Extracorporeal life support (ECLS) is increasingly being used in the setting of non-neonatal cardiac and respiratory failure. 1–8 The implementation of this technique often requires the interhospital transport of critically ill patients, sometimes over considerable distances. These transports depend upon the availability of advanced medical transport teams and have a high risk of mortality, even in the most experienced hands. Early in our experience with ECLS for non-neonatal respiratory failure, several patients died from hypoxemia during attempted transport or immediately upon arrival to our hospital. In addition, some patients who were candidates for ECLS because of cardiac dysfunction were unable to be transported because of hemodynamic instability. It became obvious that, for some patients, ECLS would have to be initiated before transport to our facility. Therefore, we developed a protocol for the ECLS assisted transport of patients with severe respiratory or cardiac failure, and in this article, we present our experience to date.
The technique of ECLS and the management strategies used at our institution have been previously described. 1–3,5 A number of modifications to the existing ECLS protocol and equipment were required to allow transport on extracorporeal support, and these are outlined below.
Patient referral was initiated by telephone from the referring physician. If the patient met ECLS criteria, an air transport team was dispatched to the referring center for evaluation. This team consisted of two flight nurses, a respiratory therapist, and an ECLS surgical fellow. After patient evaluation at the referring hospital, this team consulted with the attending ECLS surgeon regarding the safety of conventional transport. In most cases, the patient was able to be hemodynamically stabilized with conventional therapy and achieve adequate gas exchange on a transport ventilator (T-Bird VSO2, Bird Products, Palm Springs, CA), limiting peak inflation pressure (PIP) to ≤ 45 cm H2O and obtaining SaO2 ≥ 90%. Conventional transport was subsequently performed via ambulance, helicopter, or fixed-wing aircraft, depending on vehicle availability, distance, and weather.
If the patient could not be sufficiently stabilized to allow conventional transport, the ECLS transport protocol was initiated. Before departing for the outside facility, a number of prerequisites had to be met, including the following: (1) coordination of transportation to and from the outside facility, (2) assembly of all personnel necessary for transport, (3) preparation of an extracorporeal circuit and supplies needed for cannulation, and (4) notification of a surgical team to be available at the outside facility with blood products and operating room supplies (Appendix 1). Given the logistic complexity of these transports, it was not uncommon for several hours to be needed to prepare for a transport. During that delay, the ECLS team continued to provide management suggestions to the referring physician.
Vehicle Selection and Equipment Configuration
The choice of transport vehicle (ambulance, helicopter, or fixed-wing aircraft) was based on crew and vehicle availability, transport distance, and weather. Essential equipment for all vehicles included a working electrical inverter capable of providing uninterrupted 110 V alternating current, at least two oxygen ports (one for the ventilator, one for sweep gas to the artificial lung), and a means to secure all equipment to prevent motion during travel. Regardless of the chosen mode of transportation, it was often not possible to accommodate the flight crew, ECLS specialists, and physicians in a single vehicle. Once the patient had been cannulated and placed on bypass, crowding was an even larger issue. Therefore, the coordination of transport between facilities often included finding alternative means of getting all crew and equipment home.
Most patients were transported by ground. For these cases, a custom built critical care ambulance was used. Unique features of this vehicle include a seating capacity for six crew members in back and two in front, two 4 kW electrical generators, a 110 V electrical invertor, 4 M-tanks of oxygen with both Ohio and Chemtron gas connectors, an 800 lb capacity hydraulic lift, and a 250 gal diesel fuel tank allowing extended range. The interior of the ambulance is large enough to accommodate most routine and emergency patient care for extended periods of time.
The aircraft initially used for personnel transport was an Aerospatiale Twinstar helicopter. More recently, a Bell-230 helicopter has been used for this purpose. The large interior of this aircraft accommodates most adult patients supported on the ECLS transport circuit. The helicopter is equipped with 4 D-tanks of liquid oxygen and a 110 V power supply. A medically configured Cessna Citation V jet aircraft was used for transports > 250 miles and required the use of a commercial flight service (Marlin Air, Willow Run, MI). Ambulance transport to and from the airport was required. Equipment and personnel were the same as used for helicopter transport.
