Retrospective observational single-centre | January 2011 to October 2015/Italy | 63 | 46 (73%) | 59 (mean) /55 (87%) | 41 (65%) | 63 (100%) | NA | 3 (mean) from collapse | 56 (mean) from CPR initiation | CAG: 63 (100%) PCI: NA | 13 (21%) at hospital discharge | 12 (19%) at hospital discharge |
| Kuroki et al. Resuscitation 201715 | Retrospective, observational, single-centre | January 2005 to June 2016/Japan | 119 (ACS) | 37 (31%) | 63 (mean) /108 (91%) | 39 (33%) | NA | NA | NA | 34 (mean) from collapse | 119 (100%) | 38 (32%) at 30 days | 38 (32%) at 30 days |
| Wengenmayer et al. Crit Care 201720 | Retrospective, observational, single-centre | October 2010 to May 2016/Germany | 133 | 59 (44%) | 59 (mean) /NA | 38.6% | NA | NA | 2.6 (mean) from collapse | 60 (mean) from collapse or CCPR initiation | CAG: 81% PCI: NA | 14.3% at hospital discharge | NA |
| Lamhaut et al. Resuscitation 201729 | Retrospective, observational, single-centre | 2011 to 2015/France | 156 | 156 (100%) | 51 (mean) /128 (82%) | 81 (58%) | NA | NA | 3.4 (mean) from collapse | 87 (mean) from CCPR initiation | 42 (30%) | NA | 21 (13%) at hospital discharge |
| Kagawa et al. Circulation 201216 | Retrospective, observational, multicentre | January 2004 to May 2011/Japan | 86 (ACS) | 42 (49%) | 63 (median) /70 (81%) | 46 (53%) | 77 (90%) | 67 (80%) | 1 (median) from collapse | 49 (median) from collapse | 61 (71%) | 25 (29%) at 30 days | 21 (24%) at hospital discharge |
| Avalli et al. Resuscitation 201217 | Retrospective, observational, single-centre | January 2006 to February 2011/Italy | 42 | 18 (43%) | OHCA: 46, IHCA: 67 (median) /23 (55%) | 28 (67%) | NA | 34 (81%) | 1 (median) from collapse | OHCA: 77, IHCA: 55 (median) from CCPR initiation | 17 (40%) | 12 (29%) at 28 days 11 (26%) at 6 months | 10 (24%) at 6 months |
| Author/journal | Study design | Study period/study country | Sample size, n | OHCA, n | Age/male | Initial shockable rhythm, n | Witnessed, n | Bystander- initiated CPR, n | Time to CCPR initiation, min | Time to ECPR initiation, min | PCI, n | Survival, n | Favourable neurological outcome, n |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Kawakami et al. | Prospective, observational, multicentre | June 2014 to December 2017/Japan | 695 | 695 (100%) | 59 (median) /585 (84%) | 503 (72%) | 542 (78%) | 344 (49%) | 9 (median) from call receipt | 56 (median) from call receipt | 285 (41%) | 151 (28%) at 30 days | 77 (11%) at 30 days |
| Yannopoulos et al. Lancet 20207 | Randomized, ECPR vs. CCPR, single-centre | August 2019 to June 2020/USA | 36 | 36 (100%) | 59 (mean) /25 (83%) | 36 (100%) | 24 (80%) | 25 (83%) | 6 (mean) from call receipt to EMS arrival | 59 (mean) from call receipt | CAG: 13 (87%) PCI: NA | 6 (43%) at hospital discharge | 3 (21%) at hospital discharge |
| Bartos et al. Circulation 202018 | Retrospective, observational, single-centre | December 2015 to February 2019/USA | 160 | 160 (100%) | 57 (mean) /126 (79%) | 160 (100%) | 121 (76%) | 105 (66%) | NA | 60 (mean) from professional CCPR initiation | NA | NA | 52 (33%) at hospital discharge and 6 months |
| Bougouin et al. Eur Heart J 202019 | Prospective, observational, multicentre | May 2011 to January 2018/France | 525 | 525 (100%) | 50 (mean) /442 (84%) | 358 (68%) | 508 (97%) | 424 (81%) | >5 from collapse: 66 (18%) | 90 (median) from CCPR initiation | 159 (54%) | 44 (8%) at hospital discharge | 32 at hospital discharge |
