Simulation-Based Learning

Approximately 4 million babies are born in the United States every year; of these, around 10% require some degree of resuscitation, with 1% needing extensive resuscitative efforts such as chest compressions, intubation, and delivery of medication.⁷⁰ Many different types of health care professionals are responsible for caring for newborns at the time of birth and in the days and weeks that follow. These professionals include, but are not limited to, neonatologists, pediatricians, family practitioners, obstetricians, midwives, neonatal nurse practitioners, nurses, and trainees at all levels in these disciplines. Given the large number of births, the frequency of resuscitation, and the diversity of professionals bearing responsibility for caring for patients in the neonatal period (the first 28 days of life), the need for effective means of acquisition and maintenance of the skill sets necessary to deliver safe and competent care is of tremendous importance. Traditionally the apprenticeship model of assuming graduated responsibility for the care of real patients has been used to address this need. Unfortunately, the assumption underlying this model—that placing a trainee in a supervised clinical environment for a set period of time will allow him or her to experience a sufficient number and breadth of clinical cases to ensure the ability to practice independently and safely in the community—does not always prove to be true. Similarly, maintenance of skill cannot be guaranteed by the routine, non-mentored delivery of patient care. Therefore a new paradigm of skill acquisition and maintenance is required.

Whereas teaching is something that is (passively) done to trainees, learning is something that trainees must (actively) do themselves.⁴¹ Because not everything that is taught is necessarily learned, programs that best facilitate skill acquisition in trainees are those that focus on learning, rather than on teaching. Traditional didactic programs are passive by nature, and the settings in which they are held are typically isolated from realistic cues, distractors, and time pressure; thus, such programs are unable to prepare learners adequately for all of the challenges inherent when working in the real environment. Although learning in a real environment during the actual delivery of patient care may appear ideal at first glance, a deeper analysis reveals otherwise. The most obvious problem with using the real environment as the primary source of skill acquisition and maintenance is that any mistake could prove lethal to patients. The pace of actual clinical care conducted in the real environment with real patients is often too fast to allow trainees to take full advantage of the learning opportunities therein. Moreover, typically there is no way to ensure that all important learning opportunities will present themselves in the real environment during the time that the trainee is present. Finally the real environment is also a very expensive environment and is populated with a number of professionals whose job description may not include providing learning opportunities for trainees.³⁹ Learning is best facilitated when the learning opportunities are tailored to meet the needs of the learners. Training models that offer the same content in the same fashion to all learners (thus implying that competency can be attained and maintained simply by spending a particular, often arbitrary, amount of time at a task) fail to recognize that adults have different strengths, weaknesses, and life experiences and acquire and maintain different skills at different rates. Some characteristics of effective adult learning strategies include the following:

Any discussion of learning in health care must start with what it is that can be learned. There are three “skill sets” that may be acquired and refined by health care professionals:

Content knowledge is the skill set most familiar to learners and is typically the major (or only) skill set that is formally evaluated, usually through written or online tests. Technical skills such as intubation are critical to neonatal care. Despite their importance, such skills are most commonly practiced at skills stations using models that poorly represent neonatal anatomy and physiology and are evaluated by a subjective assessment of performance that is isolated from the time pressure intrinsic to the real environment. Behavioral skills (including but not limited to leadership, teamwork, and effective communication) are critically important to successful patient outcomes (Box 7-1). Unfortunately these important skills are rarely, if ever, specifically addressed in learning programs directed at health care professionals. Many patient care tasks actually incorporate elements of all three of these skill sets. Intubation of the newborn is one such example. Far from being simply a technical skill, effective and safe intubation requires coordination and integration of multiple cognitive skills (knowing the indications for intubation and the signs of successful and unsuccessful intubations), sequential discrete technical skills (assembling, testing, and inserting the laryngoscope), and a number of behavioral skills (effectively communicating observations and needs, evenly distributing the workload, and delegating responsibilities), all of which must also be accomplished in a time-efficient manner (Box 7-2).

