Characteristics of Bleeding Complications in Patients with Severe COVID-19 Requiring Veno-venous Extracorporeal Membrane Oxygenation in Japan

Abstract

Background Complications during veno-venous extracorporeal membrane oxygenation (VV-ECMO) are associated with in-hospital mortality. Asian patients on extracorporeal membrane oxygenation (ECMO) have higher risks of bleeding and in-hospital mortality than Caucasian patients. This study aimed to characterize and identify bleeding complications and their associated factors related to in-hospital mortality in patients with severe coronavirus disease 2019 (COVID-19) requiring VV-ECMO in Japan.

Methods In this retrospective observational analysis, the prospective nationwide multicenter registry was used to track real-time information from intensive care units throughout Japan during the COVID-19 pandemic. VV-ECMO patients' registry data between February 1, 2020 and October 31, 2022 were used.

Results This study included 441 patients; 178 (40%) had bleeding complications in the following sites: 20% at the cannulation site, 16% in the gastrointestinal tract, 16% in the ear–nose–throat, 13% at the tracheostomy site, 9% intrathoracic, 6% intracranial, and 5% in the iliopsoas. Anticoagulation was discontinued in >50% of patients with intracranial, iliopsoas, and gastrointestinal tract bleeding. ECMO was discontinued in one-third of patients with intracranial, intramuscular, and iliopsoas hemorrhages. Multivariable logistic regression analysis revealed that only gastrointestinal tract bleeding was associated with in-hospital mortality (odds ratio: 2.49; 95% confidence interval: 1.11–5.60; p = 0.03).

Conclusion Incidence of bleeding complications was 40% in the Japanese population. Gastrointestinal tract bleeding emerged as a significant predictor of adverse outcomes, necessitating further research into preventive strategies and optimized care protocols. These findings can guide the management of VV-ECMO patients with COVID-19.

Keywords

bleeding - coronavirus disease 2019 - gastrointestinal tract - respiratory distress syndrome - mortality

Introduction

The number of patients requiring veno-venous extracorporeal membrane oxygenation (VV-ECMO) for the management of refractory acute respiratory distress syndrome has increased, especially during the coronavirus disease 2019 (COVID-19) pandemic.[1] [2] [3] [4] [5]

Bleeding complications associated with VV-ECMO are common and potentially lethal.[6] In particular, intracranial hemorrhage is reported in many countries as a bleeding complication associated with in-hospital mortality.[7] Although the incidence of bleeding complications in patients requiring VV-ECMO is >30% in European countries,[8] the incidence in patients with COVID-19 requiring VV-ECMO is reportedly higher than that in patients with non-COVID-19 acute respiratory distress syndrome. This is attributed to higher doses of anticoagulation regimens, severe acute respiratory distress syndrome coronavirus 2-associated vasculitis, microbleeds associated with critical illness, and other COVID-19-specific factors.[9] [10]

Bleeding complications during VV-ECMO have been reported to be associated with in-hospital mortality and the need for renal replacement therapy for acute kidney injury, infection, and poor neurological outcomes.[11] Additionally, Asian patients—including those who are Japanese—on extracorporeal membrane oxygenation (ECMO) have a higher risk of bleeding and in-hospital mortality than Caucasian patients.[12] [13] However, no study has evaluated the characteristics of bleeding complications during COVID-19-related VV-ECMO in the Japanese population using nationwide cohort data.

Therefore, this study aimed to characterize the bleeding complications in Japanese patients with severe COVID-19 requiring VV-ECMO and identify their associated factors.

Methods

Study Design and Patients

In this retrospective analysis, a prospective nationwide multicenter registry, the Cross Intensive Care Unit Searchable Information System (CRISIS) database was used to track real-time information from intensive care units throughout Japan during the COVID-19 pandemic. The CRISIS collects data from 738 of 1,223 Japanese facilities, including intensive care units, cardiac care units, and tertiary emergency medical and critical care centers in Japan. Although there is no officially approved ECMO center in Japan, participating facilities were registered at certified institutions by the Japanese Society of Intensive Care Medicine, the Japanese Association for Acute Medicine, and the Japanese Society of Respiratory Care Medicine. These facilities are staffed by board-certified doctors of emergency and critical care medicine, anesthesiology, and intensive care medicine.