| Cesana et al. Eur Heart J Acute Cardiovasc Care 201811 | Retrospective observational single-centre | January 2011 to October 2015/Italy | 63 | 46 (73%) | 59 (mean) /55 (87%) | 41 (65%) | 63 (100%) | NA | 3 (mean) from collapse | 56 (mean) from CPR initiation | CAG: 63 (100%) PCI: NA | 13 (21%) at hospital discharge | 12 (19%) at hospital discharge |
| Kuroki et al. Resuscitation 201715 | Retrospective, observational, single-centre | January 2005 to June 2016/Japan | 119 (ACS) | 37 (31%) | 63 (mean) /108 (91%) | 39 (33%) | NA | NA | NA | 34 (mean) from collapse | 119 (100%) | 38 (32%) at 30 days | 38 (32%) at 30 days |
| Wengenmayer et al. Crit Care 201720 | Retrospective, observational, single-centre | October 2010 to May 2016/Germany | 133 | 59 (44%) | 59 (mean) /NA | 38.6% | NA | NA | 2.6 (mean) from collapse | 60 (mean) from collapse or CCPR initiation | CAG: 81% PCI: NA | 14.3% at hospital discharge | NA |
| Lamhaut et al. Resuscitation 201729 | Retrospective, observational, single-centre | 2011 to 2015/France | 156 | 156 (100%) | 51 (mean) /128 (82%) | 81 (58%) | NA | NA | 3.4 (mean) from collapse | 87 (mean) from CCPR initiation | 42 (30%) | NA | 21 (13%) at hospital discharge |
| Kagawa et al. Circulation 201216 | Retrospective, observational, multicentre | January 2004 to May 2011/Japan | 86 (ACS) | 42 (49%) | 63 (median) /70 (81%) | 46 (53%) | 77 (90%) | 67 (80%) | 1 (median) from collapse | 49 (median) from collapse | 61 (71%) | 25 (29%) at 30 days | 21 (24%) at hospital discharge |
| Avalli et al. Resuscitation 201217 | Retrospective, observational, single-centre | January 2006 to February 2011/Italy | 42 | 18 (43%) | OHCA: 46, IHCA: 67 (median) /23 (55%) | 28 (67%) | NA | 34 (81%) | 1 (median) from collapse | OHCA: 77, IHCA: 55 (median) from CCPR initiation | 17 (40%) | 12 (29%) at 28 days 11 (26%) at 6 months | 10 (24%) at 6 months |
ACS, acute coronary syndrome; CAG, coronary angiography; CCPR, conventional cardiopulmonary resuscitation; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; IHCA, in-hospital cardiac hospital; NA, not available; OHCA, out-of-hospital cardiac arrest; PCI, percutaneous coronary intervention.
Key studies on time management of ECPR and clinical outcome
| Author/journal | Study design | Study period/study country | Sample size, n | OHCA, n | Age/male | Initial shockable rhythm, n | Witnessed, n | Bystander- initiated CPR, n | Time to CCPR initiation, min | Time to ECPR initiation, min | PCI, n | Survival, n | Favourable neurological outcome, n |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Kawakami et al. | Prospective, observational, multicentre | June 2014 to December 2017/Japan | 695 | 695 (100%) | 59 (median) /585 (84%) | 503 (72%) | 542 (78%) | 344 (49%) | 9 (median) from call receipt | 56 (median) from call receipt | 285 (41%) | 151 (28%) at 30 days | 77 (11%) at 30 days |
| Yannopoulos et al. Lancet 20207 | Randomized, ECPR vs. CCPR, single-centre | August 2019 to June 2020/USA | 36 | 36 (100%) | 59 (mean) /25 (83%) | 36 (100%) | 24 (80%) | 25 (83%) | 6 (mean) from call receipt to EMS arrival | 59 (mean) from call receipt | CAG: 13 (87%) PCI: NA | 6 (43%) at hospital discharge | 3 (21%) at hospital discharge |
| Bartos et al. Circulation 202018 | Retrospective, observational, single-centre | December 2015 to February 2019/USA | 160 | 160 (100%) | 57 (mean) /126 (79%) | 160 (100%) | 121 (76%) | 105 (66%) | NA | 60 (mean) from professional CCPR initiation | NA | NA | 52 (33%) at hospital discharge and 6 months |
| Bougouin et al. Eur Heart J 202019 | Prospective, observational, multicentre | May 2011 to January 2018/France | 525 | 525 (100%) | 50 (mean) /442 (84%) | 358 (68%) | 508 (97%) | 424 (81%) | >5 from collapse: 66 (18%) | 90 (median) from CCPR initiation | 159 (54%) | 44 (8%) at hospital discharge | 32 at hospital discharge |