Simulated clinical scenarios coupled with debriefings (in which discussion of what went well and what could be improved upon) provide rich learning experiences that equal or exceed those of other learning methodologies. Although undoubtedly some learning takes place during active participation in scenarios, trainees perceive that most of the learning occurs during the debriefing when time is allotted for self- or facilitated reflection on performance. This perception is supported by years of anecdotal experience using simulation in high-risk domains as well as decades of research in the science of adult learning.

Much has been made of the importance of the concept of fidelity in simulation-based learning. Simulation fidelity is typically thought of in terms of its physical, biologic, and psychologic elements. Physical fidelity refers to the realism of the physical space in which training occurs; this space is made to look real by including appropriate working medical equipment, fluids, pharmacologic agents, beds, and the other elements necessary for patient care. Biologic fidelity includes the patient simulators and standardized patients as well as the human beings acting as confederates during the simulation, playing roles designed to assist the evolution of the scenario. Patient simulators have been described as high, medium, and low fidelity; unfortunately there is no standardized definition of simulator fidelity in health care. In reality, no physical patient simulator currently in use bears close resemblance to a human being, either in terms of anatomy or physiology. The use of the term high fidelity when describing the current generation of patient simulators more likely refers to high complexity or high cost rather than any intrinsic similarity to a living human being. Finally all of the previously mentioned elements interact with the mindset brought into the scenario by the learners to create a sense of realism or psychologic fidelity. The overall goal of simulation-based learning is to provide learning experiences that closely mimic the conditions encountered when working in the real environment. The major difference between the simulated environment and the real environment is the absence of real human patients. Although debate continues in the health care simulation community as to how much fidelity is necessary and although some may argue that the higher the fidelity of the scenario to real life, the better the learning opportunity, it should be understood that as long as sufficient attention is paid to providing the key (not all) visual, auditory, and tactile cues for learners, allowing them to form a shared mental model of the nature of the situation that they are facing, they will have the opportunity to work effectively to resolve the clinical problems that become manifest during the scenario and therefore achieve the learning objectives.⁹

Simulation-based learning provides many obvious advantages over more traditional training methodologies. Because patient simulators replace human beings, there is no risk to patients; invasive procedures can be practiced without the fear of patient harm or medical liability. Unlike what happens in the real environment, learning opportunities using simulation can be scheduled at convenient times and structured so that specific learning objectives are consistently achieved. Simulation-based learning is an ideal methodology for allowing learners to practice integration of multiple skill sets while working under highly realistic and often stressful conditions. Rather than being directed solely at the individual, simulation easily accommodates the learning needs of multidisciplinary teams. Simulation-based learning activities can easily be scaled in intensity to meet the needs of learners at all levels of experience, and they can be used to foster both the acquisition and maintenance of particular skills. It can also be hypothesized that learners who participate in simulation-based exercises likely will be better prepared and will need less supervision when entering or re-entering the real environment (Box 7-3).

Simulation-based learning opportunities are the standard for skill acquisition in industries such as commercial aviation, aerospace, nuclear power, and the military, in which the risk of death or severe injury to human beings is very real. The use of simulation in these domains is characterized by an emphasis on probing human beings and the systems they design (both technological and social) for weaknesses and then designing and testing solutions to address those weaknesses. This type of learning is the result of a culture that not only fosters, but also demands, a willingness to learn from mistakes made during simulated events, thus decreasing the chances of repeating such mistakes when working in the real environment.