Data from patients registered in CRISIS who met the following inclusion criteria were analyzed: age ≥18 years, laboratory-confirmed diagnosis of COVID-19 (using real-time polymerase chain reaction/next-generation sequencing), and undergone VV-ECMO for refractory acute respiratory distress syndrome. Patients with missing information on study variables (characteristics of bleeding complications and in-hospital mortality) were excluded. The registry data between February 1, 2020 and October 31, 2022 were used in the analysis. CRISIS was approved by the Institutional Review Board of Hiroshima University (approval number: E-1965) and each participating institute. In addition, this study was approved by the Institutional Review Board of Yokohama City University (approval number: B200700034), and the need for informed consent was waived due to its retrospective nature. Instead, an opt-out statement was posted on the Web site. The study was conducted according to the principles of the Declaration of Helsinki.

Data Collection and Definitions

The following patient data were collected from the CRISIS database: age, sex, body mass index, pre-ECMO ratio of arterial oxygen partial pressure to fractional inspired oxygen, pre-ECMO positive end-expiratory pressure, number of ventilatory days before ECMO, outcome of ECMO (weaning success or deceased while on ECMO), duration of ECMO, duration of ventilator use, and in-hospital mortality.

Additionally, the following data were retrospectively collected using a predesigned standardized case record form for this study linked with the CRISIS database ([Supplementary Fig. S1], available in the online version): ethnicity, pre-existing coagulation disorder, anticoagulant drugs administered during ECMO (unfractionated heparin, argatroban, or nafamostat mesylate [NM]), anticoagulation management index, bleeding complications: anatomical sites (vascular access, gastrointestinal, ear–nose–throat, tracheostomy site, intrathoracic, intracranial, iliopsoas, intramuscular, intraabdominal, and others), onset time, diagnostic procedures, and hemostatic intervention types.

Bleeding complications were defined as instances in which bleeding required a clinical intervention such as transfusion. Physicians at each facility determined bleeding complications based on the information from electronic medical records.

Statistical Analysis

First, we described the incidence (n, %), onset timing (median, interquartile range), diagnostic procedures (n, %), and intervention types (n, %) for each bleeding complication. The onset timing was compared among the bleeding complications using the Kruskal–Wallis test, and Steel–Dwass analysis was added to examine the onset timing.

The factors associated with each bleeding complication were also examined. The chi-square test or Fisher's exact test was used to compare categorical variables, and the Mann–Whitney U test was used to compare continuous variables between the groups.

Finally, to identify factors associated with in-hospital mortality, a multivariable logistic regression analysis was performed using independent variables with p-values of <0.05 in univariate comparisons and variables reported in previous studies. The variation inflation factor was checked to avoid multi-collinearity (variance inflation factor >10 as a violation). Moreover, a sensitivity analysis was performed, excluding variables with >5% missing study variables. The association of hemostatic interventions (surgical, endoscopic, or transcatheter arterial embolization) with in-hospital mortality was also examined in the bleeding complications group.

All statistical tests were two-tailed, and statistical significance was set at a p-value of <0.05. All statistical analyses were performed using JMP 17 (SAS Institute, Inc., Cary, North Carolina, United States).

Results

Study Participants

In total, 643 patients were enrolled in this study, among whom 202 were excluded owing to missing information on study variables, resulting in a final sample size of 441 patients from 57 facilities ([Fig. 1] and [Supplementary Fig. S2], available in the online version).

Fig. 1 Flowchart of patients on VV-ECMO included in this study. VV-ECMO, veno-venous extracorporeal membrane oxygenation.

Characteristics of Each Bleeding Complication

In total, 178 (40%) patients had bleeding complications ([Table 1]). Their incidences were as follows: cannulation site bleeding, 22%; gastrointestinal tract bleeding, 16%; ear–nose–throat bleeding, 16%; tracheostomy site bleeding, 13%; intrathoracic hemorrhage, 9%; intracranial hemorrhage, 6%; and iliopsoas hemorrhage, 5%. Anticoagulation was discontinued in more than half of the patients with intramuscular, iliopsoas, gastrointestinal tract, intrathoracic, and intracranial bleeding. Twelve of the 49 patients with gastrointestinal tract bleeding required endoscopic hemostasis, and 8 of the 16 patients with iliopsoas hemorrhage required transcatheter arterial embolization. ECMO was discontinued in one-third of the patients with intracranial, intramuscular, and iliopsoas hemorrhages for hemostasis. Cannulation site, ear–nose–throat, and tracheostomy site bleeding were treated surgically in most patients.

