a b s t r a c t

Purpose: In out-of-hospital cardiac arrest (OHCA) patients resuscitated with veno-arterial extracorporeal mem- brane oxygenation (VA-ECMO), known as extracorporeal cardiopulmonary resuscitation (ECPR), bleeding is a common complication. The purpose of this study was to assess the risk factors for Bleeding complications in ECPR patients.

Methods: We retrospectively analyzed the data for OHCA patients admitted to our hospital and resuscitated with ECPR between October 2009 and December 2016. We compared patients with and without major bleeding (i.e. the Bleeding Academic Research Consortium class >= 3 bleeding) within 24 h of hospital admission. Patients, whose bleeding complication was not evaluated, were excluded.

Results: During the study period, 133 OHCA patients were resuscitated with ECPR, of whom 102 (77%) were in- cluded. In total, 71 (70%) patients experienced major bleeding. There were significant differences in age (median 65 vs. 50 years, P b 0.001), prior antiplatelet therapy (25% vs. 3%, P = 0.008), hemoglobin (median 11.6 vs. 12.6 g/dL, P = 0.003), platelet count (median 125 vs. 155 x 10³/uL, P = 0.001), and D-dimer levels on admission (median 18.8 vs. 6.7 ug/mL, P b 0.001) among patients with and those without major bleeding. Multivariate anal- ysis showed significant associations between major bleeding and D-dimer levels (odds ratio, 1.066; 95% confi- dence interval, 1.018-1.116). Area under receiver-operating characteristic curve, which describes the accuracy of D-dimer levels in predicting major bleeding, was 0.76 (95% confidence interval, 0.66-0.87).

Conclusion: D-dimer levels may predict major bleeding in ECPR patients, suggesting that hyperfibrinolysis may be related to bleeding.

(C) 2017

Introduction

witnessed arrest, shockable initial cardiac rhythm, and bystander- initiated cardiopulmonary resuscitation (CPR) are key predictors of out-of-hospital cardiac arrest (OHCA) survival [1]. In a recent study, which analyzed the nationwide, prospective, population-based registry of OHCA patients in Japan, Kitamura et al. reported the prognosis of 43,762 witnessed OHCA patients with shockable initial cardiac rhythm; 13,660 (31%) achieved return of spontaneous circulation (ROSC) before arrival at the hospital and 8886 (20%) had a favorable neurological out- come [2]. However, the prognosis of refractory cardiopulmonary arrest (CPA) patients, who do not achieve ROSC with conventional CPR, is poor and the Probability of survival decreases with every minute of CPR [3]. Therefore, physicians should consider CPR methods that differ from the conventional methods to improve patients’ outcomes.

* Corresponding author at: 1-1-6, Tsukumodai, Suita-city, Osaka 565-0862, Japan.

E-mail address: [email protected] (T. Otani).

Recently, the effectiveness of CPR with veno-arterial extracorporeal membrane oxygenation (VA-ECMO) for refractory CPA, known as extra- corporeal CPR (ECPR), has been reported [4-7]. The VA-ECMO system for OHCA patients entails the provision of organ support, cerebral and myocardial perfusion, and oxygenation, thus increasing the probability of ROSC, decreasing the mortality rate, and improving the cerebral per- formance outcomes [4-7]. The 2015 American Heart Association guide- lines recommended ECPR in OHCA cases in Class IIb, where the time without blood flow is brief and when the condition that led to the cardi- ac arrest is reversible, amenable to Heart transplantation, or amenable to revascularization [8].

However, during VA-ECMO support, anticoagulation is needed to prevent thrombosis in the VA-ECMO circuit; consumption of platelet and coagulation factors due to tissue injury and contact activation in the artificial membrane can increase bleeding risk [9]. Therefore, re- gardless of primary diseases, bleeding is known as a major complication during VA-ECMO management [9-13]. In addition, Post cardiac arrest patients sometimes experience hyperfibrinolysis caused by tissue hypo- perfusion and hypoxia, and chest compression induced trauma [14-17].

https://doi.org/10.1016/j.ajem.2017.11.016

0735-6757/(C) 2017

target temperature management”>Therefore in ECPR patients, the frequency of bleeding complications might be higher than that among other ECMO users. However, the fre- quency of bleeding complications in ECPR patients is not well reported and the factors that relate to bleeding complications are poorly understood.