The National Aeronautics and Space Administration (NASA) was established in 1958 to conduct aeronautical research and administer the human and robotic exploration of space. Space travel is an inherently risky business; how else can one describe a process that places human beings in a rocket filled with tons of liquid fuel and then ignites that fuel in the hope that it will propel those humans into the vacuum of space? The value of simulation was made clear during the Apollo 13 mission, launched on April 11, 1970, as the third mission to land humans on the moon. To prepare for this mission, the three members of the prime crew, Jim Lovell (captain), Tom Mattingly (command module pilot), and Fred Haise (lunar module pilot) trained in NASA’s flight simulators for months. One week before launch, Mattingly was exposed to the measles; because he was not immune to the disease, his backup, Jack Swigert, was given simulator time with Lovell and Haise in the week preceding launch to “ensure that Lovell, Swigert and Haise could function with unquestioned teamwork through even the most arduous and time-critical simulated emergency conditions.”⁵ A decision by the flight surgeon only 1 day before launch scrubbed Mattingly from the prime crew and placed Swigert in the left-hand seat as command module pilot for the mission. Fifty-six hours into the flight, as Apollo 13 was approximately 200,000 miles away from the earth en route to the moon, an explosion in the service module’s cryogenic oxygen system resulted in the uncontrolled venting of oxygen into space, creating a situation that threatened not only the success of the mission, but also the lives of the crew (oxygen was the source of the crew’s breathing air and substrate for the fuel cells that generate electrical power). As has been thoroughly documented, the crew did return safely to earth. What is less well known is that, years before Apollo 13 launched, the procedures that allowed the crew to survive and recover from this devastating event were devised by engineers charged with envisioning every possible failure and then designing procedures to address these failures.³¹ These procedures were tested under current mission parameters in the flight simulators almost continuously during the crisis and when deemed reliable were relayed to the crew.¹¹ In the formal post-mission debriefing, the crew referred to their experiences in flight simulation in excess of 40 times.⁶ Based largely on its successful use in other domains and its inherent face validity, simulation is being employed as a learning methodology with increasing frequency in health care.³⁷

In 1999, the Institute of Medicine (IOM) published To Err Is Human: Building a Safer Health System, a report on human error and patient safety in the United States.⁵¹ In this report, the authors estimated that between 44,000 and 98,000 Americans die each year as a result of medical errors. Although this figure has been highly debated, it is based on extrapolation of the data contained in studies out of Colorado, Utah, and New York published in peer-reviewed literature.^10,55,69 The 1999 report was followed in 2001 by another from the IOM, Crossing the Quality Chasm: A New Health System for the 21st Century, in which the type of interventions, including training methodologies, necessary to improve patient safety were discussed.¹⁵ Subsequently in 2004, the Joint Commission (JC) published a Sentinel Event Alert describing ineffective communication as a major cause in almost 75% of the 47 cases of neonatal mortality or severe neonatal morbidity (lifelong serious neurologic compromise) reported to that agency; since that time, an additional 62 cases have been added.⁴⁹ In response to these root cause analyses, the JC recommended that all health care organizations responsible for delivering newborns “conduct team training in perinatal areas to teach staff to work together and communicate more effectively” and “for high-risk events, such as shoulder dystocia, emergency cesarean delivery, maternal hemorrhage, and neonatal resuscitation, conduct clinical drills to help staff prepare for when such events actually occur, and conduct debriefings to evaluate team performance and identify areas for improvement.”

Simulation-based learning is grounded in adult learning theory, supported by rational conjecture, and felt to be essential for achieving expert performance.32,33,60,75 Simulation-based learning in its broadest sense has been used for decades in other domains in which the risk to human life is high and is a core component of maintenance of certification programs in those domains. Historically in health care, the emphasis has been on evidence-based practice and the gold standard of evidence has consisted of the prospective, randomized, controlled, sufficiently powered clinical trial in which the results focus on patient outcomes. However, there is an evolving debate about the type and extent of evidence required to adopt quality assurance/improvement efforts in health care. At one end of the spectrum are clinicians and investigators who insist that quality initiatives must be subject to the same rigorous testing that precedes the introduction of new pharmacologic therapies and medical instrumentation in order to prove that they actually improve quality and ensure patient safety. Alternatively, others note that requiring randomized controlled trials to assess the safety of innovations with high face validity may place humans at undue risk and, therefore, prove impossible to conduct. In fact, some authors are of the opinion that not to use simulation-based training methodologies, relying instead solely on practice on real patients, is ethically indefensible.⁷⁷ Thus, we are left with a situation in which the need for more definitive evidence, although desirable, is felt by at least some members of the health care education and training community not to be necessary or practical. Although numerous studies across multiple health care domains have shown that simulation produces short-term improvement in skills (even in highly experienced professionals) and the body of objective data supporting the use of simulation in health care continues to expand, no evidence has been published that definitively asserts its effects on patient outcome or its return on investment.* That stated, it must also be emphasized that no one working in other high-risk domains such as commercial aviation, aerospace, the military, and nuclear power would consider conducting a prospective, randomized, controlled trial with subjects who are randomized to the “no simulation” group; indeed, simulation-based training remains standard operating procedure for these professionals.⁶⁷