Table 1 Characteristics of each bleeding complication

Variable [frequency (%)/median (IQR)]	Cannulation site[a]		GI tract		Ear–nose–throat		Tracheostomy site[b]		Intrathoracic		Intracranial		Iliopsoas		Intramuscular		Intra-abdominal		Other
Incidence[d]	69	[22%]	49	[16%]	49	[16%]	40	[13%]	29	[9%]	18	[6%]	16	[5%]	13	[4%]	3	[1%]	22	[7%]
Diagnostic procedures
Physical examination	65	[94%]	47	[96%]	49	[100%]	39	[98%]	19	[65%]	8	[44%]	5	[31%]	9	[69%]	2	[67%]	17	[77%]
Laboratory	14	[20%]	20	[41%]	3	[6%]	3	[7%]	5	[17%]	0	[0%]	8	[50%]	6	[46%]	1	[33%]	6	[27%]
CT	0	[0%]	0	[0%]	2	[4%]	1	[3%]	13	[45%]	17	[94%]	14	[88%]	9	[69%]	2	[66%]	1	[5%]
Others	0	[0%]	0	[0%]	0	[0%]	0	[0%]	4	[14%]	0	[0%]	1	[6%]	1	[7%]	0	[0%]	0	[0%]
Intervention types
Surgical intervention[c]	53	[77%]	0	[0%]	35	[71%]	38	[95%]	1	[4%]	0	0%]	0	0%]	5	[38%]	1	[33%]	6	[27%]
Discontinuation of. anticoagulant	9	[13%]	30	[61%]	8	[16%]	14	[35%]	16	[55%]	9	[50%]	10	[63%]	10	[77%]	1	[33%]	5	[22%]
Endoscopic hemostasis	0	[0%]	12	[24%]	0	[0%]	0	[0%]	1	[3%]	0	[0%]	0	[0%]	0	[0%]	0	[0%]	0	[0%]
TAE	0	[0%]	6	[12%]	2	[4%]	1	[3%]	2	[7%]	0	[0%]	8	[50%]	4	[31%]	2	[66%]	1	[5%]
Discontinuation of ECMO	1	[1%]	1	[2%]	0	0%]	1	[3%]	2	[7%]	3	[17%]	3	[19%]	4	[30%]	0	0%]	2	[9%]
Others	2	[3%]	6	[12%]	6	[13%]	1	[3%]	4	[14%]	1	[6%]	0	[0%]	0	[0%]	0	[0%]	1	[5%]

Abbreviations: CT, computed tomography; ECMO, extracorporeal membrane oxygenation; ENT, ear, nose, and throat; GI, gastrointestinal; TAE, transcatheter arterial embolization.

^a Bleeding from a peripheral cannulation site such as the neck, groin, or axilla.

^b Bleeding not only from tracheostomy site but also from oral and airway after tracheostomy.

^c Including compression hemostasis and skintight sutures.

^d Incidence was calculated as each bleeding complication/all bleeding complications.

Onset Timing of Each Bleeding Complication

A significant difference was observed in onset timing among the bleeding complications (p < 0.001; [Fig. 2]). Cannulation site bleeding was more common immediately after ECMO introduction than other complications. Gastrointestinal tract bleeding and intracranial and intramuscular hemorrhage occurred later than catheter site complications. These sites of hemorrhage were observed to extend beyond 3 weeks from the initiation of ECMO ([Supplementary Table S1], available in the online version).

Fig. 2 Onset timing of each bleeding complication. This figure compares the timing of the onset of each bleeding complication. GI, gastrointestinal.

Factors Associated with Bleeding Complications

[Table 2] shows the factors associated with each bleeding complication. No differences were observed in patient characteristics. Unfractionated heparin was the most commonly used anticoagulant, and an activated partial thromboplastin time (APTT) of 40 to 60 seconds was commonly achieved. Although awake ECMO was more common for iliopsoas and tracheostomy site bleeding (p = 0.02), no differences in rehabilitation or prone position rates were found between the different bleeding complications.