For all transports, a combination of ECLS personnel (in all cases, two physicians and two ECLS specialists to prime and operate the circuit) and support personnel were required. For ground transport, this typically included two paramedics (one driver and another to assist with patient loading, etc.) and at least one flight nurse to coordinate non-ECLS patient care. For air transports, the support crew consisted of one pilot and two flight nurses.
The major equipment and supplies used during ECLS transport are listed in Appendix 2. The standard ECLS circuit was modified to minimize its physical dimensions and weight. The circuit was mounted on a platform above the patient and secured to the base of a stretcher, which allowed the movement of the patient and circuit as a single unit (Figure 1). In more recent cases, especially during air transport, a centrifugal blood pump (Sarns Delphin II, 3M Corp., Ann Arbor, MI) was used in place of a roller pump (Cobe Roller Pump, Cobe Cardiovascular, Arvada, CO) to simplify and miniaturize the circuit. The centrifugal pump provides active suction, eliminating the need for gravity drainage. Potential disadvantages of the centrifugal pump are hemolysis if suction is excessive, reverse flow if power fails, and difficult manual operation. A Kolobow designed spiral coil solid silicone rubber membrane lung (Avecor Cardiovascular, Plymouth, MN) was used for gas exchange, with the particular model chosen based on patient size. Microporous devices were avoided because of the risk of air embolus and plasma leakage during air transport. A battery pack was included to run the pump during periods of disconnection from the vehicle power supply. The remaining equipment consisted of support supplies, drugs, specialized surgical instruments, gas regulators, cannulae, and extra materials necessary for priming the circuit. These supplies were placed into large travel packs for easy transport.
Patient Management Protocols
The majority of the initial patients were cannulated for venoarterial support because it was thought that the additional cardiac support allowed better hemodynamic control and safety during transport. After some experience, venovenous bypass was attempted in patients with primarily respiratory failure, and no complications were observed related to this method of support. The recent approach has been to use the mode of support that would be used if the patient were being placed on bypass at our institution, based on the diagnosis and physiologic status of the patient at the time of cannulation. Particular care was taken during patient loading and unloading from the ambulance as well as during travel to and from the intensive care unit (ICU). During this time, one member of the team had the sole responsibility of holding the ECLS cannulae to prevent inadvertent removal. Upon arrival to our center, a member of the ECLS team met the ECLS transport team with a fully charged battery pack and a full tank of oxygen for transport to the ICU. The remainder of the patient management during transport was conducted according to previously described protocols. 1–3,5
Consent and Hospital Privileges
Informed consent was obtained from members of the immediate family of all patients before the initiation of bypass. To eliminate the need for the acquisition of temporary privileges at each hospital, and to remove responsibility for ECLS related patient care from the referring physician, an “umbrella” policy was developed by the University of Michigan legal staff. For all licensure, privilege, and liability purposes, this policy allowed the ECLS staff to act as if they were at our institution, rather than in another hospital or even another state. Following initial patient assessment, the referring physician was asked to sign orders indicating that full responsibility for the patient had been transferred to the physicians at the University of Michigan.
