This book is about decision making and crisis management in anesthesia. What is a crisis ? It is “a time of great danger or trouble whose outcome decides whether possible bad consequences will follow.” For our purposes, the time of great danger is typically a brief, intense event or sequence of events that offers a clear and present danger to the patient. Almost by definition, a crisis requires an active response to prevent injury to the patient; it is unlikely to resolve on its own. Of course, the best way to deal with a crisis is to prevent it from occurring in the first place. An old saying is that “it is easier to stay out of trouble than to get out of trouble.”
Skilled crisis management in anesthesia is no mystery. It demands that the anesthesia professional, while under stress and time pressure, optimally implements standard techniques of diagnosis and treatment for the patient. Medical knowledge and skills are essential components of the decisions and actions performed during crises, but they are not enough. To actually make things happen quickly and safely for patient management, the anesthesia professional must manage the entire situation , including the environment, the equipment, and the patient care team. These management skills include aspects of cognitive and social psychology and even sociology and anthropology. In this chapter we delineate the underlying conceptual foundations of patient safety, and in the next chapter we provide specific practical principles about crisis management. Chapter 3 reviews how to train clinicians to enact these principles, and Chapter 4 covers the art and science of debriefing about crisis management after real patient care events or after simulation scenarios. The remainder of this book (Catalog of Critical Events in Anesthesiology) offers specific recommendations for the recognition and management of a large variety of crisis situations.
Anesthesiology, by Its Nature, Involves Crises
Why is a book on medical crisis management addressed to anesthesia professionals (this term encompasses anesthesiologists, nurse anesthetists, and anesthesia assistants)? What makes anesthesiology and a few other medical domains (such as intensive care medicine, emergency medicine, obstetrics, neonatology, and surgery, to name a few) different from most other medical fields? The answer, to a large extent, is that the clinical environment of anesthesiology is dynamic, and this dynamism interacts very strongly with the complexity of the environment. The combination of complexity and dynamism makes crises much more likely to occur and more difficult to deal with. Thus the expert anesthesia professional must be skilled and therefore trained in crisis management. Following the work of Woods , and of Orasanu and Connolly, we address some of the aspects of anesthesia that make it a “complex, dynamic world,” namely, that it is event-driven and dynamic, complex and tightly coupled, uncertain, and risky (for the patient).
Event-Driven and Dynamic
The anesthetized patient’s state changes continuously. Unpredictable and dynamic events are frequent. The initiation of many events is beyond the anesthesia professional’s control, such as when the surgeon inadvertently transects a major vessel or when a patient with a previously unknown allergy suffers anaphylaxis.
Complex and Tightly Coupled
In technologic systems, complexity stems from a large number of interconnected components. The patient is the main “system” of interest to the anesthesia professional. Patients are intrinsically very complex, and they contain many components, the underlying functions of which are imperfectly understood. Unlike industrial or aviation systems, patients are not designed, built, or tested by humans, nor do they come with an operator’s manual.
Some physiologic systems are buffered from changes in others, whereas certain core components, such as oxygen (O 2 ) delivery and blood flow, are tightly coupled and interact strongly. , Anesthesia ablates some protective and compensatory physiologic mechanisms and will force the patient’s systems to become more connected. The patient’s physiology may also become tightly joined to external systems such as ventilators or infusions of hemodynamically active drugs.
Although the medical equipment connected to the patient is not as complex as that found in aircraft or spacecraft, it often consists of a proliferation of independent devices with multiple, nonstandardized interconnections. Devices are typically designed in isolation by engineers so that interactions between devices, or among the equipment, the patient, and the human operator, may not be adequately addressed in the design phase. These factors increase the complexity of the domain.
The patient as a system contains inherent uncertainties. The medical world knows very little about the underlying causes of specific physiologic events, although the general physiologic principles involved can be described. The true state of the patient cannot usually be measured directly but must be inferred from ambiguous patterns of clinical observations and data from electronic monitors. These data are imperfect because, unlike industrial systems (which are designed and built with sensors in key areas to measure the most important variables), separate, predominantly noninvasive methods are used to measure the variables that are easiest to monitor. Most physiologic functions are observed indirectly through weak signals available at the body surface and thus are prone to various sorts of electrical and mechanical interference. Even the invasive measurements are vulnerable to artifacts and uncertainties of interpretation.