It is important to recognize the bleeding risk in the acute phase when ECPR is performed to prevent and/or treat the bleeding. The pur- pose of this retrospective study was to verify the frequency of bleeding complications in OHCA patients resuscitated with ECPR. Additionally, we aimed to determine the factors related to bleeding in ECPR patients.

Methods

Study design and Inclusion/exclusion criteria for ECPR

Between October 2009 and December 2016, we retrospectively ana- lyzed the data for OHCA patients >= 16 years old admitted to our hospital (a tertiary care referral center) who were resuscitated with ECPR. Pa- tients were provided ECPR if ROSC >= 5 min was not achieved within 20 min via conventional CPR. Patients were excluded if they had a ‘do not attempt resuscitation’ order, did not provide informed consent, had an existing terminal illness, had an exogenous cause of arrest (e.g. trauma), were physically inactive, or if pericardial effusion was ob- served by echocardiography and acute aortic dissection was suspected as the cause of arrest. Finally, patients who were evaluated as having no chance of survival by physicians (e.g. non-shockable initial cardiac rhythm and asystole was continued during CPR) were also excluded. All patients received out-of-hospital resuscitation from emergency medical service (EMS) teams, according to the Japanese CPR guidelines [18].

Exclusion criteria among ECPR cases for inclusion in this study

After initiation of VA-ECMO, patients who were transferred or died within a Short period of time, and whose bleeding complications were not evaluated, were excluded from this analysis.

Study definitions and outcome variables

‘Major bleeding’ was defined as the Bleeding Academic Research Consortium (BARC) type >= 3 bleeding (i.e. overt bleeding plus hemoglo- bin drop of >= 3 g/dL) within 24 h of hospital admission [19]. One unit of Packed red blood cells (140 mL) was defined as 1 g/dL hemoglobin, to correct for transfusion in the patients. ‘Fatal bleeding’ was defined as BARC type 5 bleeding [19]. Neurological outcomes were evaluated using the Cerebral Performance Category on day of hospital dis- charge; CPC 1-2 were regarded as favorable and CPC 3-5 were regarded as unfavorable.

The VA-ECMO system and ECPR management

All ECPR cases underwent VA-ECMO in the coronary intervention laboratory, using a Capiox emergency bypass system, a Capiox-SX mem- brane oxygenator, and a Terumo EBS centrifugal pump (Terumo Inc., Tokyo, Japan). The femoral artery and vein were cannulated with 13.5-Fr and 19.5-Fr catheters, using the percutaneous Seldinger tech- nique, while maintaining conventional CPR. This method is the same as that detailed in our previous report [20]. The pump flow was initially set at 2.5-3.0 L/min, and 3000 units of unfractionated heparin (UFH) was administered immediately after ECPR initiation to maintain an acti- vated partial thromboplastin time (aPTT) 1.5-2.5 times that of the base- line aPTT. An intra-aortic balloon pump was inserted for all the cases after the Coronary angiography or percutaneous coronary inter- vention (PCI).

The method of CAG and the definitions of the etiologies of arrest

All ECPR cases underwent CAG after VA-ECMO initiation. The meth- od and the interpretation of the CAG results have been previously re- ported [20]. All acute myocardial infarction (AMI) cases underwent PCI. Dual antiplatelet therapy (aspirin 200 mg and clopidogrel 200 mg or prasugrel 20 mg were administrated via nasogastric tube) and the in- jection of an additional 5000 units of UFH, were performed before PCI. During PCI, UFH was also injected to maintain activated thrombin time above 250 s. If there were no abnormal findings in CAG and right ventricle dilation was found using echocardiography, pulmonary angi- ography was performed. Pulmonary embolism was defined either as a filling defect or as amputation of a pulmonary artery branch [21]. Thrombolytic treatment was not offered in this study population. The other definitions of the cardiac etiologies have been described previous- ly [20].

target temperature management

All ECPR patients were received Targeted temperature management (TTM) using a heat exchanger attached to the VA-ECMO circuit. Pa- tients, who were hemodynamically unstable regardless the use of vaso- pressor and Mechanical circulatory support, had their core temperature maintained at 36 ?C for 24 h. Other patients, who were hemodynamical- ly stable, had their core temperature maintained at 34 ?C for 24 h followed by rewarming over a 2 day period, warming at 1 ?C per day.