Between May 1990 and January 1999, 100 patients (68 adults and 32 children) were placed on extracorporeal life support at a referring institution and then transported to our center. These patients represent 15.1% of the non-neonatal cases transferred from referring hospitals for ECLS evaluation, and 12.7% of our non-neonatal ECLS cases over the same time period (Figure 2). The mean age (± standard deviation) of adult patients was 37.8 ± 10.5 years, and the mean age of pediatric patients was 46.4 ± 67.2 months. Diagnoses included adult respiratory distress syndrome (n = 78), cardiac failure (n = 7), sepsis (n = 7), asthma (n = 5), respiratory distress syndrome (of newborn) (n = 2), and airway compromise (n = 1). Pre-ECLS physiologic data are demonstrated in Table 1. Before the initiation of bypass, all adult patients were conventionally ventilated, with FiO2, 0.98 ± 0.05; PIP, 48 ± 13 cm H2O; and positive end-expiratory pressure (PEEP), 15 ± 6 cm H2O. Twenty of the thirty-two pediatric patients (62.5%) were conventionally ventilated, with FiO2, 0.97 ± 0.06; PIP, 52 ± 13 cm H20; and PEEP, 12 ± 7 cm H20. The remainder of these patients were managed with either high-frequency oscillatory ventilation (n = 8, 25%) or hand bagging (n = 4, 12.5%). Parameters of gas exchange for the entire population included PaO2,64 ± 36 mm Hg; PaCO2,49 ± 19 mm Hg; and alveolar-arterial oxygen gradient [(A-a) DO2], 602 ± 45 mm Hg. Overall, patients were moderately acidemic, with a mean arterial pH value of 7.28 ± 0.1.
Patients were supported on venovenous bypass in 53 cases and venoarterial bypass in 47 cases. The mode of transport was by ground ambulance in 80 cases, helicopter in 15 cases, and fixed-wing aircraft in 5 cases, and is shown by year in Figure 3. The range of ECLS assisted transports is demonstrated in Figure 4 and included a median transport distance of 44 miles (range 2–790 miles). The median total transport time was 5 hours and 30 minutes (range: 1 h 33 min to 16 h 6 min), and the median time spent at the referring hospital was 3 hours and 10 minutes (range: 55 min to 14 h 16 min).
All 100 patients were transported to our hospital safely after successful cannulation, with a subsequent mean total bypass time of 231 ± 212 hours. Three additional patients died at the referring hospital before cannulation was accomplished. Two of these patients died of hypoxemia or acidemia induced cardiac arrest between the time that the decision to transport the patient on ECLS was made and the beginning of cannulation. The third patient experienced cardiac arrest as a result of a perforated right atrial appendage after attempted percutaneous cannulation of the right internal jugular vein. Despite conversion to venoarterial bypass, thoracotomy, and repair of the atrial tear, the patient could not be resuscitated. Transport related complications occurred in 17 cases (Table 2) and were most commonly related to temporary failure of the electrical power supply to the circuit. There were 10 episodes of failure of the ambulance power supply and 4 incidents of subsequent failure of the ECLS circuit battery, which resulted in temporary hand cranking of the circuit pump until the original power supply was restored. Circuit tubing or connector leakage occurred in three cases. There was one case each of membrane lung thrombosis (after the initiation of bypass but before transport), membrane lung leakage, circuit tubing rupture, and hyperventilation caused by excessive membrane lung surface area, which required blending of the ventilating mixture with carbogen (95% O2, 5% CO2). In no patient did these complications adversely effect outcome.
Transport and survival data are listed in Table 3. Of the patients transported on ECLS, 69 (44/68 adult, 25/32 pediatric) were successfully weaned from extracorporeal support, and 66 (41/68 adult, 25/32 pediatric) survived to discharge. Bypass was discontinued for futility in 31 patients (for poor neurologic function in 17 patients, poor prognosis of primary disease in 9 patients, uncorrectable congenital heart disease in 3 patients, and intractable ventricular arrhythmia in 2 patients). Of the 69 patients successfully weaned from bypass, 3 (all adults) died, before discharge, of multiple organ failure resulting from the following diseases: blastomycosis (n = 1), pneumococcal sepsis (n = 1), and urosepsis (n = 1).
