Qatar Medical Journal - 1 - Extracorporeal Life Support Organisation of the South and West Asia Chapter 2017 Conference Proceedings, February 2017

Bleeding is a common complication in patients undergoing extracorporeal membrane oxygenation (ECMO) management.¹ It requires immediate management to achieve hemostasis, replace blood, and compensate volume loss. Refractory hemorrhage can be lethal and can lead to massive transfusions with all their known complications. Refractory bleeding and massive transfusions in ECMO patients are associated with high mortality even after decannulation. Management of bleeding in ECMO patients requires thorough evaluation with multi-disciplinary approach that addresses surgical causes of bleeding, correction of coagulopathy, and the balanced use of anticoagulation factors to prevent circuit clotting, avoid excessive bleeding, and replace different blood products as needed. Adjusting anticoagulants and the use of fresh frozen plasma (FFP) with correction of thrombocytopenia can control common bleeding events. Management of refractory hemorrhage may require exploration for surgical bleeding and administration of platelets, packed red blood cells (pRBCs), cryoprecipitate, anti-fibrinolytics, and selective coagulation factors. In some cases, however, the bleeding is diffuse and cannot be controlled surgically. The use of activated factor VII (rFVIIa) at different described doses for patients on ECMO with refractory bleeding has been tried.^2,3 There are many reports indicating successful use with live-saving outcome.³ Unfortunately, there are also some conflicting results with the use of rFVIIa regarding failure to control bleeding or the risk of intravascular thrombosis or circuit clotting. Furthermore, there are reports about catastrophic outcome or fatal thrombosis when rFVIIa was used in ECMO cases.⁴ Therefore, the medication is currently recommended as off-label prescription. It should be used with extreme caution with clear patient/family awareness about potential complications. The recommended doses are not established and range from 24 to 174 μg/kg.² Some centers will administer lower doses of rFVIIa (25–50 μg/kg) and, if more than one dose is required, it is not administered more often than every 2–4 hours. Some centers recommended the use of prothrombin complex concentrate (PCC), which contains unactivated factors II, VII, IX, and X, and therefore, potentially have less risk of thrombosis. To correct a prolonged prothrombin time (PT) and activated partial thromboplastin time (APTT) during ECMO run in patients with active bleeding, PCC 25-50 international units/kg can be administered.⁵

In summary, bleeding during ECMO remains a serious problem, which increases mortality risk. rFVIIa has been used successfully but with awareness of the risk of thrombosis. Clinical trials comparing alternative anticoagulation regiments are needed to determine efficacy, dosing, and safety of rFVIIa in patients suffering from refractory bleeding while on ECMO. Unfortunately, until such evidence is available, the ECMO care team is left with few evidence-based interventions to prevent and treat serious bleeding.

This presentation will discuss the use of rFVIIa in ECMO patients with focus on its benefits in controlling refractory bleeding, and the risk of thrombosis and circuit clotting associated with it. The discussion will include suggestions for recommended doses, how to monitor for thrombosis, and the potential risk/benefit of using rFVIIa in the management of life-threatening bleeding in patients on ECMO.

The application of veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is becoming an increasingly frequent procedure in emergency medicine.¹ Correct appraisal of clinical conditions and patient selection are critical not only in terms of expected outcome, but also in terms of adequate temporary support configuration, optimization of resources, and organization of daily hospital activities, since inappropriate implant may remarkably impact all these aspects.

According to recent analysis of ECMO use in the USA,¹ VA-ECMO was most commonly applied in post-cardiac surgery candidates, however with a constant decline for the period from 2007 to 2011. Indeed, from 57% of the total ECMO cases, post-cardiotomy VA-ECMO went to 38%, but with a significant rise in cases of cardio-pulmonary failure (from 3.9 to 11.1%).¹ This situation is certainly further changing during the last 5 years as VA-ECMO in cardiogenic shock and cardiac arrest is becoming more popular. Cardiac arrest: Patient selection in this setting is rather variable. However, potential cardiac etiology (acute myocardial infarction – AMI) represents almost 50% of the causes, and witnessed cardiac arrest, immediate cardio-pulmonary resuscitation (CPR) by bystanders, adequate CPR (possibly with mechanical cardiac massage devices), and not prolonged time from cardiac arrest to ECMO application ( < 60 min) represent positive predictors in the setting of extracorporeal cardio-pulmonary resuscitation (ECPR).² Recent publications have shown that ECPR might achieve respectful outcome also in elderly patients.^3,4,5 Therefore, age should not disqualify potential candidates. Absence of severe acidosis and still acceptable end-tidal CO2 may represent additional elements for patient selection in this setting. Post-cardiotomy: ECMO results in this setting appear rather unsatisfactory with in-hospital survival below 30%.⁵ Prompt initiation of ECMO and avoiding unnecessary prolonged cardio-pulmonary bypass and metabolic as well as coagulation derangement are among the most relevant key factors. The absence of severe cardiac or co-morbidities prior to surgery and the expected temporary impairment of the cardio-respiratory systems represent additional indications for post-cardiotomy ECMO. Cardiogenic shock: This setting represents the most expanding field of ECMO application. Indications may derive from refractoriness of cardio-circulatory impairment to conventional therapies (drug and IABP). Moreover, in this setting, advanced age and severe co-morbidities may represent relative contraindications. Furthermore, in this field, the potential for cardiac recovery or the possibility to bridge the patient to other treatments, if recovery is considered unlikely, represents indication for ECMO use. Respiratory distress: VA-ECMO is generally indicated if right ventricular dysfunction is present or develops, and if VV-ECMO does not provide sufficient peripheral oxygenation. In the last case, hybrid configuration (VVA or VAV) might be more suitable and advisable. Uncommon indications: Sepsis, trauma (most often treated with VV-ECMO), fulminant myocarditis, Tako-Tsubo, or complicated interventional procedures (TAVI, PCI, balloon valvuloplasty) represent areas in which ECMO is increasingly applied.^6,7 Despite wider use in these settings, more clinical evidences are still needed to conclusively define patient selection criteria and ultimate benefits. Finally, a developing indication might be prophylactic VA-ECMO in critically ill patients undergoing either cardiac surgery or interventional cardiological procedures. In these patients, in whom post-procedural severely complicated course is expected, timely and pre-crash support may help to provide a smooth peri-procedural period. Short-time controlled support of cardio-circulatory system and peripheral organ perfusion, with avoidance of metabolic and hemodynamic derangement, may also affect emergency VA-ECMO implant should patient deterioration occur, and allow a better complicated patient management with better final outcome.

Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) is increasingly applied for the treatment of cardiogenic shock despite its high complication rate.¹ The lack of left ventricular unloading is one of the serious problems associated with the poor outcome of VA-ECMO. Therefore, hemodynamic management during VA-ECMO should address the higher afterload caused by the retrograde blood flow and the consequent left ventricular distension. In fact, the blood stasis can result in ventricle or pulmonary thrombosis. Moreover, a high end-diastolic pressure can cause pulmonary venous congestion and lung injury, as well as subendocardial malperfusion and consequently impair recovery.

Possible strategies to unload the left ventricle include inotropic support or intra-aortic balloon pump implantation, as described in 135 cases by Gass and colleagues.² Surgical left ventricle venting can be performed with the cannulation of the left atrium or the left ventricle although this strategy is highly invasive. Blade atrial septostomy or atrial septostomy and placement of a venting cannula are also described.^3,4

Our group recently described a new strategy employing Impella on top of VA-ECMO in a large series of patients, compared with VA-ECMO only.⁵ Impella device is a small heart pump that pulls blood from the left ventricle through an inlet area near the tip and expels blood from the catheter into the ascending aorta. The device was inserted percutaneously through the femoral artery into the ascending aorta, via the aortic valve into the left ventricle. In compliance with the Declaration of Helsinki and in agreement with Italian and German data protection laws, we retrospectively collected data on patients with severe refractory cardiogenic shock from two tertiary critical care referral centers and enrolled 157 patients (January 2013 to April 2015): 123 received VA-ECMO support and 34 had concomitant treatment with VA-ECMO and Impella implanted simultaneously. The decision for an additional implantation of Impella was undertaken as the attending physician recognized signs of echocardiographic, radiological, and clinical signs of impaired left ventricle unloading or left ventricle stasis (stone heart, pulmonary edema, impending clotting on the left ventricle, significant aortic regurgitation). Impella was left running at P8 speed in order to produce a forward flow of 2.0 L without complications. A propensity-matching analysis was performed in a 2:1 ratio, resulting in 42 patients undergoing VA-ECMO alone (control group) compared with 21 patients treated with VA-ECMO and Impella. Patients in the VA-ECMO and Impella group had significantly lower hospital mortality (47% vs. 80%, P <  0.001) and a higher rate of successful bridging to either recovery or further therapy (68% vs. 28%, P <  0.001) compared with VA-ECMO patients. Other results are presented in Table 1.

In conclusion, among different strategies to unload the left ventricle during VA-ECMO, Impella can be considered a feasible option. Nevertheless, randomized studies are warranted to validate this strategy.

The use of mechanical circulatory assistance remains controversial during refractory septic shock in adults.¹ However, a profound myocardial dysfunction can occur during bacterial septic shock.^2–4 In this context, extracorporeal membrane oxygenation (ECMO) is highly effective as salvage therapy for children with refractory septic shock.^5,6 We reported the largest cohort of adults who received VA-ECMO for refractory cardiovascular dysfunction in the context of severe bacterial septic shock.⁵ Despite multiorgan failure at ECMO initiation and simplified acute physiology score (SAPS) 3-predicted mortality of 79%, >70% of these patients survived with complete recovery of cardiac function. The hemodynamic profile we describe herein (low cardiac index, elevated filling pressure, profound myocardial depression, and elevated systemic vascular resistance) is certainly a rare entity in the spectrum of septic shock, which resembles that of the classic paradigm of cardiogenic shock. In this setting, the infusion of very high catecholamine doses used to increase cardiac output and maintain perfusion before ECMO initiation might have contributed to the vicious circle that led to vasoconstriction and refractory cardiovascular failure. Considering the reversibility of myocardial depression associated with septic shock,^7,8, we hypothesized that ECMO could help salvage these dying patients by restoring adequate perfusion to vital organs to reverse multiple organ failures and by buying time to achieve infection control by antibiotics. Indeed, all our survivors could be explanted without recourse to cardiac transplantation or switching to another cardiac assist device, and all recovered with a normal myocardial function within a few weeks. These results seem to be far better than those obtained with ECMO for cardiogenic shock, with reported survival rates around 40%, that required cardiac transplantation or switching to a LV (left ventricular) assist device for about 10% of the survivors.⁹ Considering these promising results, ECMO might be considered a valuable therapeutic option for patients with refractory cardiovascular dysfunction in the context of septic shock, although more data and larger patient cohorts are needed to confirm the findings presented herein. Ethical statement: In accordance with the ethical standards of our hospital's institutional review board, the Committee for the Protection of Human Subjects, informed consent for demographic, physiologic, and hospital-outcome data analyses was not obtained because this observational study did not modify existing diagnostic or therapeutic strategies. Survivors gave oral consent to participate in the telephone interview, conducted by the same investigator, who asked the questions in the questionnaire in the same order.

Background: Mortality of patients on extracorporeal membrane oxygenation (ECMO) remains high. Diagnosis of infection during extracorporeal life support (ECLS) is still challenging, and prevention strategies vary widely from center to center.^1–3 These facts led us to analyze the occurrence rate, site, and organism in our ECLS patients in order to implement infection control measures to reduce the incidence of infections during ECLS.⁴

Our objective was to analyze our Extracorporeal Life Support Organization (ELSO) registry center data specifically focused on incidence of infection, typical microorganisms, time of manifestation, and site of cultures in the settings of tertiary pediatric cardiac intensive care unit mainly utilizing transthoracic cannulation and VA-ECMO, and compare with the ELSO database.¹Methods: We conducted a retrospective study analyzing 25 neonatal and pediatric ECMO cases in relation to infection from January 2014 to December 2015, in comparison to the ELSO database age and modality specific data. We obtained ethical approval from our institution. We examined the prevalence of infection, the time of the first positive cultures, the site of the positive cultures, and the underlying microorganisms and compared with ELSO data whenever feasible. Results: There is no specific data on the incidence of infection in the ELSO database with open chest/transthoracic cannulation; our incidence was 0.44. The Candida species was the highest offending organism (24% vs. ELSO 12% concerning the entire ECMO population), followed by Klebsiella 20%, E. coli 16%, and Pseudomonas 12%. The first positive culture was taken on the 8th day of ECMO (median). By site, the highest prevalence of infection is as follows: ventilator-associated pneumonia (VAP), 41%, followed by bloodstream infection (BSI), 22%, and then catheter-associated urinary tract infection (CAUTI), 12%. Conclusions: The highest prevalence of Candida infections is most probably due to the combined antibiotic and steroid therapy for patients with capillary leak syndrome. This may prompt that routine antifungal prophylaxis can be added after 1 week of ECMO for this patient group. Alternatively, the early detection with fungal polymerase chain reaction (PCR) assay should be evaluated.⁵ The high occurrence of VAP may indicate the need of reinforcing enteral feeding, oral decontamination protocol along with VAP bundle, and investigation of alternative source of contamination. As Gram-negative Enterobacteriaceae and Pseudomonas were in second line as typical multidrug-resistant (MDR) organisms, those should be covered whenever a need for empiric antibiotic therapy arises.