| Cesana et al. Eur Heart J Acute Cardiovasc Care 201811 | Retrospective observational single-centre | January 2011 to October 2015/Italy | 63 | 46 (73%) | 59 (mean) /55 (87%) | 41 (65%) | 63 (100%) | NA | 3 (mean) from collapse | 56 (mean) from CPR initiation | CAG: 63 (100%) PCI: NA | 13 (21%) at hospital discharge | 12 (19%) at hospital discharge |
| Kuroki et al. Resuscitation 201715 | Retrospective, observational, single-centre | January 2005 to June 2016/Japan | 119 (ACS) | 37 (31%) | 63 (mean) /108 (91%) | 39 (33%) | NA | NA | NA | 34 (mean) from collapse | 119 (100%) | 38 (32%) at 30 days | 38 (32%) at 30 days |
| Wengenmayer et al. Crit Care 201720 | Retrospective, observational, single-centre | October 2010 to May 2016/Germany | 133 | 59 (44%) | 59 (mean) /NA | 38.6% | NA | NA | 2.6 (mean) from collapse | 60 (mean) from collapse or CCPR initiation | CAG: 81% PCI: NA | 14.3% at hospital discharge | NA |
| Lamhaut et al. Resuscitation 201729 | Retrospective, observational, single-centre | 2011 to 2015/France | 156 | 156 (100%) | 51 (mean) /128 (82%) | 81 (58%) | NA | NA | 3.4 (mean) from collapse | 87 (mean) from CCPR initiation | 42 (30%) | NA | 21 (13%) at hospital discharge |
| Kagawa et al. Circulation 201216 | Retrospective, observational, multicentre | January 2004 to May 2011/Japan | 86 (ACS) | 42 (49%) | 63 (median) /70 (81%) | 46 (53%) | 77 (90%) | 67 (80%) | 1 (median) from collapse | 49 (median) from collapse | 61 (71%) | 25 (29%) at 30 days | 21 (24%) at hospital discharge |
| Avalli et al. Resuscitation 201217 | Retrospective, observational, single-centre | January 2006 to February 2011/Italy | 42 | 18 (43%) | OHCA: 46, IHCA: 67 (median) /23 (55%) | 28 (67%) | NA | 34 (81%) | 1 (median) from collapse | OHCA: 77, IHCA: 55 (median) from CCPR initiation | 17 (40%) | 12 (29%) at 28 days 11 (26%) at 6 months | 10 (24%) at 6 months |
| Author/journal | Study design | Study period/study country | Sample size, n | OHCA, n | Age/male | Initial shockable rhythm, n | Witnessed, n | Bystander- initiated CPR, n | Time to CCPR initiation, min | Time to ECPR initiation, min | PCI, n | Survival, n | Favourable neurological outcome, n |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Kawakami et al. | Prospective, observational, multicentre | June 2014 to December 2017/Japan | 695 | 695 (100%) | 59 (median) /585 (84%) | 503 (72%) | 542 (78%) | 344 (49%) | 9 (median) from call receipt | 56 (median) from call receipt | 285 (41%) | 151 (28%) at 30 days | 77 (11%) at 30 days |
| Yannopoulos et al. Lancet 20207 | Randomized, ECPR vs. CCPR, single-centre | August 2019 to June 2020/USA | 36 | 36 (100%) | 59 (mean) /25 (83%) | 36 (100%) | 24 (80%) | 25 (83%) | 6 (mean) from call receipt to EMS arrival | 59 (mean) from call receipt | CAG: 13 (87%) PCI: NA | 6 (43%) at hospital discharge | 3 (21%) at hospital discharge |
| Bartos et al. Circulation 202018 | Retrospective, observational, single-centre | December 2015 to February 2019/USA | 160 | 160 (100%) | 57 (mean) /126 (79%) | 160 (100%) | 121 (76%) | 105 (66%) | NA | 60 (mean) from professional CCPR initiation | NA | NA | 52 (33%) at hospital discharge and 6 months |
| Bougouin et al. Eur Heart J 202019 | Prospective, observational, multicentre | May 2011 to January 2018/France | 525 | 525 (100%) | 50 (mean) /442 (84%) | 358 (68%) | 508 (97%) | 424 (81%) | >5 from collapse: 66 (18%) | 90 (median) from CCPR initiation | 159 (54%) | 44 (8%) at hospital discharge | 32 at hospital discharge |
