Technology and Intervention in the Neonatal Intensive Care Unit
JOHN M. DRISCOLL, JR., AND CHARLES J. H. STOLAR
Over the past 15 years the Neonatal Intensive Care Unit at Columbia Presbyterian Medical Center has developed a system of medical care, transportation for sick mothers and infants, professional staff education, and performance evaluation for a regional network of New York area hospitals. During the past eight years, they have successfully introduced a new technology, extracorporeal membrane oxygenation (ECMO), to save the lives of a group of infants with very high mortality rates.
This case study describes the process of implementing the new technology, identifying factors facilitating success as well as limitations, and continuing unresolved problems. Implementation was aided by the basic effectiveness of the technology to save lives, which was not known beforehand, as well as by the dedication and skill of two pediatric surgeons. The process was hindered by continuing staff resistance related to lack of control over the technology and physical isolation from the work unit. Expanding the program as a regional resource continues to be limited by the unanticipated costs of the program coupled with a shortage of nursing staff, increased demand for other health services, and general budget constraints.
The Neonatal Intensive Care Unit (NICU) at Babies Hospital Columbia Presbyterian Medical Center (CPMC) is a super tertiary
care unit that serves a network of hospitals on the west side of Manhattan and several suburban hospitals in New Jersey, Connecticut, and north of New York City. A super tertiary unit provides a total range of medical services to infants and virtually never transfers an infant to another unit for a medical service. On the west side of Manhattan, there are other neonatal units that can provide certain intensive medical services, but require transfer for selected surgical and medical problems. In the suburban hospitals, sharply limited services are available and transfer to the NICU is therefore more frequent. The network involves provision of medical care, transportation of sick mothers and infants, education of professional staffs, and evaluation of performance.
The Neonatal Intensive Care Unit is the central component of the network and contains several units that provide care to infants currently in the hospital, as well as to new patients recently born either in the center or at a network hospital. Inherent in the network design is the guarantee of bed availability and transportation 24 hours per day. In addition to the requests for transfer from the network hospitals, there are requests from other hospitals for a specific service unique to the tertiary care unit, for example, pediatric cardiology or pediatric surgery.
Regionalization of care is essential to meet the needs of any area. In most large metropolitan areas requests for beds far exceed the capabilities of any one unit. In New York City there are six other medical schools, with units to help meet regional requirements. To establish realistic expectations with regard to regional needs, physicians identified hospitals that contacted their NICU regularly as a primary referral hospital. These hospitals form the base of the network, with approximately 25,000 births and with an additional 5,000 births at CPMC annually. Until the recent past, regional requirements have been met within the current distribution of 53 beds.
As a tertiary, university hospital, there is also an educational arm to the clinical service. Education includes medical, nursing, and paraprofessional, as well as graduate and postgraduate components. The education of house staff and postgraduate fellows in various subspecialties specifically requires the availability of a wide range of patients. As accreditation bodies have selectively reduced training programs, the survival of remaining residency programs is critically dependent on patient availability. This educational need creates yet another demand on the unit, that is, ''open'' admission for infants outside network hospitals with spe-
cific problems, such as pediatric cardiology. A hospital can have the best available program in a given subspecialty, but if the NICU continually rejects admissions because of a shortage of beds or staff the adverse impact is obvious. So, as a tertiary unit both hospital and regional needs must be met for newborn care as well as the specific needs of the different subspecialty services and their educational component.
The organization of the NICU rests in the hands of its director who is responsible to the chief of the Pediatric Service (education, policy) and to the president of the hospital (patient care, administration, and regionalization). The director establishes policies jointly with nursing colleagues that direct the provision of patient care, guarantee a standard of care, ensure effective flow of patients through the system, and meet educational goals of the university as well as regional needs. Thus, changes that affect the level of care at the bedside, subspecialty training, or standards at the network hospital flow through this combined medical/nursing hierarchy. The current team of neonatologists has been together for 15 or more years, which provides a continuity not available in other units.
The nursing staff is highly organized, very experienced, skilled, and demanding of itself and its medical colleagues. They form the backbone of the unit and are the primary implementers and guarantors of care. Their inclusion in the process, whether it be evaluation or change is essential to the success of the clinical service. The realized philosophy of joint professional responsibilities for administration of the unit is in large part responsible for a high retention rate of nurses (more than 60 percent with more than 5 years of service) and successful realization of established goals.