The ever-expanding body of knowledge of the basic processes underlying normal and abnormal neonatal physiology has allowed those clinicians responsible for caring for newborns to generate evidence-based clinical practice guidelines under the auspices of the International Liaison Committee on Resuscitation (ILCOR) and its member organizations (Neonatal Resuscitation Program of the American Academy of Pediatrics, the American Heart Association, the Heart and Stroke Foundation of Canada, the Inter-American Heart Foundation, the European Resuscitation Council, the Australian and New Zealand Committee on Resuscitation, and the Resuscitation Councils of Southern Africa). The International Liaison Committee on Resuscitation was founded in 1992 to facilitate international collaboration on issues involving neonatal, pediatric, and adult cardiopulmonary resuscitation and emergency cardiovascular care.¹² As knowledge about the physiologic processes underlying neonatal cardiorespiratory decompensation and the list of therapeutic interventions grow, so too do the expectations for mastery of this knowledge and associated skill sets that are placed on those responsible for caring for the neonate in distress. In recognition of this fact, the ongoing ILCOR process includes a review of the evidence behind the use of simulation and debriefing in resuscitation training. In 2010, ILCOR noted that a number of studies have demonstrated that the use of simulation-based learning methodologies enhances performance during simulated resuscitations. Although acknowledging that the interpretation of data generated by these studies is often complicated by their inherent heterogeneity and limitations, ILCOR nevertheless recommended that simulation, briefing, and debriefing techniques should be used during training when caring for simulated patients and in the course of clinical activities involving real patients to facilitate the acquisition and maintenance of the skills necessary for effective neonatal resuscitation.^50,63

Neonatal care occurs in environments that are extremely dynamic and complex, and the nature of the work performed in those environments requires that correct decisions be made and appropriate interventions be carried out, often while working as a member of a multidisciplinary team in the context of intense time pressure. Simulation is an ideal learning methodology to allow teams of learners to practice working in these types of environments.⁴ The first simulation-based learning program in neonatal-perinatal medicine (and one of the first in all of health care) is the NeoSim program developed at the Center for Advanced Pediatric and Perinatal Education (CAPE) located at Packard Children’s Hospital on the campus of Stanford University in Palo Alto, California. Launched in 1997, NeoSim has been a very successful innovation in training in the cognitive, technical, and behavioral skills necessary for optimal care of the newborn in distress.^44,61 Subsequently, the NeoSim program was adopted as the basis for extensive ongoing revision of the Neonatal Resuscitation Program (NRP) of the American Academy of Pediatrics (AAP), the national standard for training in neonatal resuscitation.³⁸ Since 1987, the NRP of the AAP has set a national standard and an international example for training in the resuscitation of the newborn and has enjoyed tremendous success by claiming more than 3.2 million trainees and more than 27,000 instructors in the United States alone. The NRP’s Textbook of Neonatal Resuscitation has been translated into 25 languages, and the NRP has been taught in 124 different countries around the world. A major shift in learning methodology is currently happening within the NRP. Instructors are being required to shift their role from that of a teacher responsible for imparting knowledge to trainees, to that of a facilitator who fosters acquisition of skills by learners as they accept primary responsibility for their own education. The Steering Committee also developed a list of the characteristics desired in a cost-effective human neonatal patient simulator and published this online in 2005 as a request for proposals to industry. This marked the first time in the history of health care simulation that a professional body, rather than an industry, drove development of a realistic patient simulator based on established learning objectives. Development of a career-long learning program in neonatal resuscitation that is relevant to professionals from multiple disciplines at all levels of experience and is embedded with robust learning opportunities and valid performance metrics is the ongoing focus of the NRP as it continually adapts to stay relevant and provide optimal simulation-based learning experiences.⁷