Management of bleeding and blood transfusions

If significant bleeding occurred, UFH administration was stopped immediately until bleeding was controlled. Red blood cells were trans- fused to maintain Hemoglobin levels above 8 g/dL. If significant bleeding continued, fresh frozen plasma was transfused to maintain Fibrinogen levels above 100 mg/dL and platelets were transfused to maintain plate- let count above 50,000 cells/uL. After bleeding was controlled, UFH was re-started to maintain aPTT 1.5-2.5 times that of the baseline aPTT. All ECPR cases underwent computed tomography after VA-ECMO initia- tion, CAG, and PCI to check for bleeding complications.

Statistical analysis

Continuous variables were presented as medians (interquartile range: IQR) and categorical variables were presented as frequencies and percentages. We compared the baseline characteristics, pre-hospi- tal clinical course, and blood samples on admission among patients with and without major bleeding. Continuous variables were compared using the Mann-Whitney U test, and categorical variables were com- pared using the chi-squared or Fisher’s exact tests. Multiple logistic re- gression analysis (forced-entry method) was performed to evaluate the correlation between major bleeding and factors that were found to be significant (P b 0.05) in the preliminary univariate analyses, and pre- sented odds ratio (OR) and 95% confidence intervals (CI) for each of the factors assessed. We constructed a receiver-operating characteristic (ROC) curve to illustrate the cut off values for platelet count and D- dimer levels on admission to differentiate between those with and without major bleeding, and we calculated the area under the ROC curves (AUC) to evaluate the predictive value of major bleeding. All sta- tistical analyses were performed using SPSS software (version 21.0; SPSS Inc., Chicago, IL, USA). P b 0.05 was considered statistically significant.

Ethics

Written informed consent for the CAG, the PCI, and for ECPR was ob- tained from the patients’ family members. The study protocol conformed to the concepts of the Declaration of Helsinki and its

amendments, and was approved by the Institutional Review Board of our hospital.

Results

Patients’ characteristics

During the study period, 102 patients met the eligibility criteria for inclusion in the study (Fig. 1), of which 71 (70%) experienced major bleeding. The cause of arrest was AMI in 52 (51%) patients who sequen- tially underwent PCI (Table 1). The etiologies of cardiac arrest among the remaining patients included ischemic cardiomyopathy (n = 8), other cardiomyopathy (n = 6), primary arrhythmia (n = 7), accidental hypothermia (n = 4), pulmonary embolism (n = 3), coronary vaso- spasm (n = 1), valvular disease (n = 2), sepsis (n = 2), hypoxia due to acute heart failure (n = 1), hypoxia due to chronic obstructive pul- monary disease (n = 1), diabetic ketoacidosis (n = 1), and in some pa- tients the cause of arrest was unknown (n = 14).