CPMC is a world renowned institution that has been totally supportive of the development of the NICU and the changes that have ensued over the past 20 years. They have placed the unit in the enviable position of being able to deliver superior tertiary care by providing the finances required to have the most sophisticated equipment and the manpower to operate the technology. In the new age of Diagnostic Related Groups in New York State, the financial burden for the hospital has increased geometrically, however, and fiscal constraints are being enforced.
In summary, the NICU is a critical part of a large metropolitan hospital as well as a regional network, with a busy clinical service that has a large obstetric service with approximately 5,000 deliveries annually, and pediatric subspecialties that jointly pro-
duce more than 700 annual admissions. It has a large medical staff with total subspecialty support and a nursing staff second to none. The unit also plays an important regional leadership role in the innovation and evaluation of care.
MOTIVATION FOR CHANGE
The area of our 53-bed operation that was most directly affected by the introduction of new technology was direct patient care in the 9-bed intensive care unit. The technology to be introduced was extracorporeal membrane oxygenation (ECMO). Introduction of the new technology occurred late in the historical evolution of the intensive care unit and late in the history of the current medical/nursing leadership. Fifteen to 20 years ago, significant technological and organizational changes related to the modernization of newborn intensive care had occurred. In the interval, changes were limited to modification of the technology and minor changes at the organization level related to the introduction of the regional concept of neonatal care. Introduction of ECMO was the first major change in patient care in at least 15 years.
The motivation for change was pure and simple. There was a group of infants, born after full pregnancy, who progressed to the point of respiratory failure and died. There was no successful treatment and these infants were the leading contributors to neonatal mortality among full-term infants. The two medical conditions that contributed most frequently to this mortality were infants born with a congenital absence of the diaphragm, so-called congenital diaphragmatic hernia, and infants born with a lung condition known as meconium aspiration, that progressed over the first 24 to 48 hours of life to the state of respiratory failure. In addition, a third condition, called primary pulmonary hypertension of the newborn, also resulted in respiratory failure and had a high mortality rate.
Historically, the pediatric surgeons were particularly frustrated by the postoperative course of the infant with a diaphragmatic hernia who initially did well, a "honeymoon period," only to die 24 to 48 hours later. They began to investigate alternative forms of therapy for such infants. Neonatologists had unsuccessfully tried different ways of ventilating infants and had tried pharmaco-
logic manipulations in an attempt to reverse the relentless respiratory failure of infants with all three conditions. The mortality rate with maximal conventional therapy was in excess of 90 percent. Five years ago all respiratory therapy was unsuccessful in saving this group of infants and professionals were continually frustrated by their inability to treat term infants in respiratory failure with these three conditions. Therefore, the motivation for change was related to the need to lower mortality and the availability of a new technology for which preliminary data suggested improved mortality and limited longitudinal morbidity.
Transitional circulatory difficulties are not new. In 1748 Catagon had observed, "The infant requires some intermediate time of abstinence and rest to compose and recover from the struggle of birth in the change of circulation." Newborn infants with intractable respiratory failure are unable to make the transition from the fetal to the newborn circulatory-respiratory system. ECMO is an aggressive life-support system that employs partial heart and lung bypass for long periods. It essentially enables a period of lung rest by providing an alternative source of gas exchange and oxygen delivery for an extended period of time during which the pathologic processes causing respiratory failure resolve of their own accord.
The concept of prolonged extracorporeal support is not new. Its modern history began 35 years ago with the description of mechanical oxygenation and perfusion by Gibbons. Subsequently, there have been modifications in the initial technology leading to today's current state of the art. In the mid-1960s the work of a number of investigators culminated in a National Heart, Lung, and Blood Institute study of the efficacy of extracorporeal membrane oxygenation for adult respiratory distress syndrome. This study failed to prove any advantage of ECMO over conventional ventilator management for adults afflicted with this disorder. However, in the interval, initially White (White et al., 1971) and subsequently Bartlett (Bartlett et al., 1977) began treating newborn infants with respiratory distress syndrome with ECMO. In 1976, Bartlett and his colleagues described the first series of ECMO-treated infants that had a significant number of survivors.