Veno-arterial extracorporeal membrane oxygenation (ECMO) has been used successfully for several years in refractory cardiogenic shock. Of note, the survival rate is markedly affected by the underlying patient condition, and especially their capacity to recover. Combes et al.¹ demonstrated a short- and a long-term survival rate (11 months of median follow-up time) of, respectively, 42 and 36% in a large series of patients with cardiogenic shock of various origins, including fulminant myocarditis (30%), post-cardiotomy (24%), post-myocardial infarction (15%), and shock after heart transplantation (15%). In that study,¹ implantation during cardiopulmonary resuscitation (CPR) was associated with a marked increase in the risk of death (OR 20.68 (1.09–392.03)).

As the probability of achieving return to spontaneous circulation (ROSC) decreases rapidly when the duration of cardiopulmonary resuscitation CPR exceeds 10 min and dramatically after 30 min,² some clinicians questioned whether ECMO could also be used to restore flow in patients not responding to classical resuscitation procedures. Data of ECMO for out-of-hospital cardiac arrest (OHCA) or extracorporeal CPR (ECPR) are scarce and conflicting. Several case reports and small series suggested some benefits but these were subject to publication bias. The first large study of 51 patients with OHCA showed a very low survival rate, with only two survivors.³ Recently, some other teams reported more favorable outcomes. In 42 patients with OHCA, Kagawa et al.⁴ reported a 30-day survival of 24%, with a favorable neurological outcome of 21% in the group of OHCA patients who received intra-arrest percutaneous coronary intervention (PCI). One of the explanations for the variability of the results is the duration from cardiac arrest to ECMO initiation. Leguen et al.³ reported a 4% of survival with a time to ECMO of 120 min, while studies with more favorable outcomes reported a shorter time to ECMO (40 min [25–51]) in the trial by Kagawa et al.,⁴ and systematic implementation of intra-arrest PCI. Indeed, the difference in outcome between in-hospital and OHCA results depends more on the duration of cardiac arrest than on the location of the cardiac arrest itself.

Importantly, all series mentioned that selection of candidates for ECPR seems to be crucial, especially in the setting of OHCA: witnessed cardiac arrest, with a no-flow period < 5 min, whenever possible bystander CPR or a very rapid response team, a pre-hospital care policy prompting to alert a specialized hospital with an ECMO team, and rapid transport after initiation of CPR (scoop and run strategy). In addition, good quality CPR should be provided from the start up to the initiation of ECMO flow.^5,6

In conclusion, when used, all efforts should be made to minimize the time from cardiac arrest to ECMO flow, as the latter is a critical determinant of outcome. Organ donation might be considered in patients with poor neurological outcomes, but experiencing full recovery of organ function.

Aim: Refractory cardiogenic shock (CS) complicates 5–7% of cases with ST-elevation myocardial infarction (STEMI), and is a leading cause of hospital death after myocardial infarction.¹

CS complicating acute myocardial infarction (AMI) continues to have a high mortality of 40–50% despite early revascularization and adjunctive therapies.² Extracorporeal membrane oxygenation (ECMO) technology has advanced significantly and is readily available at the bedside. This is a viable option for short-term support in the setting of acute cardiac ischemia. According to the 2003 USA National Registry of cardiopulmonary resuscitation (CPR), in-hospital cardiac arrest has a poor prognosis with an overall survival to hospital discharge rate of 17% with conventional CPR.³ One of the most common causes of cardiac arrest is ventricular fibrillation (VF) secondary to ischemia, which carries an improved prognosis if successfully defibrillated, with the rate of survival to hospital discharge being 34%.³ In cases with refractory ischemic VF, definitive therapy with percutaneous coronary intervention (PCI) may not be possible without anoxic brain injury secondary to hemodynamic collapse. CPR was introduced in the 1960s as a lifesaving method in patients with cardiac arrest.³ To supplement CPR in select patients, ECMO is used successfully for witnessed in-hospital cardiac arrest.³ In the setting of an AMI, bridging to a revascularization procedure is important in improving neurological outcome and overall survival. We report the profile and the outcome of patients in refractory VF resistant to defibrillation on ECMO support. Subsequent to revascularization, the patient's cardiac rhythm converted back to sinus rhythm with a single defibrillation shock with excellent neurological recovery. Methods: Since January 2014, we have been reviewing patients who had suffered from progressive severe refractory CS post STEMI undergoing emergency PCI on percutaneous veno-arterial (VA)-ECMO support. Results: For 11 male patients (mean age 50 ± 18 years), the mean duration of support was 7 ± 4 days. Of these patients, 9 (81%) were weaned successfully from ECMO. However, two patients on ECMO support died: one due to massive gastrointestinal bleeding and the other due to septic shock. Three other patients also died; one due to occluded stent on third day post-ECMO removal, one due to intracranial hemorrhage on second day post-ECMO removal, and one due to septic shock on fourth day post-ECMO removal. The 30-day survival was 54% (6/11 patients) without any neurological deficit. Conclusion: VA-ECMO has shown to be an option to bridge patients in CS and/or refractory VF to allow for a successful revascularization procedure and ultimately good neurological and survival outcome.

Background: Extracorporeal cardiopulmonary resuscitation (ECPR) is the rapid deployment of extracorporeal membrane oxygenation (ECMO) – or cardiopulmonary bypass – to provide immediate cardiovascular support for patients who have cardiac arrest unresponsive to conventional cardiopulmonary resuscitation (CPR) measures.