Comparison between patients with and without major bleeding

We compared the baseline characteristics, pre-hospital clinical course, and blood samples on admission for patients with and without major bleeding (Table 1). Among those who experienced major bleed- ing, age (median 65 vs. 50 years, P b 0.001), the frequency of prior anti- platelet therapy (25% vs. 3%, P = 0.008), and D-dimer level on admission (median 18.8 vs. 6.7 ug/mL, P b 0.001) were higher, while he- moglobin levels (median 11.6 vs. 12.6 g/dL, P = 0.003) and platelet count (median 125 vs. 155 x 10³/uL, P = 0.001) on admission were lower when compared to those without major bleeding. The TTM at 34 ?C was performed less frequently among those with major bleeding (21% vs. 45%, P = 0.013). Of those with major bleeding, 35 (49%) and 17 (24%) experienced bleeding at the VA-ECMO puncture site and in the gastrointestinal tract, respectively. CPR related bleeding occurred for 20 (28%) and 10 (14%) of patients in the thorax (hemothorax, thoracic wall, and mediastinal hemorrhage) and the abdomen (intra-Abdominal hemorrhage due to liver injury), respectively. Additionally, 5 (7%) of pa- tients experienced nasal bleeding and 7 (10%) experienced alveolar hemorrhages. Among those with major bleeding, 53 (75%) required red blood cell transfusions and the median quantity of blood required for transfusion was 560 mL (IQR 0-1400 mL) of red blood cell. Platelet

transfusions were performed for 4 patients; all patients required 250 mL of platelets to be transfused. BARC type 3b bleeding (overt bleeding plus hemoglobin drop of >= 5 g/dL) occurred in 59 (83%) pa- tients and fatal bleeding occurred in 5 (7%) patients. Age, prior anti- platelet therapy, hemoglobin, platelet count, D-dimer levels, which significantly differed between those with and without major bleeding were evaluated using a multivariate analysis. Age (OR, 1.053; 95% CI, 1.009-1.100), platelet count (OR, 0.984; 95% CI, 0.971-0.997), and D-

dimer levels (OR, 1.066; 95% CI, 1.018-1.116), were significantly associ- ated with major bleeding (Table 2).

Accuracy of platelet count and D-dimer levels as a predictor for major bleeding

Fig. 2 shows the ROC curves describing the diagnostic accuracy of platelet count and D-dimer levels in predicting major bleeding. The AUC was 0.70 (95% CI, 0.59-0.81) and the cut-off value of 148

x 10³/uL for platelet count corresponded to a sensitivity of 0.58 and a specificity of 0.79 as predictors for major bleeding (a). The AUC was

0.76 (95% CI, 0.66-0.87) and the cut-off value of 7.6 ug/mL for D- dimer levels corresponded to a sensitivity of 0.82 and a specificity of

0.61 as predictors for major bleeding (b).

Discussion

There was a significant association between major bleeding and D- dimer levels on admission in ECPR cases, and therefore D-dimer levels may be useful as a predictor for major bleeding.

Bleeding complications in ECPR cases

The bleeding risk during ECMO is relatively high (46-60%); there have been several reports about the frequency of bleeding complica- tions in ECMO cases [10-12]. Aubron et al. reported that among 149 ECMO cases (n = 111 for VA-ECMO and n = 39 for ECPR), 89 (60%) ex- perienced bleeding that was related to prior anticoagulation use before ECMO initiation [10]. In our study, there was no relationship between prior anticoagulation use and bleeding, however there were only 7 (7%) patients with prior anticoagulation use, thus the small sample size may have impacted this finding. Staudacher et al. reported that among 93 VA-ECMO cases (n = 32 for both VA-ECMO and ECPR), 56

Fig. 1. Consort diagram for the patients with out-of-hospital cardiac arrest detailing study eligibility criteria. Abbreviations: OHCA, out-of-hospital cardiac arrest; ECPR, extracorporeal cardiopulmonary resuscitation; ROSC, return of spontaneous circulation; CPR, cardiopulmonary resuscitation; ECMO, extracorporeal membrane oxygenation.

Table 1

Comparison of clinical and demographic characteristics of out-of-hospital cardiac arrest patients resuscitated using veno-arterial extracorporeal membrane oxygenation, with and without bleeding.