As previously mentioned, ECMO is a form of partial heart and lung bypass. The usual means of perfusion is by inserting a catheter to drain the venous blood from the right atrium and return-
ing it through another catheter to the ascending aorta. This is accomplished by a extrathoracic cannulation performed in the neonatal intensive care unit. Once the appropriate catheters are in place the infant then receives partial cardiopulmonary bypass with the membrane oxygenator assuming the function of the damaged lung and the infant's prior respiratory support gradually decreased to minimal levels. Just before initiation of ECMO, patients are treated with maximal conventional ventilator and pharmacologic support and believed to have an expected mortality rate between 85 and 95 percent. Once the infant has been stabilized on ECMO, this support is no longer necessary. The ventilator is generally turned to minimal settings that are sufficient to inhibit development of small areas of collapse in the lung and to maintain pulmonary toilet while the infant is on ECMO.
As a consequence, the management of the infant undergoing ECMO is considerably more straightforward than one might expect. In fact the condition of most acutely ill infants in the nursery quickly becomes stable. All intravenous lines are removed, leaving only an umbilical artery catheter for monitoring and sampling. All medications and intravenous nutrition are given into the extracorporeal circuit. Drug doses are appropriately increased and fluid and electrolytes provided by standard requirements. The infants remain on ECMO for variable periods of time. As the lung injury is repaired, the flow of oxygen required to maintain the infant decreases and the oxygen level in the blood increases. Once the infant has begun to improve and is stabilized, the bypass is weaned and terminated. The perfusion catheters are removed after a mean ECMO duration of approximately 90 hours.
Introduction of ECMO involved a great deal of risk. There was considerable concern about the potential harm of ECMO before its introduction into the unit. It was a major break from previous therapy. There was potentially a significant compromise of blood flow to the brain, and the data available at that time about longterm outcomes were scant. There were no other units in the metropolitan area with ECMO; the nearest units were in Ann Arbor and Pittsburgh. This was only the third unit in the world to introduce ECMO and the only role model for the application of ECMO was Dr. Bartlett's program at the University of Michigan.
In addition, mastery of the technology rested in the hands of the pediatric surgeons. For the first time in the recent history of
the neonatal unit, both the neonatologist and the nursing staff were placed in the position of not having direct control over the technology that would be applied to their patient. However, the unit had a background in risk taking in the recent past with the development and application of respiratory technology to the premature infant with birth weights of less than two pounds. The application of technology in this instance involved risk taking in treating these very small infants with a previously high mortality rate and once again an unknown morbidity rate. The positive prior experience with the application of respiratory technology was an important factor in the ultimate acceptance of this more recent technology into the armamentarium of the unit.
Before the introduction of ECMO, the obvious supporters of the technique were the pediatric surgeons who had perfected it in the animal laboratory, with cautious support from the leadership of neonatology. The anesthesiologist assigned to the unit was also supportive of ECMO as a logical extension of the respiratory care developed under his leadership. There were marked reservations from the junior attending staff and the nursing staff was appropriately cautious about the introduction of a new technology, but both were willing to be risk takers in the controlled environment of the ICU.
The decision process also included a review of the fiscal aspects of the new program. There were insufficient funds within the hospital, and grant support from NIH was not available. Therefore, the initial support for the program came from the pediatric surgical service, which provided not only the pediatric surgeons but also the technical support and the funds necessary to initiate the program. Without the financial, human, and technical support of the Division of Pediatric Surgery, the new technology would never have been introduced.
A major factor outside of the organization that played a facilitative role in the acceptance of the technique was the positive experience being reported by Bartlett and his colleagues at the University of Michigan. Their reputation as careful investigators and outstanding clinicians provided substantial support for the adoption of this technique. Outside businesses played little role in the introduction of the new technology since the membrane oxygenator was already available. There was no pressure being applied to the unit from industry for adoption of the new technique.
A major driving force in the eventual adoption of ECMO was a perceived need to assume leadership in exploring the application of this technique to the care of newborn infants. Staff members recognized that the likelihood of other units using this technique was high and that it was better to assume a leadership position early in the process. If the technique was found to be either lacking or helpful, they would have accumulated first-hand experience. It was also absolutely clear that there was no alternative therapy for these infants; without an alternative therapy, the death of a certain group of infants would continue.
PROCESS AND CONSIDERATION
After discussion among the professionals involved that lasted over six months, a decision was made to implement the technology. At that point the need to develop appropriate administrative channels and a process for admitting infants and treating them efficiently and effectively was developed at the medical and nursing levels with the active participation and leadership of surgical colleagues.