We have been using ECLS to manage severe cardiac and respiratory failure in adults and children since 1988. The most recent reports of our experience have documented a 54% and 73% survival in adult and pediatric respiratory patients, respectively. 1,5 Between 1988 and 1990, we received 107 non-neonatal referrals for ECLS evaluation. Eleven (10.3%) of these patients either died during conventional transport or were denied transport because of cardiorespiratory instability. The issue of mortality during transport of patients with severe respiratory failure was examined by Boedy et al.9 in 1990. They reviewed the outcome of neonatal respiratory management at their institution during a 52 month period in which they received 167 referrals for ECLS. Of the patients in this referral groups, 46 group died, and 18 of these deaths were related to transport. The authors recommended earlier referral of these neonates.
In an attempt to more safely manage these critically ill patients, some have advocated combining the techniques of critical care transport and extracorporeal life support. Previous series have demonstrated the ability to transport patients on ECLS. We transported two patients on ECLS from New Mexico to California early in our experience. 10 In 1991, Cornish et al.11 reported a series of 13 neonates with respiratory failure who were placed on bypass at referring hospitals. Eleven of these patients were successfully transported using both military aircraft (n = 10) and ambulance (n = 1) with four long-term survivors. Heulitt et al.12 also reported a series of 11 successful neonatal ECLS transports, with 9 long-term survivors. Another series described the ECLS assisted helicopter transport of 13 neonates and 1 adult. 13 In 1994, Bennett et al.14 reported the long-term survival of two of five adult patients with cardiorespiratory failure after transport on bypass. Rossaint et al.15 reported the successful transport of seven adults and one child after stabilization on ECLS, with six patients surviving to discharge. In the present study, we have demonstrated the ability to safely transport both pediatric and adult patients with severe respiratory or cardiac failure using ECLS.
Previous reports have suggested the possibility of safe ECLS assisted transport in neonates, but we have rarely used ECLS transport in this patient population. There has been a significant decrease in the number of neonates placed on ECLS in recent years, which is likely due to improvements in neonatal respiratory care such as vasodilation with nitric oxide, high-frequency oscillatory ventilation, and exogenous surfactant administration. 16–18 In addition, referral patterns for neonatal intensive care have become very efficient, and we typically receive these patients in transport before they require the initiation of ECLS. Our neonatal transport team is experienced in transporting unstable patients quickly and safely using either conventional ventilation or hand bagging techniques. These techniques are less effective in the unstable pediatric and adult population, and the movement of these patients usually requires more time, resulting in the more frequent occurrence of an unsafe transport situation and the need for ECLS assistance. For these reasons, ECLS transport has existed largely in the non-neonatal population at our institution. Progress in conventional adult transport, such as the T-Bird VSO2 ventilator (Bird Products, Palm Springs, CA), which is both compact (15 kg) and able to deliver pressure control and inverse-ratio ventilation, may expand the number of adult and pediatric patients who can be transported using conventional means. 19 In fact, our yearly ECLS transport experience peaked in 1996, the year before we acquired the T-Bird ventilator.
In general, referrals are received from a 250 mile radius around our institution. Referring physicians are advised to follow our protocol for severe respiratory failure, which includes diuresis, prone positioning, and pressure controlled inverse ratio ventilation. 20 With this regimen, many referral patients are stable enough to transport on a ventilator. If arterial hypoxemia (SaO2 < 90%) cannot be achieved using acceptable end inspiratory pressures (EIP ≤ 40 cm H2O), or if the patient is hypotensive despite inotropes and pressors, then the decision is made to initiate bypass before transport. When alveolar ventilation is the major problem (i.e., status asthmaticus), ECLS is utilized if PaCO2 is over 100 mm Hg or if there is evidence of compromising air-leak or severe barotrauma. In cases of primary cardiac failure, patients are candidates for ECLS transport if they cannot maintain a stable mean arterial blood pressure ≥ 60 mm Hg despite pressors and inotropes. The decision to send the ECLS team is usually made before transport, based on the criteria above. Occasionally the ventilator transport team determines at the referral hospital that the patient is too unstable, and the ECLS team follows for transport on ECLS.