Table 2 Factors associated with each bleeding complication

Variable [frequency (%)/median (IQR)]	No bleeding complication		Cannulation site		GI tract		Ear–nose–throat		Tracheostomy site		Intrathoracic		Intracranial		Iliopsoas		Intramuscular		Intra-abdominal		Other		p-Value
	(n = 263)		(n = 69)		(n = 49)		(n = 49)		(n = 40)		(n = 29)		(n = 18)		(n = 16)		(n = 13)		(n = 3)		(n = 23)
Patients' characteristics
Age (years)	56	[48–64]	56	[48–63]	64	[55–70]	60	[51–66]	57	[49–64]	59	[50–67]	61	[55–68]	64	[52–70]	56	[49–64]	66	[58–73]	57	[53–63]	0.11
Male	213	[81]	60	[87]	41	[85]	41	[82]	33	[79]	20	[83]	13	[72]	13	[81]	11	[92]	1	[33]	19	[83]	0.58
Ethnicity, Japanese	194	[95]	63	[93]	42	[91]	46	[92]	40	[100]	29	[100]	18	[100]	16	[100]	13	[100]	3	[100]	18	[78]	0.55
Body mass index (kg/m2)	28	[25–33]	28	[24–31]	27	[25–29]	27	[25–30]	28	[25–31]	27	[23–29]	26	[24–31]	28	[23–32]	27	[24–33]	28	[26–29]	28	[25–29]	0.84
HFNC  before ventilator	98	[41]	30	[43]	16	[32]	18	[37]	13	[33]	9	[39]	4	[22]	5	[31]	4	[31]	3	[100]	8	[35]	0.74
Time to ECMO  from ventilator use (days)	1	[0–4]	2	[1–6]	3	[0–6]	1	[1–6]	4	[1–7]	3	[0–9]	4	[2–8]	6	[2–10]	7	[2–10]	1	[0–1]	6	[1–11]	0.66
ECMO management
APTT management 40–60 s	114	[57]	53	[78]	29	[62]	36	[72]	36	[90]	14	[51]	14	[77]	10	[63]	8	[62]	1	[33]	13	[57]	0.22
APTT management 60–80 s	48	[24]	17	[25]	13	[27]	9	[18]	3	[8]	7	[24]	2	[11]	3	[19]	2	[16]	0	[0]	5	[22]	0.34
ACT management 160–200 s	54	[27]	10	[14]	9	[18]	9	[18]	2	[4]	5	[13]	3	[10]	3	[17]	2	[15]	1	[33]	3	[13]	0.65
ACT management 180–220 s	23	[11]	8	[12]	6	[12]	6	[12]	3	[8]	5	[17]	2	[11]	2	[12]	1	[8]	0	[0]	2	[9]	0.95
TEG management	1	[0.5]	3	[4]	3	[6]	1	[2]	2	[5]	4	[14]	0	[0]	0	[0]	0	[0]	0	[0]	1	[5]	0.26
Heparin use	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	100	[100]	>0.99
Argatroban use	2	[3]	2	[3]	2	[5]	5	[10]	4	[10]	1	[3]	1	[5]	0	[0]	1	[8]	1	[33]	3	[13]	0.15
Nafamostat mesylate use	13	[19]	13	[19]	14	[29]	9	[19]	9	[23]	7	[24]	4	[22]	0	[0]	2	[15]	0	[0]	4	[17]	0.49
Prone position	169	[64]	34	[50]	30	[61]	30	[61]	23	[58]	11	[38]	14	[77]	11	[69]	10	[77]	2	[66]	16	[70]	0.06
Rehabilitation[a]	19	[7]	3	[4]	1	[2]	1	[2]	4	[10]	1	[3]	0	[0]	1	[6]	0	[0]	1	[33]	2	[9]	0.07
Awake ECMO[b]	40	[19]	27	[39]	14	[29]	9	[18]	21	[53]	4	[14]	5	[28]	7	[44]	2	[18]	1	[33]	4	[17]	0.02
Duration of ECMO (days)	8	[5–14]	19	[11–38]	20	[11–41]	19	[9–31]	21	[10–28]	22	[11–35]	14	[0–34]	14	[5–43]	17	[5–43]	55	[5–106]	19	[11–41]	0.85

Abbreviations: ACT, activated clotting time; APTT, activated partial thromboplastin time; ECMO, extracorporeal membrane oxygenation; GI, gastrointestinal; HFNC, high-flow nasal cannula; TEG, thromboelastography.

^a Rehabilitation means sitting on the edge of the bed during ECMO.

^b Awake ECMO is a state of consciousness and spontaneous breathing.

Factors Associated with In-Hospital Mortality

The in-hospital mortality rate was 32.2% for all patients, 60.0% for those with bleeding complications, and 40.1% for those with nonbleeding complications (p < 0.001). [Table 3] shows the characteristics of the survivors and nonsurvivors at the time of hospital discharge. The following variables were used for multivariable logistic regression analysis: age, body mass index, duration of ECMO, duration of ventilator use, NM use, and incidence of gastrointestinal tract, ear–nose–throat, or intrathoracic bleeding. Age (odds ratio: 1.04; 95% confidence interval: 1.01–1.07; p = 0.004), duration of ECMO (1.03; 1.01–1.05; p < 0.001), and incidence of gastrointestinal tract bleeding (2.49; 1.11–5.60; p = 0.03) were significantly associated with in-hospital mortality.