Once the decision had been made at the administrative level, then both the nursing and the medical staffs were jointly involved in the development of the policies and procedures that would govern both nursing and medical practices. The nursing and medical staffs were also involved in the development of the "ECMO Room" and there was an active attempt to develop a team concept. This was an attempt to develop the necessary collegial relationship essential for the success of an ongoing program.
The development of these policies and procedures, however, intentionally did not involve the neonatology and the nursing staffs in learning hands-on application of the technology. This was entrusted to the surgical team who assumed total responsibility for the application and maintenance of the technology. This decision was made to ensure the highest standard of care, alleviate anxiety, and facilitate interaction among professionals who previously did not work together on a daily basis.
To notify colleagues in other neonatal intensive care units and practicing pediatricians about the availability of the new technique, letters were sent to the directors of other neonatal programs in the tristate area. Following this mailing a young pediatric surgeon who had directed the program in the laboratory and who was responsible for its clinical application, began visiting neonatal intensive care units in the tristate area to discuss ECMO
in general and the development of the program in our newborn intensive care unit. Extensive correspondence and lectures provided colleagues in the area with an information base on which they could make decisions about their infants with respiratory failure. Based on both the internal and external discussions, a number of planning considerations emerged before implementation.
For infants with congenital diaphragmatic hernia, should surgeons at outside institutions refer those infants before correction of the defect in the diaphragm to facilitate their care postoperatively if they should require ECMO? Or should the infants have the initial repair and then be transferred if they became candidates for ECMO? In units where university hospitals had training programs that mandated certain experiences for trainees, an automatic transfer of such infants created a serious training shortfall. In smaller units without pediatric surgical training programs, the same sort of educational consideration did not prohibit the transfer of those infants before surgical repair of the lesion. As the program has evolved, referring units have adopted different approaches to the resolution of this issue, ranging from transfer of all patients before surgery to transfer of no patients before surgery.
Medical colleagues in referring hospitals were also confronted with a decision about the timing of transfer that would allow a sufficient interval for the infant to be moved from one institution to another while he or she is critically ill and still allow time for introduction of ECMO upon arrival in our unit. In this particular situation, it is important to note that 50 percent of the infants transferred for ECMO never receive the treatment either because of mortality, a contraindication for the application of the technique, or because they improve with the application of the standard ventilatory therapy in our unit. In such a process, communication between units is essential to the continued success of this operation and the timing of each case had to be handled individually.
Internally there was a decision about how many infants could be treated at one time. It was decided before the introduction of the technique that resources allowed acceptance of only one ECMO candidate at a time. It was also agreed that there would not be "back-up candidates" in the unit until it was quite clear that an infant on ECMO was going to be successfully weaned from the
machine. The interrelationship between the ECMO centers in the Northeast also made it easier for patients who could not be accepted for therapy to be referred to other units that had beds available.
Another organizational factor that needed to be modified before implementation of the technology was the transport system. As previously mentioned, because the NICU is a regional center, transport for infants from network hospitals is provided 24 hours per day. In this instance infants were received from other tertiary units both within and outside our tristate area as well as from other ECMO programs outside the metropolitan area. Some tertiary units had their own transport system and thus were able to transport such infants with their own transport team. For tertiary units that did not have transport systems, it was necessary to provide transport. This requirement demanded interaction with New York City Premature Transport Program, which had the obligation to provide transport for sick infants within the city itself. Permission from the administration of the Preemie Transport System enabled the NICU to assume responsibility for these transports because of the unique nature of the technology.
For infants transferred from outside the metropolitan area, appropriate air transport had to be arranged. Experience of an organ transplant unit provided an appropriate air transport model. After careful consideration, a decision was reached that the NICU role in the transport of patients from outside of New York who would arrive by air was to provide an ambulance, a nurse, and a physician to meet the transporting teams at the airport and arrange transport of the infant from the airplane to the transport ambulance. Transfer of responsibility for the patient, however, would not occur until they arrived in our unit. Transfer of responsibility for an acutely ill infant would more safely take place in the intensive care unit than at the airport site.
The site for ECMO was another organizational factor that was considered before implementation of the technology. Because of the bulkiness of the equipment and the uniqueness of the situation, it was decided to use a room directly adjacent to the unit but yet separate from the unit. This would allow the physicians and the single nurse who was assigned to this infant 24 hours a day to work somewhat in isolation yet enjoy the full support of the unit. This "isolation" from the unit remains an administrative problem to this moment and will be discussed later under unanticipated effects.