The subspecialty of neonatal-perinatal medicine is unique in that one patient (the fetus) exists inside of another patient (the mother); in the case of multiple gestations, two or more neonatal patients await birth inside of the pregnant female patient. The possibility of a sick mother delivering (a) sick newborn(s) creates a situation in which optimal preparation occurs only when the neonatology and obstetric teams train together.⁴⁰ The multitude of events that can complicate human birth and the neonatal period make this an especially appealing target for simulation-based learning.⁶¹ Thus it makes sense that the neonatologists, obstetricians, and nurses from labor and delivery units and newborn nurseries who work closely together in the delivery room caring for patients should also conduct joint simulation-based learning exercises.^22,23,56

An example of an obstetric simulation-based learning program that is directly relevant to fetal and neonatal care is the Obstetrics Emergency Training Programme at Southmead Hospital, Bristol, United Kingdom. Draycott and colleagues used a variety of patient simulators in several simulated environments to show that simulation-based learning resulted in enhanced content knowledge and improved technical management of shoulder dystocia.18–21,29 In addition, the same group found in a retrospective multicenter cohort observational study of 19,460 infants that (1) the incidence of infants born with 5-minute Apgar scores of 6 or lower decreased from 86.6 to 44.6 per 10,000 births (p < .001), and (2) hypoxic ischemic encephalopathy (HIE) decreased from 27.3 to 13.6 per 10,000 births (p < .032) over a 5-year period following the introduction of a training program consisting of a review of fetal heart rate tracings and hands-on drills in the management of shoulder dystocia, postpartum hemorrhage, eclampsia, twin delivery, breech presentation, and maternal and neonatal resuscitation.³⁰ This study remains one of the best examples of improvement in clinically relevant outcomes in association with an educational intervention in health care.

Human birth is characterized by nearly continuous changes in the physiology, anatomy, and spatial relationships among various physical structures in both mother and baby, and simulation of the process of labor and delivery is therefore a technically complex endeavor. Purely mechanical devices are not able to simulate vaginal birth in a manner akin to real life in a cost-effective manner or to allow practice of highly invasive procedures such as cesarean section. Development of hybrid technologies that combine materials similar to the plastics used for physical patient simulators with visual displays and haptic interfaces capable of generating the images and tactile sensations associated with patient care will create learning opportunities that are currently impossible to achieve in the absence of a real patient. Thus combinations of physical whole body simulators with virtual reality interfaces designed to compensate for the physical simulators’ limitations will play a major role as pediatric and obstetric simulation evolves. In addition to physical and hybrid patient simulators, highly interactive web-based virtual environments will allow multiple professionals located in geographically distinct regions to participate in simulated clinical scenarios tailored to meet their specific learning needs. Following the example of the NRP, close collaboration among physicians, computer scientists, biomedical engineers, medical artists, and others will allow the technical challenges of simulating birth to be overcome in a timely and cost-efficient manner.36,38,39,43

What are the features of the best simulation-based learning programs that make them effective in facilitating the acquisition and maintenance of cognitive, technical, and behavioral skills? Each element of a successful program, including the scenarios, is built around specific learning objectives that are tailored to meet the needs of the learners. Every effort is made to ensure that the scenarios to which the learners are exposed are realistic in detail, challenging in scope, and relevant to their practice. Finally expert feedback and facilitated debriefings assist the learners in their critical self-reflection on performance. Effective simulation is not dependent on the purchase and use of highly complex and expensive patient simulators costing tens or even hundreds of thousands of dollars; more important are carefully designed scenarios that align with the needs of the learners, provision of important (not necessarily all) cues, and conduct of skillfully led debriefings.

The rationale for employing simulation-based learning in neonatal-perinatal medicine is clear. The management of serious neonatal pathology is one example of the classic low-frequency, high-risk event that lends itself well to simulation-based learning.⁴³ Many health care professionals who care for newborns have the opportunity to manage serious or rare disease processes on an infrequent basis. Even for those for whom a sufficient number of opportunities do exist, one must question whether it is acceptable to essentially practice on real living patients who are not capable of providing informed consent on their own. Although parents do act as surrogate decision makers for children below the age of consent, few want to contemplate that their child will be the first one on whom someone will perform their first spinal tap, first intubation, or first thoracostomy tube placement. Therefore it may be argued that the ethical imperative for simulation is stronger in pediatrics in general and in neonatal-perinatal medicine in particular than in any other field of health care.⁷⁷