	All		Major bleeding		No major bleeding	P value
	n = 102		n = 71		n = 31
Age, years	62 (47-70)		65 (54-72)		50 (40-62)	b0.001
Males	87 (85)		60 (85)		27 (87)	1
Prior antiplatelet therapy	19 (19)		18 (25)		1 (3)	0.008
Prior anticoagulation	7 (7)		6 (8)		1 (3)	0.67
Witnessed arrest	87 (85)		62 (87)		25 (81)	0.38
Bystander-initiated CPR	48 (47)		31 (44)		17 (55)	0.30
Shockable initial cardiac rhythm	72 (71)		52 (73)		20 (65)	0.37
Call (or witnessed by EMS) to hospital arrival, min	37 (30-44)		36 (30-42)		38 (32-46)	0.32
Call (or witnessed by EMS) to ECMO, min	50 (45-59)		50 (45-57)		51 (46-63)	0.44
Use of adrenaline, mg	3 (2-4)		3 (2-4)		3 (2-4)	0.59
Acute myocardial infarction	52 (51)		39 (55)		13 (42)	0.23
Hemoglobin, g/dL	12.0 (10.8-13.4)		11.6 (10.3-12.8)		12.6 (11.9-14.0)	0.003
Platelet count, x103/uL	129 (103-160)		125 (94-145)		155 (121-177)	0.001
Prothrombin time, s	13 (12-15)		13 (12-17)		13 (12-14)	0.13
Fibrinogen, mg/dL	181 (129-227)		174 (129-226)		186 (139-223)	0.44
D-dimer, ug/mL	12.7 (6.5-25.1)		18.8 (8.3-28.0)		6.7 (3.7-11.6)	b0.001
Serum lactate concentration, mmol/L	14.0 (11.0-16.7)		14.0 (10.8-16.7)		13.3 (11.6-16.6)	0.85
Initial loading dose of unfractionated heparin, units	8000 (3000-8000)		8000 (3000-8000)		8000 (3000-8000)	0.14
34 ?C target temperature management	29 (28)		15 (21)		14 (45)	0.013
Successful of weaned ECMO	45 (44)		29 (41)		16 (52)	0.31
favorable neurological outcome	20 (20)		14 (20)		6 (19)	0.97
In-hospital death	70 (69)		50 (70)		20 (65)	0.55
Bleeding sites (not mutually exclusive)
ECMO puncture site			35 (49)
Thorax			20 (28)
Abdomen			10 (14)
Gastrointestinal tract			17 (24)
Others			12 (17)
Red blood cell transfusions in 24 h, mL			560 (0-1400)
Fresh frozen plasma transfusions in 24 h, mL			0 (0-1440)
Platelets transfusions in 24 h, mL			0 (0-0)
Red blood cell transfusions			53 (75)
BARC type 3b bleeding			59 (83)
Fatal bleeding (BARC type 5)			5 (7)

Data are presented as the number (column %) of patients or median (interquartile range).

CPR, cardiopulmonary resuscitation; EMS, emergency medical service; ECMO, extracorporeal membrane oxygenation; BARC, bleeding academic research consortium.

(60%) experienced bleeding and 29 (31%) experienced bleeding which required red blood cell transfusions [11]. In a meta-analysis investigat- ing complications associated with ECMO for the treatment of cardiogen- ic shock and cardiac arrest, Cheng et al. reported that among 260 ECMO cases, 120 (46%) experienced major or significant bleeding [12]. Howev- er, the above-mentioned studies included veno-venous ECMO and non- ECPR cases, therefore they were not comparable to the patients includ- ed in our study.

There have been a few reports about complications associated with ECMO among ECPR cases only. Kagawa et al. reported that among 77 ECPR cases (n = 38 for in-hospital cardiac arrest [IHCA] and n = 39 for OHCA), 49 (64%) experienced bleeding or hematomas. The median duration of conventional CPR for IHCA was 25 min [4]. Blumenstein et al. reported that among 52 IHCA patients who received ECPR, 17 (33%) experienced bleeding or hematomas which required blood trans- fusions. The median duration of CPR was 33 min [22]. The frequency of

Table 2

Multivariate analysis of factors associated with major bleeding among out-of-hospital car- diac arrest patients resuscitated using veno-arterial extracorporeal membrane oxygenation

	Odds ratio	95% confidence interval	P-value
Age (per 1 year increase)	1.053	1.009-1.100	0.018
Prior antiplatelet therapy	8.372	0.855-82.010	0.07
Hemoglobin (per 1 g/dL increase)	0.926	0.716-1.199	0.56
Platelet count (per 1 x 103/uL	0.984	0.971-0.997	0.014
increase)
D-dimer (per 1 ug/mL increase)	1.066	1.018-1.116	0.006

bleeding complication in our study (70%) was higher than these previ- ous studies, however these studies included IHCA patients and the dura- tion of conventional CPR was shorter than that in our study, therefore longer CPR duration may affect the increase of bleeding complications. Our study is important in that it shows that the frequency of bleeding complications is associated with ECPR in OHCA patients.