Even if the anesthesia professional could know the exact patient state, the response of the patient to interventions is extremely variable. Even in “normal” patients, genetic or acquired differences in reflex sensitivity, pharmacokinetics, or pharmacodynamics can yield a wide range of responses to a given dose of a drug or to a routine action (e.g., laryngoscopy). In diseased or traumatized patients, or in the presence of acute abnormalities, these responses may be markedly abnormal, and patients may “overreact” or “underreact” to otherwise appropriate actions.
The decisions and actions taken by anesthesia professionals can determine the outcome for the patient. Even for elective surgery involving healthy patients, the risk of catastrophe is ever-present. Death, brain damage, or other permanent injury may be the end-result of many pathways that can begin with fairly innocuous triggering events. Each intervention, even if appropriate, is associated with side effects, some of which are themselves catastrophic. Furthermore, many risks cannot be anticipated or avoided. Unlike a commercial flight, which can be delayed or aborted if a problem occurs, immediate surgery may be necessary to treat a medical problem that is itself life-threatening. Balancing the risks of the anesthesia and surgery against the risk of the patient’s underlying diseases is often extremely difficult.
How Do Crises Arise?
A crisis is often perceived as sudden in onset and rapid in development, but, at least in retrospect, one can usually identify an evolution of the crisis from underlying triggering events. Figure 1-1 illustrates this process. In this model, underlying factors lead to specific triggering events, which initiate a problem. A problem is defined as an abnormal situation that requires the attention of the anesthesia professional but is unlikely, by itself, to harm the patient. Problems can then evolve and, if not detected and corrected by the anesthesia professional, they may lead to an adverse outcome for the patient. We consider this process in detail.
Problems Often Result from Latent Underlying Conditions
The events that trigger problems do not occur at random. They emerge from three sets of underlying conditions: (1) latent errors , (2) predisposing factors , and (3) psychological precursors .
Latent errors, as described by Reason, are “…errors whose adverse consequences may lie dormant within the system for a long time, only becoming evident when they combine with other factors to breach the system’s defenses. [They are] most likely to be spawned by those whose activities are removed in both time and space from the direct control interface: designers, high-level decision makers, construction workers, managers, and maintenance personnel.”
Such latent errors exist in all complex systems. Reason describes them as “resident pathogens,” which, like microorganisms in the body, remain under control until sets of local circumstances “combine with these resident pathogens in subtle and often unlikely ways to thwart the system’s defenses and bring about its catastrophic breakdown” ( Fig. 1-2 ).
In anesthesia, latent errors can result from administrative decisions regarding scheduling of cases, assignment of personnel to staff them, and the priorities given to such things as rapid turnover between cases. They can also result from the design of anesthesia equipment and its user interfaces or how drug vials and ampules are designed and labeled or supplied to the anesthesiologist. Manufacturing defects and failures of routine maintenance are also latent errors.
Organizational Culture Factors
Safety in other industries of high intrinsic hazard is known to be a property primarily of systems rather than individuals. Organizations that perform successfully under very challenging conditions, with very low levels of failure, are termed “high reliability organizations” (HROs). The first HRO to be studied was the flight deck of aircraft carriers. Others include certain military organizations, commercial aviation, electric power grids, and firms handling large-scale electronic financial transactions. Based on direct observation of HROs, investigators have determined that a key element of high reliability is a “culture of safety” or a “safety climate” permeating the organization. Several features of safety culture or climate are as follows:
A commitment to safety is articulated at the highest levels of the organization and translated into shared values, beliefs, and behavioral norms throughout all organizational levels.
The organization provides the necessary resources, incentives, and rewards to allow this to occur.
Following standard operating procedures and safety rules is a part of the behavioral norms.
