|Complications (Ref. #)||Incidence and Prevalence||Risk Factors||Risk of Mortality|
|Vascular (Online Refs. 1 and 2)||Reported prevalence of 20% to 30%.|
Limb ischemia more commonly reported, prevalence as high as 40%.
Hyperemia is less common, but prevalence is estimated to be around 10% to 20%.
Absence of distal perfusion catheter
Ipsilateral femoral arterial and venous cannulation
Axillary cannulation commonly associated with hyperemia
|Approaches 60% in some series with limb ischemia|
Less clear with hyperemia
|Neurological (Online Ref. 3)||Broad range of neurological complications have been associated, ranging from subclinical cognitive impairment, seizures, paraplegia, peripheral neuropathy, compartment syndrome, ischemic and hemorrhagic strokes, and death.|
Highly variable due to lack of standardized reporting criteria.
Adult VA-ECMO patients have incidence rate of 13.3% for all neurological complications, and 5.9% to 7.8% for ischemic and/or hemorrhagic stroke.
Imaging findings of neurological injury has been reported in nearly 50% of patients.
|Solid or gaseous microemboli and thrombosis within cannula|
Duration of ECMO support
Hemostatic imbalance between procoagulants and anticoagulants
|Nearly 90% with ICH|
|Infection (Online Ref. 4)||Bloodstream infections have reported prevalence of 3% to 18% and incidence of 2.98 to 20.55 episodes per 1,000 ECMO days in adults.|
Lower respiratory tract infections incidence is reported at 24.4 episodes per 1,000 ECMO days.
Prevalence of urinary tract infections is reported between 1% to 2%, and incidence is reported to be 1 to 13.8 cases per 1,000 ECMO days.
History of autoimmune disease
Higher SOFA score
Duration of ECMO support
|38% to 63%|
|Hemolysis (Online Refs. 5 and 6)||Improved incidence with newer pump designs|
Reported incidence between 5% to 18% (Online Ref. 6).
Plasma free hemoglobin ≥100 mg/l was observed in nearly 67% of adults, and prevalence of severe hemolysis or thrombosis requiring circuit changes was noted to be 8.9% among adults (Online Ref. 5).
Need and duration of continuous renal replacement therapy
History of inflammatory disease
Hypovolemia or inadequate preload
Technical complications (cannula malposition, kinking, excessive centrifugal pump speeds, among others)
|Associated with higher risk of mortality (∼32% for those with plasma free hemoglobin ≥100 mg/l) (Online Ref. 3)|
|Renal failure (Online Refs. 2,7–10)||Data limited by variable definitions of AKI across the studies; reported incidence between 33% to 55.6%.|
No significant difference in AKI incidence with type of cannulation.
Prevalence of post-ECMO HD is reported between 28% and 52%.
|Age >70 yrs|
Pre-operative serum creatinine >2 mg/dl
Comorbidities (diabetes, obesity, cerebrovascular accident)
Thoracic aorta repair
Incomplete sternum closure
Bleeding and hemolysis
Sepsis and DIC
|Overall hospital mortality 20% to 65%|
1-yr post-HT survival of 52.3% for those with eGFR <45 ml/min/1.73 m2 or on HD
|Bleeding (Online Refs. 11–16)||Highly variable due to lack of standard definitions.|
Prevalence is 30% to 56%.
10 events per 100 ECMO days.
Common sites are thorax, GI tract, and cannula site.
|ECMO causes qualitative and quantitative platelet defects, destruction of large von Willebrand factor multimers, and fibrinolysis|
Number of anticoagulation levels above target range
Platelet count <50,000/μl mm
HAT score (1 point each for hypertension, age >65 yrs, and VA-ECMO type) predicts bleeding; increasing score predicts increasing transfusion requirements, especially for platelets and fresh frozen plasma
|Higher mortality associated more with number of red blood cell units transfused than bleeding itself|
|SIRS (Online Refs. 17 and 18)||Some degree of systemic inflammation occurs in most ECMO recipients and in about 30% after decannulation.||Infection|
Duration of ECMO
Age (very young and very old)
|Not enough data to determine (less significant in absence of sepsis)|
|Quality of life (Online Refs. 19–21)||In general, mental and physical activity are satisfactory but not normal.|
In 24 adult ECMO survivors using EQ-5D, physical activity was more impaired than mental function, and mental issues were 2 to 3 times more common than in normal subjects.
In 28 long-term adult ECMO survivors (median follow-up of 11 months), 36-Item Short Form Health Survey scores were significantly lower than matched healthy controls for physical role, general health, and social functioning, but higher than those reported for patients on chronic HD, with advanced HF, or after recovery from acute respiratory distress syndrome.
In 67 patients who survived ECMO for CS post-myocardial infarction, HRQOL was evaluated after median follow-up of 32 months. Mental health was satisfactory but persistent physical and emotional-related difficulties were reported: 34% with anxiety, 20% with depression, and 5% with PTSD symptoms.
Factors for improved quality of life include younger age and nonischemic disease
References for Table 5 can be found in the Online Appendix. Figures in table reprinted with permission from Tramm R, Ilic D, Sheldrake J, et al. Recovery, risks, and adverse health outcomes in year 1 after extracorporeal membrane oxygenation. Am J Crit Care 2017;26:311–9.
AKI = acute kidney injury; DIC = disseminated intravascular coagulopathy; GI = gastrointestinal; HD = hemodialysis; HF = heart failure; HRQOL = health-related quality of life; HT = heart transplant; ICH = intracranial hemorrhage; PTSD = post-traumatic stress disorder; SIRS = systemic inflammatory response syndrome; SOFA = Sequential Organ Failure Assessment; VV-ECMO = venovenous extracorporeal membrane oxygenation; other abbreviations as in Tables 1 and 3.