There were no significant relationships between major bleeding and treatments after ECMO initiation (e.g. PCI, loading dose of UFH). The TTM of 34 ?C was used less frequent in those with major bleeding (21% vs. 45%, P = 0.013), because hemodynamic instability was more frequent in cases with bleeding and was therefore not included in mul- tivariate analysis. Staudacher et al. reported that the dual antiplatelet therapy after ECMO initiation was not related to the occurrence of bleeding complications [11]. These results suggest that bleeding compli- cation in ECPR cases are not related to treatments after ECMO initiation, but rather to the patients’ clinical and demographic characteristics on admission.

Factors related to bleeding complications in ECPR

Multivariate analysis showed that age, platelet count, and D-dimer levels were significantly associated with major bleeding (Table 2). In- creasing age is also known to increase bleeding risk under antiplatelet or anticoagulation therapy [23,24]. During ECMO support, consumption of platelet and coagulation factors occurs, and the use of anticoagulants is required, which may increase the risk of bleeding especially in elderly patients.

Post cardiac arrest patients sometimes complicate disseminated in- travascular coagulation (DIC) and hyperfibrinolysis [14-16]. Tissue

Fig. 2. Receiver-operating characteristic curves for platelet count and D-dimer levels on admission for prediction of major bleeding. The area under the receiver-operative characteristic curves were 0.70 (95% confidence interval, 0.59-0.81) in platelet count (a) and 0.76 (95% confidence interval, 0.66-0.87) in D-dimer levels (b) as a predictor for major bleeding.

hypoperfusion and hypoxia during cardiac arrest and reperfusion injury after ROSC induce the release of inflammatory cytokines such as tumor necrosis factor-?. Consequently, the tissue factors dependent on coagu- lation are activated, leading to massive thrombin generation and fibrin formation. These changes lead to excessive release of tissue plasmino- gen activator, resulting in a transient increase in activated protein C, causing hyperfibrinolysis [14]. Wada et al. reported that among 338 re- suscitated OHCA patients, 208 (62%) and 73 (22%) experienced DIC and hyperfibrinolysis, respectively [15]. In this study, platelet count and D- dimer levels, which are known markers of DIC, are related to bleeding (Table 2). In addition, the AUC describing the diagnostic accuracy of the D-dimer levels in the prediction of major bleeding was 0.76 (95% CI, 0.66-0.87), which was higher than that of platelet count (Fig. 2). D- dimer, a Fibrin degradation product that is created during the fibrinoly- sis process after fibrin formation, is an indicator of hyperfibrinolysis [15]. Thus, hyperfibrinolysis after cardiac arrest was considered to be re- lated to bleeding complications in ECPR cases. However, fibrin/fibrino- gen degradation products and the plasmin-?2 plasmin inhibitor complex, which are known to be more ideal indicators for the evalua- tion of hyperfibrinolysis than D-dimer levels, were not investigated in this study [14-16]. Therefore, further studies are needed to identify the relationship between bleeding complications in ECPR patients and hyperfibrinolysis. In addition, Hayakawa et al. reported that trauma pa- tients sometimes experience elevated D-dimer levels, which are associ- ated with consumptive coagulopathy and hyperfibrinolysis, and high D- dimer levels at hospital admission are strong predictors of the need for massive transfusions [25]. ECPR cases, which require chest compres- sions for a long duration, sometimes complicate CPR related trauma [17]. In our study, there were 20 and 10 cases with thorax and abdomen trauma, respectively, caused by CPR among those with major bleeding; D-dimer levels in this study may be affected by not only post-cardiac ar- rest situation but CPR related trauma.