Once the decision to use ECLS assisted transport is made, organization and preparation are of the utmost importance. The potential risks inherent in ECLS are magnified when one leaves the familiar environment of the home institution. The use of a strict protocol ensures an organized approach. Facsimile transmissions are used to rapidly convey information regarding operating room supplies, personnel, and blood requirements to the referring institution. A well trained team of two surgeons, two ECLS specialists and two flight nurses is essential; each member assumes an important role during each phase of the transport. Patients were initially cannulated for venoarterial bypass so that enhanced support capabilities were available in the event of hemodynamic compromise while patient access was limited. After some experience, this approach was found to be unnecessary and venovenous bypass was used, unless cardiac support was otherwise indicated by the particular pathophysiology. In most cases, this allowed the use of a percutaneous approach to cannulation, except in children under 3 years of age. The percutaneous approach appeared to save time, result in less cannulation site bleeding, and reduce the complexity of the cannulation process.
Complications of this technique include both those inherent to ECLS and those specific to transporting a patient on cardiopulmonary bypass. One death occurred during cannulation, as a consequence of a perforated right atrial appendage. This event underscores the need for the presence of experienced personnel both at the surgical and ECLS specialist level during transport related cannulations, as the staff at the referring institution may have little or no experience with the techniques and potential emergencies associated with the initiation of bypass. For this reason, an ECLS trained staff surgeon, an ECLS surgical fellow, and two ECLS circuit specialists are always present during cannulation. Once bypass has been successfully initiated, the two most potentially serious transport related complications are ECLS circuit failure/disruption and inadvertent decannulation. These can only be avoided by careful attention to details and proper training of the ECLS team. We stress the use of only team members in the loading and unloading of patients in the ambulance and in patient transport to and from the ICU. During these times, particular attention is focused on the cannulae/tubing to ensure security and avoid snagging on obstacles and on the roller pump to ensure continued function. Transport team members also need to be experienced with ECLS emergency management techniques. The fact that none of the complications occurring during transport had an adverse effect on outcome reflects the skill of the ECLS specialists in both avoiding and handling circuit related emergencies.
The institutional cost of ECLS transport ranges from $10 (gasoline for a local ambulance transport) to $20,000 (fuel cost for two fixed-wing aircraft and all personnel), depending on the method of accounting. All the equipment, personnel, and transport vehicles (except fixed-wing aircraft) are already owned or paid for by the University of Michigan, whether transport takes place on a given day or not, so the actual new cost is that of the fuel and disposables. However, when the total costs are amortized over the number of transports of all types, we estimate that the institutional cost of ECLS transports is in the range of $10,000.
Survival to discharge for this group of patients was 66% (60% of adults, 78% of pediatric patients), which compares favorably with the overall ECLS survival at our institution of 52% and 70% in adult and pediatric respiratory failure groups, respectively. This relatively high survival rate is interesting considering the fact that these patients were the most acutely ill at the time of bypass initiation. Overall survival was not elevated by the inclusion of seven patients with primary cardiac failure, as this group had a 57.1% survival rate. We could speculate that transport on ECLS provides better cardiorespiratory support during transfer, thereby limiting the magnitude of tissue hypoxemia experienced during this period and improving the chance for eventual survival. However, it is also possible that the patients with acute, fulminant processes who require transport on ECLS are the same patients who have a relatively rapid and complete resolution of their disease processes. Although the potential for complications is great, we have been pleased with the overall results in our initial 100 patients. We continue to encourage our colleagues to refer patients with severe cardiorespiratory failure earlier in their course to allow transport with conventional support and avoid the need for ECLS assisted transport. However, we have realized that properly selected patients who are at high risk for transfer can be safely transported on ECLS with a high survival rate.
The authors thank the flight crew and dispatchers of the University of Michigan Survival Flight Service and the ECLS specialists for their crucial role in the development of an ECLS transport protocol at our institution.
1. Kolla S, Awad SS, Rich PR, et al: Extracorporeal life support for 100 adult patients with severe respiratory failure. Ann Surg 226: 544–566, 1997.