Table 3 Factors associated with in-hospital mortality

Variable [frequency (%)/median (IQR)]	Nonsurvival		Survival		Univariate analysis	Multivariable analysis
	(n = 142)		(n = 299)		p-Value	Odds	95% CI	p-Value
Patients' characteristics
Age (years)	62	[55–69]	55	[48–63]	<0.001	1.04	[1.01–1.07]	0.004
Male	117	[82.4]	247	[82.]	0.96
Ethnicity, Japanese	122	[95.3]	239	[95.2]	0.97
Body mass index (kg/m2)	27	[24–30]	29	[25–33]	0.008	1.00	[0.96–1.05]	0.89
History of coagulation disorder	1	[0.7]	4	[1.4]	0.50
HFNC before ventilator	58	[47.2]	100	[37.0]	0.06
Time to ECMO from ventilator use (days)	3	[0–7]	1	[0–4]	0.17
ECMO management characteristics
Prone position during ECMO	86	[60.1]	193	[64.6]	0.66
Rehabilitation	26	[20.1]	65	[25.6]	0.23
Awake ECMO
Management index of anticoagulated therapy
APTT management 40–60 s	77	[61.1]	163	[64.4]	0.53
APTT management 60–80 s	33	[26.2]	54	[21.3]	0.29
ACT management 160–200 s	25	[19.8]	56	[22.3]	0.61
ACT management 180–220 s	19	[15.1]	26	[10.3]	0.17
TEG management	3	[2.4]	4	[1.6]	0.69
UFH use	100	[100]	100	[100]
Argatroban use	6	[4.3]	8	[2.8]	0.41
Nafamostat mesylate use	19	[13.6]	21	[7.3]	0.04	1.19	[0.49–2.87]	0.70
Outcomes
Duration of ECMO (days)	22	[7–39]	9	[5–15]	<0.001	1.03	[1.01–1.05]	<0.001
Duration of ventilator (days)	37	[23–58]	20	[12–36]	<0.001	0.99	[0.98–1.00]	0.16
Type of bleeding complications
Cannulation site	25	[17.6]	36	[12.4]	0.12
GI tract	30	[21.3]	17	[5.7]	<0.001	2.49	[1.11–5.60]	0.03
Ear–nose–throat	23	[16.2]	24	[8.0]	0.01	1.56	[0.71–3.38]	0.26
Tracheostomy site	17	[12.0]	22	[7.4]	0.12
Intrathoracic	17	[12.0]	11	[4.0]	0.001	2.27	[0.87–5.93]	0.09
Intracranial	9	[6.3]	9	[3.0]	0.11
Iliopsoas	7	[4.9]	9	[3.0]	0.41
Intramuscular	5	[3.5]	8	[2.7]	0.76
Intra-abdominal	1	[0.7]	2	[0.7]	0.96
Others	10	[7.0]	12	[4.0]	0.18

Abbreviations: ACT, activated clotting time; APTT, activated partial thromboplastin time; CI, confidence interval; ECMO, extracorporeal membrane oxygenation; GI, gastrointestinal; HFNC, high-flow nasal cannula; IQR, interquartile range; NIV, noninvasive positive pressure ventilation; P/F, PaO₂/FIO₂ ratio; s, seconds; UFH, unfractionated heparin.

Sensitivity analysis of the factors associated with in-hospital mortality, excluding variables with >5% missing study variables, showed similar results as the main analysis ([Supplementary Table S2], available in the online version).

The performance of hemostatic interventions for bleeding complications during ECMO was not associated with in-hospital mortality (50.3% vs. 49.7%, p = 0.09).

Discussion

In this study, the incidence of all bleeding complications was 40% in Japanese patients with severe COVID-19 requiring VV-ECMO, and the most common bleeding complication was cannulation site bleeding, followed by gastrointestinal tract bleeding, ear–nose–throat bleeding, and tracheostomy site bleeding. Gastrointestinal tract bleeding was the only bleeding complication associated with in-hospital mortality. Additionally, the performance of hemostatic interventions for bleeding complications was not associated with in-hospital mortality.