There is improved survival with isolated cardiac lesions.¹ Cardiac disease (adjusted for confounding factors) was associated with improved survival when compared with non-cardiac diseases (odds ratio 6.3, 2.01–19.80).²

Conventional CPR versus ECPR has a lower survival to discharge 8.2–22% and about 6–11% for critically ill patients. The survival of out-of-hospital cardiac arrest is less than 3%.³ The long-term survival is 53% with ECPR versus 17% with conventional CPR.⁴ ECPR in witnessed in-hospital cardiac arrest in areas of advanced life support system and effective CPR with single organ dysfunction with minimum time elapse in logistics like ECPR in cath lab is associated with much better patient outcome and revival to hospital discharge. Methods: Procedural support for angioplasty, arrhythmia ablation, pulmonary embolectomy, and bypass surgery are few examples of crash down situations, which are better managed with ECPR. A cath lab is the best place for application of ECMO in a short time. The equipment consists of the ECMO circuit with a centrifugal pump, hollow fiber oxygenator, heat exchanger, back up battery, 3/8 inch venous quick prime tubing, arterial tubing, and percutaneous arterial and venous cannulas. This is a study of 16 cases of ECPR done in a cath lab for witnessed adult cardiac arrests. The decision to initiate ECPR was done in 5 min with circuit priming within 20 min, and simultaneous cannulation performed in 15 min by another team. Results: Overall, 16 patients with cardiac disease over a period of 3 years were included in this study. The age group varied from 35 to 70 years. There were 12 males and 4 females. Six patients had poor left ventricle (LV) with heart failure, who were undergoing bypass surgery. Seven patients had acute myocardial infarction (MI) with cardiac arrest, who were considered for primary angioplasty (PAMI). Two patients had malignant arrhythmias (post-viral) and one patient had pulmonary embolism. There was 8/16 (50%) survival at least 24 h after ECMO decannulation and 5/16 (33%) survival to hospital discharge. Two patients could not be weaned off ECMO support. The most common cause of death was ischemic brain injury. All the survivors had favorable neurological outcome. Two patients had CPR of 60 min prior to ECPR. Pre-arrest factors associated with non-survival were persistent hypotension and renal insufficiency. Conclusions: ECPR promotes survival with ECMO application. Pre-ECMO quality of resuscitation will influence success percentage. Functional outcomes in survivors were reasonable with few derangements, particularly neurological impairments. All procedures were uncomplicated following ECMO application.

Acute myocardial infarction is a common cause of cardiogenic shock (approximately 75% of all patients) and out-of-hospital cardiac arrest (approximately 70% of survivors).¹ Mechanical hemodynamic support is employed prior to coronary revascularization (pre-percutaneous coronary intervention (PCI)), during or after PCI.² Revascularization procedure is characterized by a transient interruption of coronary blood flow (due to repetitive contrast dye injections, balloon inflations, atherectomy passes, and stent manipulations) resulting in a negative inotropic effect. Percutaneous left ventricular assist device implantation (Impella and TandemHeart), and intra-aortic balloon pump implantation have been described as strategies to avoid the worsening of cardiac function during PCI in the literature, especially in high-risk patients. The USpella registry has shown that pre-PCI implantation of IMPELLA 2.5 significantly improves survival of cardiogenic shock patients (Figure 1).³

Cardiac arrest is a recognized complication in the cath lab during percutaneous procedures, such as valve interventions, left auricle closure, and vascular interventions in addition to PCI. Extracorporeal cardiopulmonary resuscitation (ECPR) in terms of VA-ECMO plays a role in rescue therapy for cardiac arrest⁴ with a better outcome, when compared with conventional cardiopulmonary resuscitation (CPR), when CPR is failing.⁵ However, the technical and logistical possibility to implement ECPR in the cath lab is challenging. First, to deal with an emergency strategy out of intensive care and operative theater. In this context, trained personnel and dedicated sets of instruments and drugs could play a role. Second, the fluoroscope limits the free access to the patient. Moreover, percutaneous VA-ECMO cannulation requires a complex approach as the vascular accesses have been violated in most cases during the procedure. On the other hand, the presence of a multidisciplinary team has to be considered as an important resource.

In conclusion, an adequate cardiac support during cath lab procedures should be planned whenever possible to avoid emergencies especially in high-risk patients. Training and local protocols should be provided to overcome the procedural difficulties of ECPR.

The rapid institution of veno-arterial extracorporeal membrane oxygenation (VA ECMO) support for patients with prolonged, recurrent cardiac arrest (CA) complicated by severe shock and cases of refractory arrhythmia without return of spontaneous circulation (ROSC) is now termed ECMO-CPR (E-CPR). The use of E-CPR is increasing and there are reported benefits for both out-of-hospital and in-hospital patient populations. Recently, this service has been provided by staff from the hospital Emergency Department. It is likely that the best model of service provision for E-CPR depends on local factors.

Currently, there is no unifying or accepted definition for E-CPR¹ and many case series reports include both patients with and without ROSC^2,3. From analyses of outcomes from cardiac arrest data (without the use of ECMO), the likelihood of recovery from prolonged CA beyond 20 min is negligible for both in- and out-of-hospital populations, and this seems to be a reasonable threshold for the classification of ECMO-CPR.

There is a strong correlation in case series between the CPR time and the survival from ECMO with long-term survival without neurological injury becoming rare if CPR has extended beyond 60 min.

The SAVE-J study is the best evidence for the use of E-CPR for out-of-hospital cardiac arrests. This large prospective (non-randomized) cohort study showed that good neurological outcome at 6 months occurred in 12.3% of patients who received E-CPR and in 1.5% of patients who did not have access to ECMO following prolonged CA.² Propensity studies of retrospective case series suggest that neurological injury (but not survival) may be improved by the use of E-CPR in the in-hospital setting⁴. Attempts to delineate risk prediction algorithms for in-hospital cardiac arrest have consistently identified advanced age and an initial cardiac arrest rhythm other than ventricular tachycardia (VT) or ventricular fibrillation (VF) as highly predictive of poor outcome.

E-CPR should be predominately used for younger patients with VT/VF arrests, where there is access to early coronary investigation and intervention and the time to ECMO can be less than one hour.⁵ While temperature control is vital in the first 36 h following cardiac arrest, hypothermia is no longer considered beneficial. Partial pressure of oxygen and carbon dioxide may have effects on neurological outcome following cardiac arrest, but more investigation is currently underway. E-cpr Is Not The Same As “Fast” Va Ecmo: The technical skills, equipment preparedness, staffing, and staff training required for an E-CPR program are different from those required for non-CA ECMO initiation. An E-CPR program should work in concert with medical emergency teams that provide early patient assessment for deteriorating patients.