| Cesana et al. Eur Heart J Acute Cardiovasc Care 201811 | Retrospective observational single-centre | January 2011 to October 2015/Italy | 63 | 46 (73%) | 59 (mean) /55 (87%) | 41 (65%) | 63 (100%) | NA | 3 (mean) from collapse | 56 (mean) from CPR initiation | CAG: 63 (100%) PCI: NA | 13 (21%) at hospital discharge | 12 (19%) at hospital discharge |
| Kuroki et al. Resuscitation 201715 | Retrospective, observational, single-centre | January 2005 to June 2016/Japan | 119 (ACS) | 37 (31%) | 63 (mean) /108 (91%) | 39 (33%) | NA | NA | NA | 34 (mean) from collapse | 119 (100%) | 38 (32%) at 30 days | 38 (32%) at 30 days |
| Wengenmayer et al. Crit Care 201720 | Retrospective, observational, single-centre | October 2010 to May 2016/Germany | 133 | 59 (44%) | 59 (mean) /NA | 38.6% | NA | NA | 2.6 (mean) from collapse | 60 (mean) from collapse or CCPR initiation | CAG: 81% PCI: NA | 14.3% at hospital discharge | NA |
| Lamhaut et al. Resuscitation 201729 | Retrospective, observational, single-centre | 2011 to 2015/France | 156 | 156 (100%) | 51 (mean) /128 (82%) | 81 (58%) | NA | NA | 3.4 (mean) from collapse | 87 (mean) from CCPR initiation | 42 (30%) | NA | 21 (13%) at hospital discharge |
| Kagawa et al. Circulation 201216 | Retrospective, observational, multicentre | January 2004 to May 2011/Japan | 86 (ACS) | 42 (49%) | 63 (median) /70 (81%) | 46 (53%) | 77 (90%) | 67 (80%) | 1 (median) from collapse | 49 (median) from collapse | 61 (71%) | 25 (29%) at 30 days | 21 (24%) at hospital discharge |
| Avalli et al. Resuscitation 201217 | Retrospective, observational, single-centre | January 2006 to February 2011/Italy | 42 | 18 (43%) | OHCA: 46, IHCA: 67 (median) /23 (55%) | 28 (67%) | NA | 34 (81%) | 1 (median) from collapse | OHCA: 77, IHCA: 55 (median) from CCPR initiation | 17 (40%) | 12 (29%) at 28 days 11 (26%) at 6 months | 10 (24%) at 6 months |
ACS, acute coronary syndrome; CAG, coronary angiography; CCPR, conventional cardiopulmonary resuscitation; CPR, cardiopulmonary resuscitation; ECPR, extracorporeal cardiopulmonary resuscitation; IHCA, in-hospital cardiac hospital; NA, not available; OHCA, out-of-hospital cardiac arrest; PCI, percutaneous coronary intervention.
Importance of time management of extracorporeal cardiopulmonary resuscitation as a predictor of a favourable neurological outcome
This study’s stringent definition of ECPR was also adopted by the ARREST trial, which although small in size, was the first randomized interventional trial to compare ECPR with conventional CPR in OHCA patients.7 Extracorporeal cardiopulmonary resuscitation was performed in OHCA patients who did not achieve ROSC before hospital arrival or during the emergency department stay. The mean call-to-ECPR interval was 59 min, and 43% (6/14) of patients survived until hospital discharge. Inclusion of only patients with initial shockable rhythm and a higher proportion of bystander CPR may have been associated with a higher probability of a favourable outcome. Many studies have shown initial shockable rhythm to be associated with better outcomes after ECPR,18,19 and early recognition of cardiac arrest and initiation of bystander CPR before EMS arrival are known to improve survival and increase the likelihood of favourable neurological outcomes.3 Although our study included OHCA patients with shockable or non-shockable cardiac rhythm, with or without bystander CPR, multivariable analysis revealed that the call-to-ECPR interval and initial shockable rhythm independently predicted a favourable neurological outcome, whereas bystander CPR did not.