In a review of the current literature on patients with COVID-19 on ECMO, the incidence of bleeding complications ranged from 27 to 42%.[9] [10] [14] In France, the incidence of overall bleeding complications was 49%, cannulation site bleeding was 18%, ear–nose–throat bleeding was 12%, intrathoracic hemorrhage was 6%, intracranial hemorrhage was 8%, and gastrointestinal tract bleeding was 7.6%.[15] In the United States, the overall incidence was 28%, and the incidence of intracranial hemorrhage was 4.6%.[16] In the United Kingdom, the overall incidence was 31%, intracranial hemorrhage was 10.5%, intrathoracic hemorrhage was 7.8%, and gastrointestinal tract bleeding was 3.8%.[17] These results were different in each country; however, intracranial hemorrhage was commonly associated with in-hospital mortality.[10] [14] [15] [16] [17] On the other hand, ECMO registry data from the same Asian country, China, showed an 18% incidence of bleeding complications, cannulation site bleeding of 7.1%, intracranial hemorrhage of 2.8%, intrathoracic bleeding of 1.5%, and gastrointestinal tract bleeding of 3.5%. Intracranial hemorrhage was not documented to be associated with in-hospital mortality, as in Japan.[18] The characteristics of bleeding complications during ECMO may vary across countries.

This study showed that gastrointestinal tract bleeding was a common bleeding complication in Japan. Gastrointestinal tract bleeding was associated with in-hospital mortality, tended to develop later than other bleeding complications, and was more common in patients who received prolonged ECMO management (>21 days; [Fig. 2] and [Supplementary Table S1], available in the online version). Previous studies have suggested that prolonged ECMO management can lead to multiple organ failure, which is associated with in-hospital mortality.[19] [20] Multi-organ failure is followed by gastrointestinal mucosal damage, leading to gastrointestinal tract bleeding.[21] Therefore, gastrointestinal tract bleeding was assumed to be the most common complication associated with in-hospital mortality in Japan.

The reasons why gastrointestinal tract bleeding is more common in Japan than in other countries remain unclear. There was no clear difference in the duration of ECMO between Japan and other countries,[10] [14] [15] [16] [17] suggesting that the duration of ECMO could not be the reason for the high incidence of gastrointestinal tract bleeding. Unfractionated heparin was commonly used, and there was no difference in APTT management (40–60 seconds between Japan and other countries.[9] However, NM, which is infrequently used in other countries,[9] was prescribed in Japan for thromboprophylaxis in ECMO and continuous renal replacement therapy[22] or as a treatment for COVID-19 to prevent viral entry into cells.[23]

NM is a broad-spectrum, synthetic serine protease inhibitor used in Japan and Korea. The dosing for DIC, a continuous infusion of 0.06 to 0.20 mg/kg/h, is used.[24] The indicated dose to prevent blood coagulation is a continuous infusion at 20 to 50 mg/h.[22] Compared with unfractionated heparin, NM reduces the risk of thrombosis, with no significant difference in bleeding risk.[24] However, in patients with COVID-19, some studies have reported an association between bleeding complications and NM use for thromboprophylaxis in ECMO or as a treatment for COVID-19,[25] [26] [27] necessitating further studies.

Conversely, intracranial hemorrhage was not associated with in-hospital mortality in Japan; this may be due to the few cases of early intracranial hemorrhage. Intracranial hemorrhage usually occurs within 4 days of ECMO initiation,[28] leads to treatment discontinuation, and is associated with early mortality.[29] [30] [31] However, in the present study, two-thirds of the cases of intracranial hemorrhage occurred >10 days after ECMO initiation.

Differences in management practices and ethics may contribute to differences in the risk of in-hospital mortality from intracranial hemorrhage in Europe, the United States, and Asia[14] [15] [16] [17] [18]; this aspect warrants further research. On the other hand, as previously reported, the present study showed a higher incidence of iliopsoas hemorrhage than that in other countries.[32] [33] Awake ECMO management was also associated with iliopsoas hemorrhage.[33]

This study has some limitations. First, there may have been a selection bias owing to the voluntary registration system. Second, detailed clinical information on individual patients was unavailable. Therefore, we could not assess the severity of each bleeding complication. Information on various potential confounders, such as the use of oral anticoagulants, antiplatelet medications, and anti-ulcer drugs before ECMO induction, was lacking. Lastly, the decision to initiate or terminate ECMO or discharge from the intensive care unit was left to the judgement of the attending physician, and no standardized protocols were used.

In conclusion, the incidence of bleeding complications was 40% in the Japanese population, according to the nationwide cohort data used in this study. Gastrointestinal tract bleeding was the only bleeding complication associated with in-hospital mortality, and the performance of hemostatic interventions for bleeding complications was not associated with in-hospital mortality. The characteristics of bleeding complications during ECMO may vary across countries. Therefore, individualized management strategies should be developed to prevent bleeding complications.

Conflict of Interest

None declared.