2. Pranikoff T, Hirschl RB, Steimle CM, et al: Efficiency of extracorporeal life support in the setting of adult cardiorespiratory failure. ASAIO J 40: 339–343, 1994.
3. Anderson HL III, Delius RE, Sinard JM, et al: Early experience with adult extracorporeal membrane oxygenation in the modern era. Ann Thorac Surg 53: 553–563, 1992.
4. Anderson HL, Attori RJ, Custer JR, et al: Extracorporeal membrane oxygenation for pediatric cardiopulmonary failure. J Thorac Cardiovasc Surg 99: 1011–1021, 1990.
5. Moler FW, Custer JR, Bartlett RH, et al: Extracorporeal life support for severe pediatric respiratory failure: An updated experience 1991–1993. J Pediatrics 124: 875–880, 1994.
6. Green TP, Moler FW, Goodman DM, et al: Probability of survival after prolonged extracorporeal membrane oxygenation in pediatric patients with acute respiratory failure. Crit Care Med 23: 1132–1142, 1995.
7. Green TP, Timmons OD, Fackler JC, et al: The impact of extracorporeal membrane oxygenation on survival in pediatric patients with acute respiratory failure. Crit Care Med 24: 323–329, 1996.
8. Gattinoni L, Pesenti A, Bombino M, et al: The role of ECLS in the management of the adult respiratory distress syndrome. New Horiz 1: 603–612, 1993.
9. Boedy RF, Howell CG, Kanto WP: Hidden mortality rate associated with extracorporeal membrane oxygenation. J Pediatr 117: 462–464, 1990.
10. Bartlett RH, Gazzaniga AB, Fong SW, Jeffries MR, Roohn VR, Haiduc N. Extracorporeal membrane oxygenator support for cardiopulmonary failure: Experience with 28 cases. J Thorac Cardiovasc Surg 73: 375–386, 1977.
11. Cornish JD, Carter JM, Gerstmann DR, et al: Extracorporeal membrane oxygenation as a means of stabilizing and transporting high risk neonates. Trans Am Soc Artif Intern Organs 37: 564–568, 1991.
12. Heulitt MJ, Taylor BJ, Faulkner SC, et al: Inter-hospital transport of neonatal patients on extracorporeal membrane oxygenation: Mobile-ECMO. Pediatrics 95: 562–566, 1995.
13. Faulkner SC, Taylor BJ, Chipman CW, et al: Mobile extracorporeal membrane oxygenation. Ann Thorac Surg 55: 1244–1246, 1993.
14. Bennett JB, Hill JG, Long WB III, et al: Interhospital transport of the patient on extracorporeal cardiopulmonary support. Ann Thorac Surg 57: 107–111, 1994.
15. Rossaint R, Pappert D, Gerlach H, et al: Extracorporeal membrane oxygenation for transport of hypoxemic patients with severe ARDS. Br J Anaesth 78: 241–246, 1997.
16. Carter JM, Gerstmann DR, Clark RH, et al: High frequency oscillatory ventilation and extracorporeal membrane oxygenation in the treatment of acute neonatal respiratory failure. Pediatrics 85: 159–164, 1990.
17. Kennaugh JM, Kinsella JP, Abman SH, et al: Impact of new treatments for neonatal pulmonary hypertension on extracorporeal membrane oxygenation use and outcome. J Perinatol 17: 366–369, 1997.
18. Wilson JM, Bower LK, Thompson JE, et al: ECMO in evolution: the impact of changing patient demographics and alternative therapies on ECMO. J Ped Surg 31: 1116–1123, 1996.
19. Bartlett RH, Nelson K, Wagner C: Interfacility transport of patients with acute respiratory failure. Air Med J 5: 25–27, 1998.
20. Rich PB, Awad SS, Kolla S, et al: An approach to the treatment of severe adult respiratory failure. J Crit Care 13: 26–36, 1998.
Copyright © 2002 by the American Society for Artificial Internal Organs