Safety is valued as the primary priority, even at the expense of “production” or “efficiency.” Personnel are rewarded for erring on the side of safety, even if they turn out to be wrong.
The organization proactively manages safety and carefully monitors ongoing safety processes and operating procedures.
Communication among workers and across organizational levels is frequent and candid.
Unsafe acts are rare, despite high levels of production.
There is openness about errors and problems; they are reported when they occur.
Organizational learning is valued; the response to a problem is focused on improving system performance.
Think for a moment how your organization compares to these safety ideals and where it could improve its performance. To the extent that a health care organization or work unit maintains a culture of safety, it can reduce the occurrence of latent errors and bolster flexible defenses against the accident sequences that do get started. However, there are many challenges to a culture of safety, particularly the erosion of safety in the search for throughput and revenue. Such forces can lead to “production pressure” , —internal or external pressure on the anesthesia professional to keep the OR schedule moving along speedily, with few cancellations and minimal time between cases. When anesthesia professionals succumb to these pressures, they may fail to perform adequate preoperative evaluation and planning or neglect to conduct pre-use checks of equipment. Even when preoperative evaluation does take place, overt or covert pressure from surgeons (or others) to proceed with elective cases despite the existence of serious or uncontrolled medical problems can cause anesthesia professionals to do things that are unsafe.
In 1994 we conducted a survey of California anesthesiologists concerning their experience with production pressures. We found that 49% of respondents had witnessed an event in which patient safety was compromised owing to pressure on the anesthesiologist. Moreover, 32% reported strong to intense pressure from surgeons to proceed with a case they wished to cancel; 36% reported strong to intense internal pressure to “get along with surgeons”; and 45% reported strong pressures to avoid delaying cases. Significantly, 20% agreed with the statement, “If I cancel a case, I might jeopardize working with that surgeon at a later date.” The economic pressures are obvious.
Production pressure also leads to haste by the anesthesia professional, which is another psychological precursor to the commission of unsafe acts. In the survey, 20% of respondents answered “sometimes” to the statement, “I have altered my normal practices in order to speed the start of surgery,” while 4% answered “often” to this statement, and 20% of respondents rated pressure by surgeons to hasten anesthetic preparation or induction as strong or intense.
Comparable results were found in a survey of residents in anesthesiology. In a similar survey conducted by Johnson in 2001, such pressures and experiences were again found for anesthesiologists. Although the study has not been repeated in nearly 20 years, we think that production pressures have only increased in the interim.
We also have conducted surveys across all hospital employees in multiple institutions in studies involving tens of thousands of personnel, and have documented that production pressures exist throughout the hospital and are not unique to anesthesiology. , , Moreover, we found a threefold greater rate of responses indicative of a lack of safety culture for health care personnel (18%) than for naval aviators (6%) given matched questions concerning safety culture. , Thus health care institutions do not yet have as strong a culture of safety as they should.
Local Predisposing Factors and Psychological Precursors
The final set of underlying features consists of latent psychological precursors, which predispose the anesthesia professional or surgeon to commit an unsafe act that triggers a problem. The primary psychological precursors are traditionally referred to as performance-shaping factors , and include such elements as fatigue, boredom, illness, drugs (both prescription and recreational), and environmental factors such as noise and illumination. Factors of work culture in general, and safety culture in particular, are also important to consider. Different combinations of psychological factors are discussed in detail in a number of review articles and general strategies to deal with performance-shaping factors and safety culture are discussed in Chapter 2 .
Each problem is initiated through one or more triggering events. Historically, anesthesia professionals have been most concerned about events that they create themselves, such as esophageal intubation or drug swaps, but these are relatively rare compared with events that are triggered in other ways. Triggering events can come from (1) the patient, (2) the surgery, (3) the anesthesia, or (4) the equipment.
Many problems occur de novo owing to the underlying medical pathology of the patient. For example, studies of myocardial ischemia in the perioperative period demonstrate that ischemia often occurs without any significant change in hemodynamic status or any known causal effects of anesthesia.