Cardiac arrest is a common problem, both in the in-hospital environment and in the out-of-hospital environment with an estimated incidence of over 300,000–400,000 cases of out-of-hospital cardiac arrests every year in the USA. Globally, the outcome of cardiac arrest remains poor, particularly with respect to neurological outcomes.¹ It is estimated that approximately 10–11% of patients with cardiac arrest survive to hospital discharge. Approximately 8–9% of patients survive to a good neurological outcome with a cerebral performance class of 1 to 2 (surviving to living at home with no or minimal support for activities of daily life). Survival is more likely if cardiac arrest has been witnessed by bystanders and cardiopulmonary resuscitation (CPR) has been commenced immediately. Survival is also more likely if the cause of cardiac arrest is ventricular arrhythmia, either ventricular fibrillation or ventricular tachycardia, and if early defibrillation is available. Given the poor outcomes from cardiac arrest, there has been a growing interest in using novel approaches to try and improve the outcomes, particularly the neurologically intact outcomes from cardiac arrest. One of the potential therapeutic modalities which has arisen in the last decade has been the use of extracorporeal membrane oxygenation (ECMO) for patients with cardiac arrest (E-CPR).² Although there have been a number of retrospective and prospective uncontrolled studies, or for some studies propensity matched with a current or historical control, observational studies have demonstrated potential benefits associated with ECMO.³ Currently, there are no randomized trials in the literature, although there are at least two trials currently being undertaken. Furthermore, at present, there is no widely accepted pathway for patients which provides recovery from cardiac arrest, commencement of E-CPR, management of the cardiovascular system and targeted therapy designed to protect the brain. E-CPR continues to pose significant challenges in patient management, both in its implementation and in the management of patients following cardiac arrest and resuscitation.^2,4 The management of patients is likely to be substantially different with E-CPR compared with conventional approaches, particularly following cardiac arrest. In particular, the management of the cardiac function is different in patients with retrograde aortic blood flow, and the ideal management of the brain, including mean arterial pressure, cerebral perfusion pressure, ideal targets for carbon dioxide and arterial oxygen content, has yet to be defined. Another key challenge is neurological prognostication and diagnosis of brain stem death, which is more challenging on ECMO.⁴ An additional area that requires careful consideration is temperature management, given the results of recent trials exploring the outcomes for patients managed with hypothermia compared with normothermia.⁵ The relevant literature will be covered and an approach to the management of the post-arrest patient, particularly with respect to considerations on ECMO, will be discussed.

Since its development in the early 1990s, extracorporeal membrane oxygenation (ECMO) has become a standard therapy for a wide variety of respiratory and cardiac problems in neonates. Traditionally, the preferred method of access and support has been using a venous inflow, with arterial outflow methodology (venoarterial or VA ECMO), bypassing the cardiac circulation.¹ Ongoing improvements in catheter configuration have made the use of a dual lumen catheter, with venous inflow from above and below the atrium and directed outflow of oxygenated blood into the right atrium, possible in the majority of neonates (venovenous or VV ECMO).^2,3 The advantages of VV ECMO include improved physiology with preservation of pulsatile blood flow, reduced cardiac stun, delivery of oxygenated blood to the pulmonary circulation, maintenance of normal cerebral perfusion, and reduced risk of emboli. However, only 30% of infants are supported using VV methodology, while up to 80% are potential candidates.¹ The reasons cited for choosing VA ECMO are patient size, instability, and the need for ongoing pressor support. Although these are concerns, none, except for patient size, are absolute indications for VA support. With appropriate cannulation methods, and verification of catheter position and function, the majority of neonates can be supported using VV ECMO. This article reviews the practical considerations for using VV ECMO in neonates. Patient selection is based on diagnosis and size. Almost all respiratory, septic, and most cardiac indications can be supported with VV ECMO if the patient weighs more than 2500 g. The two main catheter types differ in size and configuration; both have been shown to be efficient when properly placed. Catheter positioning during cannulation should be confirmed with an echocardiogram for flow and mixing, and a radiograph for positioning. The Avalon catheter must have an echocardiogram for placement and has a higher reported incidence of complications.⁴ Typically, a secondary venous inflow catheter is required, either as a cephalad jugular or as a femoral line. Medical management during a VV ECMO run commonly requires pressor management; however, this is also associated with a more stable perfusion profile. There are no other changes in management required; anti-coagulation requirements and decannulation are simpler with the VV ECMO methodology. Overall, outcomes (survival and neurological status) in all major disease classes are better with VV ECMO; however, controversy exists as to whether this is due to the therapy, or a selection bias in choosing candidates.¹ In summary, with appropriate preparation of the team and selection of catheter sizes, the majority of neonates can be supported with VV ECMO. This offers the potential for improvement in survival, and reduced incidence of neurological injury. These long-term outcomes should be the primary consideration in the choice of cannulation technique.

Introduction: Strategy during extracorporeal membrane oxygenation (ECMO) support is guided by available outcome data. Boundaries limiting selection of cases are shifting with advances in ECMO technology, conventional support, and destination therapies. Evolution: Neonatal respiratory ECMO numbers continue to decline due to improved conventional therapies, whereas infants historically excluded from ECMO (for example, ex-premature infants with O2-dependent chronic lung disease and severe RSV) are now successfully supported. ECMO for bridging of children to lung transplant is now established.¹ Advances are also being made in mobile ECMO,² long duration of ECMO support, and ECMO while wide awake. Initiation: Flow requirements and patient or vessel size govern cannula choice. Percutaneous cannulation of vessels with ultrasound guidance is described in all sizes of children, including small infants.³ Serial imaging and ECHO control are needed for optimisation of cannula position and orientation. Transoesophageal ECHO is preferable to transthoracic where available. Reconfiguration of the circuit may prove necessary in some cases and is better done early in the run.⁴Troubleshooting: Protocol and bundle adherence are the keys to preventing complications. Detection requires vigilance and rigorous checks along with comprehensive handover and tight teamwork. Management is about appropriate escalation and pathway utilisation. Simulation is an essential component of staff training and a powerful tool for reinforcing these points.⁵Lung recovery: ECMO can be considered the ultimate in “lung rest” techniques. Enhancing lung recovery during the rest provided by ECMO is mainly achieved by attention to fluid balance, pulmonary toilet, and adoption of ventilator rest settings. Bronchoscopy is also particularly useful in this context.⁶ Other therapies such as prone position, steroid use, surfactant, perfluorocarbon,⁷ and even individual lung ventilation² have a role in selected cases. De-cannulation: Patience can be needed when timing the end to VV ECMO support. Removing the sweep gas from the oxygenator can be done simply without having to wean ECMO flow. In more difficult situations, 12–24 h trial off can be used. Radiological lung clearance, improving lung compliance, reduction of sweep gas requirements, and oxygen challenges are all informative in the run up to a decision to trial off. Lung biopsy⁸ or genetic test results are useful when recovery is absent to help establish futility. Earlier de-cannulation can be indicated in the setting of an intractable or severe complication. Summary: Strategy for management of VV ECMO support is evolving with improvements in ECMO techniques and advances in supporting therapies. A proactive, rigorous but flexible approach to individual cases can deliver excellence.