Outcomes of ECPR patients and the relationship with bleeding complication

In this study population, 32 (31%) survived hospital discharge, and 20 (20%) had a favorable neurological outcome, and there was no differ- ence in the outcomes between those with and without major bleeding in our study (Table 1). Conversely, Aubron, et al. reported that bleeding events in ECMO users were associated with hospital mortality [10]. This difference may be due to differences in the study population. Among Aubron’s study population, 26% (39/149) of post-surgical patients re- quired ECMO, and 38% (56/149) required surgical intervention for

bleeding [10]. In our study, the frequency of major bleeding was high (70%), however fatal bleeding was experienced by only 5 (7%) patients; the majority of the remaining forms of bleeding could be dealt with via blood transfusions only. Therefore there may be no differences in out- come with and without major bleeding.

Limitations

There were several limitations in this study. This was a single center, retrospective study with a relatively small sample size; not all refractory CPA patients received ECPR. We did not include the ECPR patients whose bleeding complication was not evaluated, nor did we consider the influence of causes of arrest on D-dimer levels. It is known that bleeding at the ECMO puncture site is not only affected by hyperfibrinolysis but also by technical problems, however we did not investigate technical problems. Finally, patients with non-witnessed ar- rest were also included in this study, and the pre-hospital time was cal- culated as the time from the call.

Conclusion

In ECPR cases, D-dimer levels on admission were higher among those who experienced major bleeding and may be useful as a predictor for bleeding, suggesting that hyperfibrinolysis may be related to major bleeding.

Sources of funding

This research did not receive any specific grant from funding agen- cies in the public, commercial or not-for-profit sectors.

Conflicts of interest

The authors declare that there are no conflicts of interest.

Acknowledgment

We thank Editage (www.editage.jp) for English language editing.