Acknowledgment

We thank the members of Japan ECMO Net for COVID-19 who contributed to the management of ECMO patients.

Ethical Approval Statement

This study was approved by the Institutional Review Board of Yokohama City University (approval number: B200700034), and the need for informed consent was waived owing to its retrospective nature. Instead, an opt-out statement was posted on the Web site. The study was conducted according to the principles of the Declaration of Helsinki.

• References

• 1 Peek GJ, Mugford M, Tiruvoipati R. et al; CESAR trial collaboration. Efficacy and economic assessment of conventional ventilatory support versus extracorporeal membrane oxygenation for severe adult respiratory failure (CESAR): a multicentre randomised controlled trial. Lancet 2009; 374 (9698) 1351-1363
• 2 Combes A, Hajage D, Capellier G. et al; EOLIA Trial Group, REVA, and ECMONet. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med 2018; 378 (21) 1965-1975
• 3 Ramanathan K, Shekar K, Ling RR. et al. Correction to: extracorporeal membrane oxygenation for COVID-19: a systematic review and meta-analysis. Crit Care 2021; 25 (01) 375
• 4 Urner M, Barnett AG, Bassi GL. et al; COVID-19 Critical Care Consortium Investigators. Venovenous extracorporeal membrane oxygenation in patients with acute covid-19 associated respiratory failure: comparative effectiveness study. BMJ 2022; 377: e068723
• 5 Ohshimo S, Liu K, Ogura T. et al; Japan ECMO Network. Trends in survival during the pandemic in patients with critical COVID-19 receiving mechanical ventilation with or without ECMO: analysis of the Japanese national registry data. Crit Care 2022; 26 (01) 354
• 6 Nunez JI, Gosling AF, O'Gara B. et al. Bleeding and thrombotic events in adults supported with venovenous extracorporeal membrane oxygenation: an ELSO registry analysis. Intensive Care Med 2022; 48 (02) 213-224
• 7 Willers A, Swol J, Buscher H. et al. Longitudinal trends in bleeding complications on extracorporeal life support over the past two decades-extracorporeal life support organization registry analysis. Crit Care Med 2022; 50 (06) e569-e580
• 8 Fanning JP, Weaver N, Fanning RB. et al; COVID-19 Critical Care Consortium. Hemorrhage, disseminated intravascular coagulopathy, and thrombosis complications among critically ill patients with COVID-19: an international COVID-19 critical care consortium study. Crit Care Med 2023; 51 (05) 619-631
• 9 Martucci G, Giani M, Schmidt M. et al; International ECMO Network (ECMONet). Anticoagulation and bleeding during veno-venous extracorporeal membrane oxygenation: insights from the PROTECMO study. Am J Respir Crit Care Med 2024; 209 (04) 417-426
• 10 Willers A, Swol J, van Kuijk SMJ. et al. HEROES V-V-HEmorRhagic cOmplications in Veno-Venous Extracorporeal life Support-Development and internal validation of multivariable prediction model in adult patients. Artif Organs 2022; 46 (05) 932-952
• 11 Schmidbauer ML, Ferse C, Salih F. et al; On Behalf Of The Ignite Study Group. COVID-19 and intracranial hemorrhage: a multicenter case series, systematic review and pooled analysis. J Clin Med 2022; 11 (03) 605
• 12 Kim HK, Tantry US, Smith Jr SC. et al. The East Asian paradox: an updated position statement on the challenges to the current antithrombotic strategy in patients with cardiovascular disease. Thromb Haemost 2021; 121 (04) 422-432
• 13 Richardson S, Verma A, Sanaiha Y. et al. Racial disparities in outcomes for extracorporeal membrane oxygenation in the United States. Am J Surg 2023; 225 (01) 113-117
• 14 Yusuff H, Zochios V, Brodie D. Thrombosis and coagulopathy in COVID-19 patients rceiving ECMO: a narrative review of current literature. J Cardiothorac Vasc Anesth 2022; 36 (8, Pt B): 3312-3317
• 15 Mansour A, Flecher E, Schmidt M. et al; ECMOSARS Investigators. Bleeding and thrombotic events in patients with severe COVID-19 supported with extracorporeal membrane oxygenation: a nationwide cohort study. Intensive Care Med 2022; 48 (08) 1039-1052