As CPR duration increases, the probability of survival with favourable functional status declines.5 Alternative strategies, for example ECPR, should be considered immediately after cardiac arrest rather than after failure of conventional CPR. Ideally, ECPR should be initiated within 60 min of cardiac arrest.20 However, the time of onset, which is often expressed as collapse, might be inaccurate in the absence of witnesses, and the duration of bystander CPR is also difficult to record accurately. Therefore, several studies have used emergency call receipt as a more reliable parameter before EMS arrival to exclude recall bias of patients or bystanders. For example, Rajan et al.4 used the response time, defined as the interval from call receipt to EMS arrival, as an alternative to bystander CPR duration. They found that it was an important determinant of outcomes, and should ideally be less than 5 min and certainly no longer than 10 min. Bartos et al.18 showed that ECPR was associated with more favourable neurological outcomes for professional CPR durations <60 min. We adopted emergency call receipt as the initial time for evaluating the duration to ECPR.
Percutaneous coronary intervention and target temperature management in out-of-hospital cardiac arrest patients who undergo extracorporeal cardiopulmonary resuscitation
In this study, implementation of PCI or TTM significantly predicted a favourable neurological outcome. Current guidelines recommend emergent PCI in patients with ROSC after cardiac arrest that is suspected to be associated with ST elevation.1,2 The appropriate timing of PCI is controversial in patients with ROSC after cardiac arrest without ST elevation, and the aforementioned guidelines recommend emergent PCI in patients with haemodynamic or electric instability.1,2 The COACT trial revealed that among OHCA patients who had initial shockable rhythm and no ST elevation after ROSC, immediate angiography was not superior to delayed angiography with respect to overall survival at 90 days.21 The TOMAHAWK trial assessed patients with OHCA of possible coronary origin who had either shockable or non-shockable rhythm and who did not have ST elevation after ROSC, and found similar results with respect to 30-day risk of death from any cause.22 In the COACT trial, 65% of patients who underwent CAG had coronary artery disease, and only 5% had acute thrombotic coronary occlusion.21 In the TOMAHWAK trial, a coronary culprit lesion was deemed to be the event trigger in 40% of patients.22 A low incidence of acute occlusion and ongoing ischaemia might make early revascularization less beneficial. Moreover, the majority of non-survivors in both studies died of neurologic complications after cardiac arrest,21,22 thereby attenuating the possible beneficial effect of coronary revascularization.
In patients with coma after cardiac arrest, current guidelines strongly recommend TTM with a constant target between 32°C and 36°C.1,2 The evidence to support this recommendation originated in trials suggesting a better neurological outcome in patients who underwent targeted hypothermia at 33°C.23,24 However, the TTM and TTM2 trials revealed that targeted hypothermia did not improve survival or neurological outcome compared with targeted normothermia at 36°C.25,26 In this study, 80% of patients with TTM underwent targeted hypothermia. Theoretically, targeted hypothermia may increase the risks of bleeding and sepsis,23–26 which are important complications in patients who have undergone ECPR. Further studies are needed to assess the beneficial and harmful effects of targeted hypothermia in ECPR patients.
Study limitations
This study has several potential limitations. First, the trial was not randomized. Second, the quality of CPR was not evaluated. Third, although several studies showed that signs of life during CPR are considered to be a favourable prognostic factor in ECPR patients,27,28 data on signs of life were unavailable in this study. Fourth, data were unavailable regarding treatment strategies after hospital arrival, angiographic characteristics of CAG, PCI procedures, complications of ECPR, and peripheral organ damage. Fifth, the frequency of IABP use was higher in patients with a favourable neurological outcome than in those without. Selection bias may have favoured using this strategy to treat patients with better presumed prognosis. Sixth, the concomitant use of ECPR and Impella support (ECPELLA) is currently available as a highly effective circulatory support system. Impella has been used in Japan since September 2017; thus, its use did not affect this study. The results might be different in the ECPELLA era.
Conclusion
A shorter call-to-ECPR interval predicted better clinical outcomes in OHCA patients who underwent ECPR for cardiac causes. Given the scarce data on time intervals to ECPR in nationwide, multicentre, prospective registries of OHCA patients, this study should facilitate prehospital and in-hospital care by multidisciplinary teams and improve patient survival.
Supplementary material
Supplementary material is available at European Heart Journal: Acute Cardiovascular Care online.
Acknowledgements
We are deeply indebted to all members and institutions of the JAAM-OHCA Registry for their contributions. The participating institutions of the JAAM-OHCA Registry are listed at the following URL: http://www.jaamohca-web.com/list/.
Funding
The authors have not declared a special grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Approval of the research protocol
The protocol was approved by the Ethics Committee of Kyoto University (R1045), and by the National Cerebral and Cardiovascular Center (R20046) as the corresponding institution. All participating hospitals approved the JAAM-OHCA registry protocol.
Conflict of interest: none declared.
Comments