Finally, it was important to the neonatal intensive care unit
staff and their surgical colleagues to be aware at the outset that there would be infants placed on ECMO who would be unresponsive to therapy immediately or late and expire in the early or late stages of treatment. For others the therapy would need to be discontinued because of complications or because after 10 or 12 days on the technique they did not respond. Each of these issues was discussed before implementation of the program, but until a team has faced the dilemma it is impossible to anticipate group reaction. However, because the neonatal team had prior experience with the discontinuation of medical therapy both in full-term and premature infants, decision making was not new and there were acceptable standards in place to provide the basis for making these decisions.
With regard to the management of setbacks and failures, it was decided that each case should be reviewed both prospectively and retrospectively to decide what had gone wrong, if anything, and why. In each instance these reviews involved the pediatric surgeons, the neonatologist assigned to the case, and the nursing staff. Where organizational defects had been detected in the management of an individual infant, the review took place at the level of the director of the newborn intensive care unit and all services to isolate the problem. To prevent its recurrence, it was necessary to make appropriate modification in the operating policies and procedures. In the beginning, quite understandably, the problems were often administrative, and changes were made to facilitate process, whether it was movement of the infant from one institution to another, shortening the time required to place an infant on ECMO once a decision had been made, or transfer of information to the referring institution. Reviews were conducted by the pediatric surgeon directing the program. They were critical and constructive, and the only goal was to improve the operation. The team was aware of the intensity of the experience, particularly the critical nature of time, rapid decision making, and risk to the infant involved.
Broad measures were also used to monitor the consequences of this innovation, including mortality and longitudinal morbidity, which are essential markers measuring the consequences of instituting a new therapy. In addition, both the pediatric surgeon involved and the hospital monitor the cost of the new innovation. The pediatric surgeon also maintains a log of all requests as a
means of monitoring the ability of the unit to accept patients in transfer.
After the initial experience with ECMO, it became obvious that one problem was the availability of sufficient additional nursing manpower to allow one-on-one nursing care for the sick infant. At times, depending on number of infants in the intensive care unit, there were not sufficient personnel to remove a nurse from the existing staff and to assign that professional to the case. Once this situation was recognized, the institution approved two new nursing positions that would be assigned to the ECMO program and that would provide the basis for an on-call system for nurses. With the on-call system, the unit then had the potential addition of a nurse for ECMO as the demand arose. Thus, the administration of nursing was not required to change schedules based on the potential needs of ECMO. Costs involved in this on-call system and in providing additional nurses were borne by the hospital.
ECMO represented the introduction of a brand new technology. The adoption did not require the acquisition of new technical knowledge for the nursing staff or the neonatologists, however, nor the abandonment of prior knowledge. Rather modification of existing knowledge was needed to aid the technology. For the first time, the nursing and medical staffs did not have total control over the clinical care of the patient. In almost every other treatment, they might relinquish control for brief periods of time, for example, during surgery, but control was resumed when the infants returned to the NICU from the operating room. This loss of control had the anticipated effects on both the nursing and the medical staffs.
At the outset there were only two physicians involved in this program. When an infant was selected for ECMO the pediatric surgeon and his fellow were at the bedside continuously while the infant was on ECMO. For some infants this was a relatively short period of time—two or three days—but for others it was for as long as 10 or 12 days. The dedication and endurance of these two physicians ensured consistency in decision making and continuity in the care of the patients. It also allowed the nursing staff to
become accustomed to dealing with the same individuals in these recurring clinical situations.
After a year of using this approach, however, it became obvious that the enormous time commitment from the pediatric surgeons could not be maintained. The physical presence of a pediatric surgeon at a patient's bedside throughout the period of ECMO was no longer possible. At that point it became necessary to introduce the perfusionists, another group of professionals previously not exposed to the NICU and vice versa. They would be responsible for the maintenance of the machine once the infant was placed on ECMO. The introduction of these new professionals was discussed at length before the change. The nursing and medical staffs were informed, and there was anxiety from all involved about the introduction of this change. A new unproven, unknown professional was introduced into a high-tech process, where there were great risks to the infant involved and where the medical and nursing staffs were not at ease with the technology.