For neonates, veno-arterial (VA) extracorporeal membrane oxygenation (ECMO) support is still commonly used for respiratory as well as for cardiac or combined failure. However, in the pediatric population, veno-venous (VV) ECMO is establishing itself as the standard mode of support for respiratory failure.¹

The need for ECMO in neonatal respiratory support has declined over the past years following introduction of alternative options. Conversely, a significant increase in VA ECMO for cardiac support has been identified (see Figure 1). The present article aims to describe the various indications for VA ECMO in the neonatal population.

Although the majority of cardiac runs for neonates still relates to congenital heart defects, widening experience and indications, such as myocarditis, sepsis, poisoning (reversible), or extracorporeal cardiopulmonary resuscitation (ECPR), have led to an increased use of VA ECMO in neonates and children.

For congenital heart defects, special considerations must be made, especially for ECMO in the context of single ventricle (SV) physiology:³ the balancing between pulmonary and systemic circulation remains crucial whether a Blalock–Taussig shunt or a Sano shunt is used. For the second and third stages of single ventricular palliation, cannulation strategies must be adapted to the underlying anatomy on a case-to-case basis.

Myocarditis due to various reasons can be supported successfully with VA ECMO, which constitutes the ultimate endpoint of the myocarditis management algorithm. Timely deployment before irreversible multi-organ damage has occurred is crucial. Further attention must be paid that myocardium can be quite stunned and left ventricular decompression is mandatory to allow adequate myocardial recovery.⁴

Sepsis has become an indication if conventional management fails,⁵ as reflected in the algorithm published by the Surviving Sepsis Campaign in 2012, which recommends starting ECMO in refractory shock.

Case reports with successful support of heart and lung function until recovery for various poisonings or during cath lab interventions have been published.⁶

ECPR with the deployment of extracorporeal life support (ECLS) during resuscitation has been mentioned in the PALS guideline since 2010 as class II recommendation. As could be expected, outcomes for survival and neurological deficits are related to centers’ experience and resuscitation time prior to ECLS installment.⁷

VA support remains common in neonates and children placed on extracorporeal membrane oxygenation (ECMO) for respiratory, cardiac or combined dysfunction.¹ Providing adequate ECMO flow to reverse tissue oxygen debt is imperative in the first few hours of ECMO. Several studies outline the poor outcome which results if lactic acidosis and pH do not improve after institution of ECMO. Following serial lactate, urine output, and other signs of organ perfusion are vital aspects of care.

For infants and children, the common route for adequate vessel size is use of the right internal jugular vein and right common carotid artery. Some centers also utilize a venous drainage cannula placed retrograde up the internal jugular vein to the level just below the jugular bulb to augment venous drainage. Whether this also reduces risk for venous congestion in the brain is unknown. Older children (usually at least 15 kg or 2–3 years old) may be supported via the femoral vessels (vein and artery) or a combination of cervical and femoral routes. There is some evidence to suggest that use of the carotid artery is associated with more risk for neurologic abnormalities such as seizures or intracranial infarction. For patients cannulated via the femoral vessels, care must be taken to avoid distal venous congestion in the extremity, which can lead to compartment syndrome. Similarly, distal lack of arterial perfusion can lead to ischemia and need for amputation in severe cases. Placement of a distal cannula which is then Y-ed into the arterial return from the ECMO circuit can provide perfusion to the limb. Neurovascular integrity of the lower extremities is important to prevent injury.²

Veno-arterial ECMO increases left ventricular afterload. This can result in sudden and severe left ventricle failure, leading to pulmonary edema or hemorrhage from back-up into the left atrium and pulmonary venous hypertension. Transmural ischemia from left ventricular distention can also occur. While afterload reduction or low-dose vasoactives to improve ventricular performance and maintain left ventricular output can be effective, some patients will require decompression of the left side of the heart. This can be done via a balloon septostomy across the atrial septum (allowing left atrial blood to then be drained into the right atrium and removed with the venous drainage cannula), by placement of a direct venting cannula into the left atrium, left ventricle, or pulmonary vein and then Y-ing this into the venous side of the ECMO circuit.³ Following left ventricle performance with ECHO and maintaining a pulse pressure of at least 10 mmHG following ECMO initiation are important aspects of VA care. For weaning during VA ECMO, decompression efforts must be ceased to allow adequate filling of the left ventricle. For post-cardiotomy patients placed on VA ECMO, early cardiac catheterization to identify correctible residual lesions should be performed. Failure to recover heart function within the first week of ECMO is associated with high mortality and discussion of listing for heart transplant or discontinuation should occur as appropriate.⁴

Extracorporeal membrane oxygenation (ECMO) is a commonly used form of mechanical circulatory support in children with congenital or acquired heart disease, and cardiac failure refractory to conventional medical treatments.¹ The cardiac output is partially or completely provided by the ECMO machine. To assess how much support is needed at any given time, a constant balance of oxygen delivery and oxygen consumption is required. Permanent monitoring of flow, hemoglobin, arterial, and mixed venous saturations are key elements to provide the needed understanding and continuous guidance of body requirements, aiming at controlling and maintaining optimal management of the ECMO patient.