References

1. Sasson C, Rogers MAM, Dahl J, Kellermann AL. Predictors of survival from out-of- hospital cardiac arrest: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes 2010;3:63-81.
2. Kitamura T, Kiyohara K, Sakai T, Matsuyama T, Hatakeyama T, Shimamoto T, et al. Public-access defibrillation and out-of-hospital cardiac arrest in Japan. N Engl J Med 2016;375:1649-59.
3. 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.
4. Kagawa E, Inoue I, Kawagoe T, Ishihara M, Shimatani Y, Kurisu S, et al. Assessment of outcomes and differences between in- and out-of-hospital cardiac arrest patients treated with cardiopulmonary resuscitation using Extracorporeal life support. Resus- citation 2010;81:968-73.
5. Kim SJ, Jung JS, Park JH, Park JS, Hong YS, Lee SW. An optimal transition time to ex- tracorporeal cardiopulmonary resuscitation for predicting good neurological out- come in patients with out-of-hospital cardiac arrest: a propensity-matched study. Crit Care 2014;18:535.
6. Ortega-Deballon I, Hornby L, Shemie SD, Bhanji F, Guadagno E. Extracorporeal resus- citation for refractory out-of-hospital cardiac arrest in adults: a systematic review of international practices and outcomes. Resuscitation 2016;101:12-20.
7. Sakamoto T, Morimura N, Nagao K, Asai Y, Yokota H, Nara S, et al. Extracorporeal car- diopulmonary resuscitation versus conventional cardiopulmonary resuscitation in adults with out-of-hospital cardiac arrest: a prospective observational study. Resus- citation 2014;85:762-8.
8. Brooks SC, Anderson ML, Bruder E, Daya MR, Gaffney A, Otto CW, et al. Part 6: alter- native techniques and ancillary devices for cardiopulmonary resuscitation: 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015;132:S436-43.
9. Murphy DA, Hockings LE, Andrews RK, Aubron C, Gardiner EE, Pellegrino VA, et al. Extracorporeal membrane oxygenation-hemostatic complications. Transfus Med Rev 2015;29:90-101.
10. Aubron C, DePuydt J, Belon F, Bailey M, Schmidt M, Sheldrake J, et al. Predictive fac- tors of bleeding events in adults undergoing extracorporeal membrane oxygenation. Ann Intensive Care 2016;6:97.
11. Staudacher DL, Biever PM, Benk C, Ahrens I, Bode C, Wengenmayer T. Dual antiplate- let therapy (DAPT) versus no antiplatelet therapy and incidence of major bleeding in patients on venoarterial extracorporeal membrane oxygenation. PLoS One 2016; 11:e0159973.
12. Cheng R, Hachamovitch R, Kittleson M, Patel J, Arabia F, Moriguchi J, et al. Complica- tions of extracorporeal membrane oxygenation for treatment of cardiogenic shock and cardiac arrest: a meta-analysis of 1,866 adult patients. Ann Thorac Surg 2014; 97:610-6.
13. Iwashita Y, Yukimitsu M, Matsuduki M, Yamamoto A, Ishikura K, Imai H. Use of a fixed, body weight-unadjusted loading dose of unfractionated heparin for extracor- poreal cardiopulmonary resuscitation. J Intensive Care 2015;3:33.
14. Adrie C, Monchi M, Laurent I, Um S, Yan SB, Thuong M, et al. Coagulopathy after suc- cessful cardiopulmonary resuscitation following cardiac arrest: implication of the protein C anticoagulant pathway. J Am Coll Cardiol 2005;46:21-8.
15. Wada T, Gando S, Ono Y, Maekawa K, Katabami K, Hayakawa M, et al. Disseminated intravascular coagulation with the fibrinolytic phenotype predicts the outcome of patients with out-of-hospital cardiac arrest. Thromb J 2016;14:43.
16. Viersen VA, Greuters S, Korfage AR, Van der Rijst C, Van Bochove V, Nanayakkara PW, et al. Hyperfibrinolysis in out of hospital cardiac arrest is associated with markers of hypoperfusion. Resuscitation 2012;83:1451-5.
17. Hellevuo H, Sainio M, Nevalainen R, Huhtala H, Olkkola KT, Tenhunen J, et al. Deeper chest compression - more complications for cardiac arrest patients? Resuscitation 2013;84:760-5.
18. Japan Resuscitation Council. Japan Resuscitation Council resuscitation guidelines 2010. Tokyo: Health Shuppansha; 2011[in Japanese].
19. Mehran R, Rao SV, Bhatt DL, Gibson CM, Caixeta A, Eikelboom J, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding Academic Research Consortium. Circulation 2011;123:2736-47.
20. Otani T, Sawano H, Oyama K, Morita M, Natsukawa T, Kai T. Resistance to conven- tional cardiopulmonary resuscitation in witnessed out-of-hospital cardiac arrest pa- tients with shockable initial cardiac rhythm. J Cardiol 2016;68:161-7.
21. Konstantinides SV, Torbicki A, Agnelli G, Danchin N, Fitzmaurice D, Galie N, et al. ESC guidelines on the diagnosis and management of acute pulmonary embolism. Eur Heart J 2014;2014(35):3033-69.
22. Blumenstein J, Leick J, Liebetrau C, Kempfert J, Gaede L, Gross S, et al. Extracorporeal life support in cardiovascular patients with observed refractory in-hospital cardiac arrest is associated with favourable short and long-term outcomes: a propensity- matched analysis. Eur Heart J Acute Cardiovasc Care 2016;5:13-22.
23. Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fi- brillation: the Euro Heart Survey. Chest 2010;138:1093-100.
24. Yeh RW, Secemsky EA, Kereiakes DJ, Normand SL, Gershlick AH, Cohen DJ, et al. De- velopment and validation of a prediction rule for benefit and harm of dual antiplate- let therapy beyond 1 year after percutaneous coronary intervention. JAMA 2016; 315:1735-49.
25. Hayakawa M, Maekawa K, Kushimoto S, Kato H, Sasaki J, Ogura H, et al. High D- dimer levels predict a poor outcome in patients with severe trauma, even with high fibrinogen levels on arrival: a Multicenter retrospective study. Shock 2016; 45:308-14.