Despite the attempts of the leadershop to alleviate this anxiety and to provide a smooth transition, there were major problems of staff confidence and implementation of the new system. This alteration required a series of meetings with the medical and nursing staffs to assure them of the new professional's competence and to explain the exact role of the new participant. The pediatric surgeon was still available 24 hours a day. The perfusionist was included in the administrative hierarchy of the unit related to ECMO patients and this contact, although infrequent, increased the confidence in another group of professionals and a smoother transition to the new system.
The purpose of the ECMO program was to determine whether the new technology would in fact reduce mortality and whether the surviving infant would have an acceptable morbidity at one and two years of age. It was also the intent of the organization to be an educational resource and provide this service on a regional basis, since the only other programs were in Ann Arbor and Pittsburgh.
ECMO has been successful in accomplishing the first two purposes. Infants who previously expired with diaphragmatic hernia and meconium aspiration now survive. The survival rate to date
involving 51 patients is 84 percent. The effectiveness of ECMO in terms of mortality is easy to measure, but an equally important question is the longitudinal development of the infants. That information is being collected as we proceed and to date approximately 80 to 85 percent of the infants who have required ECMO are considered normal at two years of age. Educationally, the program has been an asset to both neonatal and pediatric surgical programs, and as a university hospital the program has provided another arena of clinical investigation.
With regard to the third purpose, providing the technology on a region-wide basis, the program has not been totally successful. The success has been limited primarily by availability of beds and availability of nursing. The neonatal intensive care unit has had increasing demands for its services related to the crack epidemic currently afflicting NYC and particularly Washington Heights, the immediate environment of the institution. The delivery rate in the institution has increased to 5,300 on a delivery service that was designed for roughly 3,500 mothers. This unexpected increase in delivery rate has stretched the resources of the neonatal intensive care unit to the point at which it has been required to transfer sick infants born in the institution to other neonatal intensive care units within the metropolitan area.
Increased occupancy has become a regional problem with the insufficient beds available to provide for sick infants in the region. The lack of beds along with the recent implementation of Diagnostic Related Groups has created an economic environment that is threatening to every hospital in New York State and has inhibited development of an orderly plan to meet the crisis of bed availability. This problem is being dealt with by the New York City Health Department and the State Department of Health.
The increasing crisis with regard to beds has forced the NICU to deploy new resources to meet demands of patients born in the institution. Included in this process is the opening of a transitional nursery in the delivery room around the clock rather than as a triage area for the care of infants. This physical unit often requires the presence of a nurse 24 hours a day, further stretching an already stressed nursing staff. It also pulls a nurse from the nursing pool and prohibits the opening of the ECMO room when a request for transfer occurs. The number of nurses does not allow the simultaneous operation of the newborn intensive care unit, the transitional nursery, and acceptance of an ECMO patient.
In late fall 1988 the crisis reached enormous proportions, and a large number of requests for ECMO were not accepted. The rejec-
tion of requests creates not only institutional problems but also regional problems. Even though the surgical team located beds for the requested transfers, in each instance the logistic problems involved in moving a sick infant out of the metropolitan area to either Boston, Pittsburgh, or Washington, D.C., were enormous. The credibility of the program is threatened and colleagues will begin to look elsewhere for their patients. Another issue is the development of other ECMO programs to serve the tristate region as a result of NICU's inability to meet requests for transfer. This is a logical step for other institutions to take although fiscally and organizationally the need for more than one unit in the area has not been demonstrated. Although the program has achieved its intended purpose in the survival of infants placed on ECMO with good longitudinal outcome, the goal of serving as a regional resource has not been accomplished.
There are always surprises when anything novel is introduced into an organization. Perhaps most surprising at the human level is that, despite successful implementation of ECMO, a certain amount of resistance and antagonism persists. Strong initial resistance to this invasive high-risk therapy has abated as the success of the technique accumulated. Some resistance continues, however, because the process is labor intensive and the technology still resides in the hands of surgical colleagues and therefore retains some degree of intimidation on the part of both the medical and the nursing staffs.
The assumption also remains among the professionals involved that a pharmacological approach to the treatment of these infants may be possible with proper investigation and research into the development of respiratory failure. ECMO programs may no longer be necessary once the key developments in research have occurred and the appropriate drugs for treatment are in place. Thus it may be that ECMO is a transient technology that would allow survival of infants while research develops a less invasive approach to treatment of the related disorders.