• 16 Shaefi S, Brenner SK, Gupta S. et al; STOP-COVID Investigators. Extracorporeal membrane oxygenation in patients with severe respiratory failure from COVID-19. Intensive Care Med 2021; 47 (02) 208-221
• 17 Arachchillage DJ, Rajakaruna I, Scott I. et al. Impact of major bleeding and thrombosis on 180-day survival in patients with severe COVID-19 supported with veno-venous extracorporeal membrane oxygenation in the United Kingdom: a multicentre observational study. Br J Haematol 2022; 196 (03) 566-576
• 18 Li C, Cai T, Xie H. et al. Risk factors and outcomes for patients with bleeding complications receiving extracorporeal membrane oxygenation: an analysis of the Chinese Extracorporeal Life Support Registry. Artif Organs 2022; 46 (12) 2432-2441
• 19 Malas J, Chen Q, Shen T. et al. Outcomes of extremely prolonged (> 50 d) venovenous extracorporeal membrane oxygenation support. Crit Care Med 2023; 51 (07) e140-e144
• 20 Thomas J, Kostousov V, Teruya J. Bleeding and thrombotic complications in the use of extracorporeal membrane oxygenation. Semin Thromb Hemost 2018; 44 (01) 20-29
• 21 de Oliveira GLV, Oliveira CNS, Pinzan CF, de Salis LVV, Cardoso CRB. Microbiota modulation of the gut-lung axis in COVID-19. Front Immunol 2021; 12: 635471
• 22 Nichi-Iko Pharmaceutical Co. L. Pharmaceutical interview form for FUTHAN 10 INJ., FUTHAN 50 INJ. 6th ed. [Internet]. 2019 . Accessed August 10, 2024 at: https://www.nichiiko.co.jp/medicine/file/31050/interview
• 23 Hernández-Mitre MP, Tong SYC, Denholm JT. et al. Nafamostat mesylate for treatment of COVID-19 in hospitalised patients: a structured, narrative review. Clin Pharmacokinet 2022; 61 (10) 1331-1343
• 24 Minakata D, Fujiwara SI, Ikeda T. et al. Comparison of gabexate mesilate and nafamostat mesilate for disseminated intravascular coagulation associated with hematological malignancies. Int J Hematol 2019; 109 (02) 141-146
• 25 Doi K, Ikeda M, Hayase N, Moriya K, Morimura N. COVID-UTH Study Group. Nafamostat mesylate treatment in combination with favipiravir for patients critically ill with Covid-19: a case series. Crit Care 2020; 24 (01) 392
• 26 Yoshioka T, Daizumoto K, Tada K. et al. Retroperitoneal hemorrhage in a patient with coronavirus disease 2019 (COVID-19): a case report. J Med Invest 2022; 69 (1.2): 148-151
• 27 Doi S, Akashi YJ, Takita M. et al. Preventing thrombosis in a COVID-19 patient by combined therapy with nafamostat and heparin during extracorporeal membrane oxygenation. Acute Med Surg 2020; 7 (01) e585
• 28 Hunsicker O, Beck L, Krannich A. et al. Timing, outcome, and risk factors of intracranial hemorrhage in acute respiratory distress syndrome patients during venovenous extracorporeal membrane oxygenation. Crit Care Med 2021; 49 (02) e120-e129
• 29 Fletcher-Sandersjöö A, Thelin EP, Bartek Jr J. et al. Incidence, outcome, and predictors of intracranial hemorrhage in adult patients on extracorporeal membrane oxygenation: a systematic and narrative review. Front Neurol 2018; 9: 548
• 30 Arachchillage DRJ, Passariello M, Laffan M. et al. Intracranial hemorrhage and early mortality in patients receiving extracorporeal membrane oxygenation for severe respiratory failure. Semin Thromb Hemost 2018; 44 (03) 276-286
• 31 Fletcher-Sandersjöö A, Thelin EP, Bartek Jr J, Elmi-Terander A, Broman M, Bellander BM. Management of intracranial hemorrhage in adult patients on extracorporeal membrane oxygenation (ECMO): an observational cohort study. PLoS One 2017; 12 (12) e0190365
• 32 Taniguchi H, Ikeda T, Takeuchi I, Ichiba S. Iliopsoas hematoma in patients undergoing venovenous ECMO. Am J Crit Care 2021; 30 (01) 55-63
• 33 Taniguchi H, Rätsep I, Heinsar S. et al. Iliopsoas haematoma during extracorporeal membrane oxygenation: a registry report from the COVID-19 critical care consortium across 30 countries. Perfusion 2024; 39 (05) 891-895

Publikationsverlauf

Eingereicht: 18. Juni 2024

Angenommen: 02. September 2024

Accepted Manuscript online:
06. September 2024

Artikel online veröffentlicht:
27. September 2024