Surgical stimulus alone is a profound trigger of many physiologic responses, including hypertension, tachycardia, laryngospasm, and bronchospasm. Problems linked to the patient’s medical pathology may be precipitated by routine actions of the surgeon. Also, unplanned events, such as surgical compression of organs or transection of vital structures, can rapidly evolve into serious problems.
Induction and maintenance of anesthesia can precipitate problems in patients even in the absence of significant underlying medical disease. Actions or errors on the part of the anesthesia professional can directly jeopardize the patient, as when central venous cannulation causes a pneumothorax. An operation may require standard but complex maneuvers that may trigger problems, such as turning the patient to the prone position. Especially in patients under general anesthesia and with neuromuscular blockade, the body’s own protective mechanisms are blunted or obliterated, making the patient more vulnerable to the anesthesia professional’s actions.
General anesthesia is maintained and the patient’s vital functions are monitored by using electromechanical equipment. Should this equipment fail, the patient can suffer irrevocable harm. However, it is very rare for an equipment malfunction, by itself , to harm the patient immediately. Examples of this might include electrocution, fires, and airway overpressure events. More typically, an equipment failure stops a life support or monitoring function, which can in theory be performed in another way if the failure is identified and the necessary backup systems are available and functional. Equipment problems often contribute to difficulties in handling other problems, either because they usurp the attention of the anesthesia professional or because the treatment of the main problem requires the use of equipment that itself has failed.
Prevention of Problems
Reducing or eliminating the latent factors that predispose patients to problems would be the most effective strategy to improve safety. This might include changes in organizational structure, work procedures, safety culture, or staffing, However, most of these factors are the result of a complex evolution of medical practice and economics in combination with historical and political factors; identifying and changing them is a difficult, slow, and frustrating process. Moreover, there are external circumstances that cannot be controlled even in principle (e.g., trauma, disasters, terrorism). Therefore the individual anesthesiologist must adopt effective strategies for preventing problems targeted at individual cases, making specific checks for triggering factors, and taking corrective action as necessary. Such pre-case checks include (1) the patient, (2) the surgeon and anesthesia professional, and (3) the equipment.
The anesthesia professional begins by using traditional forms of medical decision making in preoperative evaluation of the patient and in planning the anesthetic. During this evaluation, the anesthesia professional considers the patient’s medical status, the urgency of the surgery, and whether any further treatment can make the patient a better risk. This is a crucial opportunity for the anesthesia professional to prevent adverse outcomes for the patient. If surgery can proceed, there may still be additional preventive measures that should be implemented to deal with specific medical conditions (e.g., rapid sequence induction [RSI] in patients with a bowel obstruction) or to prepare for specific surgical procedures (e.g., use of a double lumen endotracheal tube for thoracic surgery). In Chapter 2 , we will emphasize the necessity of developing a sound anesthetic plan for the patient that takes each of these measures into account. In many cases, however, there are competing goals, which will prevent a perfect plan from being developed. The optimal plan in such cases must be a compromise among the various risks, benefits, and costs.
The Surgeon and Anesthesia Professional
Surgeons and anesthesia professionals have a duty to perform their jobs with appropriate care and skill. They must honestly determine whether their own ability, fitness, and preparation match those demanded by the planned procedure. Chapter 2 covers in detail how anesthesia professionals can deal with possible deteriorations in their performance so as to prevent harm to the patient.
A thorough pre-use check of critical life-support equipment is considered mandatory before the induction of anesthesia. The anesthesia machine encompasses O 2 delivery, ventilation, and gaseous anesthesia delivery systems. Monitors incorporate many alarms that need to be set correctly and turned on. When intravenous (IV) infusions are a primary component of the anesthetic or hemodynamic management, these are not coupled to the anesthesia machine, and they need to be checked thoroughly as well. In addition, the anesthesia professional should ensure that appropriate backup equipment is available for all life-critical functions (e.g., a self-inflating bag for ventilation).