The ''isolation'' of the professionals treating an infant on ECMO was also not anticipated. Out of sight/out of mind mentality permeated the behavior of the staff not directly involved. The significance of this isolation cannot be overestimated. Movement of the patient into the main unit would not require further reduction of census, since NICU census is reduced by one bed when
an ECMO infant is accepted. In a sense, it is time for ECMO to join the unit and for us to determine the benefits of a move of six feet on staff morale!
There are a number of changes that would be appropriate if this system were implemented anew. More time should be devoted to the development of consensus before the implementation of a new technology. Where consensus cannot be achieved before implementation, the assignment of a core group of individuals at the professional level who are enthusiastic supporters of the technique would contribute to a smoother transition in the application of the technology. Once the technique has proven beneficial, then extension of responsibility to other members who have been previously resistent would be simplified.
It is also obvious, retrospectively, that we should have had a greater degree of institutional commitment to the enormous resources involved. However, the enthusiasm of the surgical staff to implement the program led them to assume fiscal and physical responsibilities above and beyond those for any professional or division. At a time of cost containment, and particularly with the implementation of the Diagnostic Related Groups in New York State, the fiscal problems associated with the implementation of a new program should have been more clearly anticipated and dealt with in a prospective fashion.
Now some five years after institution of ECMO, the staff more readily accepts the benefits of the procedure to infants selected for ECMO, but continues to have questions about the technology. In light of the success of ECMO, this continuing concern with regard to the technology and the lack of control of that technology on the part of neonatal nursing and medical service remain an enigma. Perhaps as we move further from the second transition to the use of perfusionists, and as those professionals interact with our staff more regularly, the concern about technology will fade. It may well be, however, that until the nursing staff and medical staff are more familiar with the technology in a hands-on fashion, this concern will not be totally alleviated.
The joint leadership of the ICU and the ECMO program will now request outside assistance in addressing the issues that in both measurable and unmeasurable ways continue to affect on the smooth operation of the ECMO program, particularly the continued staff resistance. Under immediate consideration is the issue
of how much the neonatal nursing and attending staffs should learn about hands-on application of the technology. The final decisions with regard to this issue have not been reached.
ENDURING SOCIAL EFFECTS
Retrospectively, the expectations of the professionals responsible for the implementation of the program have been realized in terms of improved survival for the infants selected for ECMO. Concerns about the potential neurologic complications for these infants have been alleviated. The favorable development of infants who have been on ECMO has served to reduce the concerns of both medical and nursing staffs about the potential complications of the procedure itself on the developing central nervous system. These outcomes could not have been accurately predicted, because the experience at the national level was still relatively limited at the time the program was implemented.
These results confirm more recent national findings. An increasing number of infants are being treated with ECMO and surviving as more centers report their experience. The most recent report from the ECMO central registry described 2,528 newborn patients with respiratory failure supported by ECMO from the 56 reporting centers. The patients were treated between 1980 and 1989. Eighty-three percent of these patients survived. It is important to remember that before the introduction of this technology, the mortality rate for these infants was 90 percent. These results indicate that ECMO and lung rest are an appropriate and successful treatment for the newborn with respiratory failure that has been unresponsive to other means of treatment.
The introduction of ECMO into an established neonatal intensive care unit has been successful in reducing mortality, with acceptable longitudinal outcomes for the high-risk infants. The unexpected fiscal and manpower costs were not reasonably estimated in the beginning and continue to be a problem for pediatric surgery, the institution, and the neonatal intensive care unit. The interaction of the professionals involved has improved substantially over the course of the program, but the lack of control of the technology by the neonatal nursing and medical staffs remains a major problem. The concurrent development of an unexpectedly high delivery rate and a nursing shortage have obstructed the goal of serving as resource for the tristate area. The recent changes and the addition of new beds may allow achievement of this last goal. In short, the introduction of ECMO has been successful in
saving lives previously lost, but costly and not yet fully implemented as a regional resource. The process is in place, however, to modify the system and to resolve remaining problems.
Bartlett, R. H., A. B. Guzzaniger, R. F. Huxtable, H. C. Schippes, M. J. O'Connor, and M. R. Jefferiar. 1977. Extracorporeal circulation in neonatal respiratory failure (ECMO). Journal of Cardiovascic Surgery 74:826–833.
White, J. J., H. G. Andrews, H. Rosenberg, and D. Mazun. 1971. Prolonged respiratory support in newborn infants with a membrane oxygenator. Surgery 70:288–296.