Although ECMO is lifesaving in many circumstances, it bares challenges and certain risks of complications. Despite bioline-coated circuits, anticoagulation is required. Measurement of activated clotting time (ACT) and activated partial thromboplastin time (APTT) are both used to guide anticoagulation, but it is unclear which method is the best, and constant monitoring of the hemoglobin level allows the early detection of internal and external bleeding.² For children, sedation and analgesia, most often in combination with paralysis, is unavoidable for a smooth ECMO run. On the other hand, with such treatment on board, it is almost impossible to assess the neurological status of the patient on ECMO. Different monitoring tools, such as cerebral near infrared spectroscopy (NIRS) and serial brain ultrasound studies, are used to estimate brain integrity and function, but further research is needed to show which methods are the most practical and useful to achieve a favorable outcome.³ Nutrition in critically ill children is known to be crucial but debated how to achieve as the gut perfusion can be impaired and therefore the risk of necrotizing enterocolitis (NEC) is increased. It is discussed if somatic perfusion by NIRS monitoring can help to avoid NEC.⁴ The kidney function is often impaired in these critically ill children and careful monitoring of kidney function and urine output leads the decision when to start which type of renal replacement therapy for renal clearance and/or fluid removal. Techniques available include peritoneal dialysis or continuous renal replacement therapy (CRRT) incorporated with the ECMO circuit. Medication dosing is under discussion in many centers as the circulatory volume is higher than normal. Adjustment of medication dosing could consequently be of great importance. Close monitoring of medications where toxic levels can be reached is crucial to avoid further damage to the patient's organ functions.

Background: Extracorporeal membrane oxygenation (ECMO) is a rapidly expanding, life-supporting therapy for critically ill patients with severe respiratory and/or cardiovascular failure. Cerebrovascular impairment can result in hemorrhagic and ischemic complications commonly seen in the patients supported on ECMO. The healthy brain is protected by cerebral autoregulation, which maintains an adequate cerebral blood flow in face of blood pressure changes.¹ Pre-ECMO factors, such as hypoxia, hypercarbia, and hypertension, can disrupt blood flow regulation, leaving the brain vulnerable to changes in blood pressure.² Cannulation of large blood vessels³ and alterations of pulsatile flow patterns during ECMO also play a role in altered cerebral autoregulation.⁴

A reliable methodology that can assess the status of cerebral autoregulation during ECMO and provide early indication of neurological injury is critical for optimization of bedside management to improve clinical outcomes. Wavelet transform coherence (WTC) is a time–frequency domain analysis that characterizes the cross-correlation and relative phase between spontaneous fluctuations in blood pressure and cerebral oxygenation measurement by oximetry.

We implemented WTC⁵ to assess the degree of cerebral autoregulation impairment in neonatal and pediatric ECMO and evaluated its usefulness as an early predictor of acute neurological complications. Further, we examined cerebrovascular parameters, blood gas changes, and anticoagulation parameters as potential causes of autoregulation impairment during ECMO. Methods: Spontaneous fluctuations of mean arterial pressure (MAP) and cerebral tissue oxygen saturation (SctO2) were continuously measured during the ECMO run. The dynamic relationship between the MAP and SctO2 fluctuations were assessed based on wavelet transform coherence (WTC) to derive an index of cerebral autoregulation impairment. The Institutional Review Board at the University of Texas Southwestern Medical Center at Dallas approved the study. Results: A total of 25 neonatal (11) and pediatric (14) patients were studied. In-phase coherence between the MAP and SctO2 fluctuations was predominant in a time scale range of 8–32 min. Significant correlations between individual autoregulation indices and neuroimaging scores were found in both neonates and children. There was a significant association between individual blood pressure variations with autoregulation indices in the children, but not in the neonates (Figure 1). Discussion and conclusion: We found that intra-ECMO autoregulation impairments derived from WTC were apparent even before clinically observable changes occur at the bedside. Furthermore, these impairments correlated with the patients' neuroimaging abnormalities. This finding remained constant for both VA ECMO and VV ECMO in contrast to evidence of increased incidence of neurological complication with carotid artery cannulation in the literature. Blood pressure variability in ECMO patients appeared to be associated with impaired autoregulation in the non-neonatal population.

Continuous assessment of cerebral autoregulation based on WTC has the potential to be a useful bedside tool to predict acute neurological events in patients on ECMO. Our study shows that high blood pressure viability, for which these patients are at risk, appears to be a cause of cerebral autoregulation impairment. This finding suggests a new approach to bedside management that may lead to a decrease in cerebral autoregulation impairment, thereby improving neurological outcomes in these patients.

Survival of patients with malignancy over the last 30 years has continually improved due to advancements in aggressive chemotherapeutic regimens as well as supportive measures toward treatment and prevention of infection.^1,2 Even though cancer remains a significant cause of mortality in children, 5-year survival for childhood cancers, including hematologic malignancies and solid tumors, is greater than 80%. In these children, critical illness occurs either as a result of their malignancies and/or complications from treatment. Given improvements in survival, a more aggressive approach to the management of these complications in pediatric patients with malignancies has been encouraged, which has enlightened the medical community to the use of extracorporeal life support (ECLS) in this complex population. Evidence of this new approach was reported in 2009 by Gow et al.³ when a survey of ECLS centers regarding utilization of ECLS for patients with malignancy found that 95% of respondent ECLS centers felt that malignancy was not a contraindication to ECLS. This is further supported by the Extracorporeal Life Support Organization (ELSO) registry, which documents that, since 1997, the use of ECLS, including extracorporeal membrane oxygenation (ECMO) for pediatric cancer patients has been steadily rising.

Up until 2007, pediatric patients with malignancy comprised 0.5% of all ECLS patients. Limitation in the use of ECLS for patients with cancer was most likely influenced by cancer-related mortality as well as the ideology of ECLS being a support modality offered only to acutely ill “healthy” patients. Since 2009, a number of reports describing ECLS use for patients with malignancy have been published, implying that its use is increasing in this population most likely due to noted improved oncologic survival compounded by more widespread ECLS use in complex patient populations. Utilization of this support modality in pediatric malignancy, although still low compared with other populations, is increasing. In 2008–2012, the ELSO registry documented 178 patients with malignancy who received ECLS, which equates to 1% and a doubling of utilization compared with that reported previously. Although there is still a small portion of ECLS patients, education regarding the use of this modality of support for patients with malignancy will likely lead to further increases in utilization.^4,5

Background: Harlequin syndrome is a rare autonomic disorder, characterized by unilateral diminished sweating and flushing of the face in response to heat or exercise.¹

Harlequin syndrome is described in patients receiving peripheral veno-arterial extracorporeal membrane oxygenation (VA-ECMO), where differential oxygen saturation is observed between the upper and lower parts of the body.

It is a phenomenon related to cannulation, where upper body hypoxia occurs due to compromised arterial return despite an initial correct bifemoral cannula insertion. Case report: A 9-month-old child presented to the Royal Hospital, Oman, with a 1-day history of tachypnea and potential case of chocking, which rapidly progressed into ARDS and refractory respiratory failure despite providing maximum supportive measures: different modes of conventional mechanical ventilation and high-frequency oscillatory ventilator (HFOV), steroids, inhaled surfactant, inhaled iloprost, and prone positioning. Inhaled nitric oxide was not available. Rigid bronchoscopy excluded foreign body aspiration and echocardiography showed normal heart structure and function. As all measures failed and oxygen index remained low, ECMO was initiated. She was initially cannulated for VA-ECMO; two femoral venous cannulas of sizes 16F and 10F were inserted into both sides for adequate drainage and a size 8F femoral artery cannula was inserted into the right side; VV-ECMO cannulas were not available. The ECMO machine delivered a flow rate of about 80 ml/kg/min, BP 90/46, HR 130, CVP 14, Hb 12.6 g/dl, and milrinone 0.5 μg/kg/min. There was differential oxygen saturation between the upper and lower parts of the body; saturation was 50–60% in the upper part and 100% in the lower part, and FiO2 was 100%.

To solve the problem, VA-ECMO was changed from peripheral to central through insertion of a venous cannula into the right atrium and an arterial cannula in the aorta with a flow rate of about 110 ml/kg/min.

The child was successfully decannulated after 10 days of ECMO support and discharged home 2 months after admission. Respiratory viral panel was positive for adenovirus. Discussion: Harlequin syndrome is a rare complication of peripheral VA-ECMO. However, it can be as high as 8.8%.²

It occurs when the heart function is preserved or recovering, but the lungs are still poorly functioning, so the native cardiac output flows against the pumped blood, usually in the aortic arch region.² The reinfusion jet flows retrograde up the aorta and may meet resistance from antegrade flow generated by the left ventricle.

Depending on the amount of native cardiac function, the location of the interface between antegrade and retrograde flow will vary, and in circumstances where there is impaired native gas exchange with a significant amount of poorly oxygenated blood ejected from the left ventricle, the oxygenated reinfused blood may not reach the aortic arch.³ Subsequently, the coronary arteries, and to a variable degree the supra-aortic vessel as well, are provided with hypoxic blood, and the heart and brain can be affected.²

Therapeutic options consist of relocation of the arterial cannula into the right subclavian artery or aorta,² or converting to central VA-ECMO. It can also be solved by converting the system into a VA-V setting, where an additional return cannula may be added to the configuration with a “Y” connection off the femoral arterial reinfusion cannula, with insertion into an internal jugular vein.³Conclusion: Harlequin syndrome is a known complication of peripheral VA-ECMO, where the upper part of the body is poorly oxygenated. It occurs when the native heart function is preserved but the lungs are poorly functioning. Therapeutic options include converting to central VA-ECMO or VA-V-ECMO.

Extracorporeal membrane oxygenation (ECMO) is a form of mechanical circulatory support that can be lifesaving in people with potentially reversible heart or lung injuries. ECMO is nearly always used urgently, when all other treatment options for cardiopulmonary injury have failed and high mortality is otherwise expected. Standard ECMO treatment involves venous drainage from the femoral vein or left atrium with artificial extra-circulatory oxygen exchange. Return to the body is through the same veins (veno-venous) or arterial system via the femoral artery or ascending aorta (veno-arterial). Compared with cardiopulmonary bypass circuit, ECMO is transportable, smaller, closed to the atmosphere, and can treat a patient for several days to weeks.

Neurological consequences of severe respiratory failure and its different management strategies in adults are likely common but uncharacterized and poorly described in the reviewed literature.

Development of severe respiratory failure (SRF) occurs in 20–25% of patients with isolated severe traumatic brain injury (TBI) and is associated with a threefold increase in mortality, or patients remaining in a vegetative state.¹ It has been attributed mostly to aspiration, infection, neurogenic pulmonary edema, and release of pro-inflammatory mediators into the systemic circulation causing ultrastructural damage in type II pneumocyte.² This decreases the pulmonary tolerance of subsequent mechanical stress due to mechanical ventilation. Actually, some data suggest that the main feature of ALI/ARDS in brain-injured patients is the presence of a poor oxygenation (reduced PaO2/FiO2 ratio) accompanied by a moderate increase in the elastance of the respiratory system even though these patients had a normal chest X-ray.

Ventilatory support for such patients could be difficult with a lot of challenges to keep optimal oxygenation and acceptable level of blood carbon dioxide. It involves the application of positive end-expiratory pressure (PEEP) to recruit collapsed alveoli, improve arterial oxygenation, and reduce elastance of the respiratory system. Although improving oxygenation is a key factor for optimizing O2 delivery to the brain, clinical studies provide contradictory information on the use of PEEP in patients with acute lung injury (ALI) complicating severe brain injury.

Furthermore, we tend to use higher tidal volumes in patients with acute brain injury because mild hypocapnia is a key factor in the clinical management of raised intracranial pressure, which is a frequent abnormality in such patients;^3,4 this would be an injurious ventilator strategy and may present a further relevant inflammatory stimulus. Moreover, it has been shown that the use of high tidal volumes for the first 48 h after ICU admission is associated with the development of ventilator-induced lung injury.⁴

In addition, one of the recognized methods in improving oxygenation in SRF patients is the use of prone position, because it improves the lung mechanics and augments oxygenation. Studies have demonstrated that patients with unstable intracranial pressure (ICP) have higher ICP in the horizontal position sideways. Therefore, the recommended position for this patient cohort is a 30-degree head-up tilt combined with a straight head position.⁴

Furthermore, TBI patients who require ECMO support for their SRF will need full anticoagulation, as the circuit poses considerable derangement in the hemostatic system with increased platelet consumption and a higher risk of intracranial bleeding. There have been several reports of ECMO without anticoagulation therapy, but so far with no evidence or good literature to support this notion.⁵

Although some controversy and difficulties exist, it seems that multiorgan clinical approach instead of single-organ approach represents the optimal way in clinical management of patients with ALI/SRF and TBI.