Problems Will Inevitably Occur Despite Attempts to Prevent Them
Despite attempts to prevent the occurrence of problems during anesthesia, experience shows that problems of varying severity occur in a large percentage of cases. The exact frequency of problems is not known. Existing studies probably have underestimated the occurrence of problems because they depended on written reports of the case by the anesthesia professionals rather than real-time, objective recording of the events. Despite their limitations, two studies offer some data concerning the frequency of problem events.
In the Multicenter Study of General Anesthesia, 17,201 patients received general anesthesia under specific protocols with random stratification to receive one of four anesthetic techniques (each of the three common volatile anesthetic agents or narcotics plus nitrous oxide). The anesthesia professionals observed the patients for the occurrence of any of a large variety of carefully defined perioperative “outcomes,” that is, adverse events ranging from minor, such as sore throat or hypotension (i.e., systolic blood pressure reduced by more than 20% from baseline), to serious, such as myocardial infarction (MI) or death. Based on our definition of a “problem,” most of the outcomes measured in this study would constitute a perioperative condition that could evolve into one that might harm the patient. The researchers observed 34,926 defined outcomes in the 17,201 patients. Clearly, some patients experienced more than one outcome while others had none, but 86% of patients faced at least one undesirable outcome. Although most events were minor and caused no injury to the patient, over 5% of patients had one or more severe events requiring “significant therapy, with or without full recovery.” This incidence is probably a lower limit for severe events because the entry criteria of the study excluded critically ill patients and emergency cases in which evolving problems of a severe nature more likely might occur.
In another study, Cooper and associates found that “impact events,” which were “undesirable, unexpected, and could cause at least moderate morbidity,” occurred in 18% of patients either in the OR or in the postanesthesia care unit (PACU) and 3% of all cases involved a “serious” event. These, too, are probably lower limits, because, for technical reasons, the study excluded patients electively destined for an intensive care unit (ICU). Extrapolating from both of these studies, it seems that at least 20% of cases involve a problem event requiring intervention by the anesthesiologist, and approximately 5% of cases involve a potentially catastrophic event. The actual frequency of problems may be higher in practice settings with greater than average case complexity.
The exact mortality rate of patients related to anesthesia care is uncertain. Deaths resulting from administration of anesthesia in relatively healthy patients undergoing routine surgery are rare, but even for this restricted category the rate is approximately 1 per 250,000 (0.4 per 100,000 cases). Overall, perioperative deaths related to anesthesia may be as high as 1 per 1400 cases. This makes the frequency of a fatal accident in surgery attributable solely to anesthesia 45 times higher than that of any fatal airline accident in the United States. There are about 25,000 to 30,000 airline flights per day in the United States and there are very few serious incidents or accidents, although the exact number is unknown. The total accident rate from all causes (not including terrorist acts) for scheduled airline flights from 2002 through 2011 was 0.29 per 100,000 departures. Air carrier accidents with one or more fatalities occurred at a rate of 0.009 per 100,000 departures. In fact, according to the National Transportation Safety Board, in the years 2007 to 2011 there was only a single fatal airline accident! ( www.ntsb.gov/data/table6_2012.html ). Relative to aviation, anesthesia has a long way to go to optimize patient safety.
How Do Problems Evolve into Adverse Outcomes?
Once a problem occurs, there are various possibilities for its future evolution. The problem may be self-limited or may continue to exist without any threat to the patient. It can increase in severity. It can trigger new problems (cross-triggering) within the patient or within the anesthesia/surgery system; the new problems may be more threatening than the original problem. Multiple small problems in several different subsystems can together create a more serious situation than any one of them alone (combination). A problem triggered by one factor may interfere with the management of problems triggered by others (triggered failure of recovery), or it may distract the anesthesia professional’s attention from other more serious problems.
Although there are no universally accepted criteria for categorizing the states of evolution of a perioperative problem, we term the next state an incident —a problem that will not resolve on its own and is likely to continue evolving. A critical incident is an incident that can directly cause an adverse patient outcome. Further information on the nature of critical incidents comes from studies by Cooper and colleagues at Massachusetts General Hospital in Boston. These studies pioneered the investigation of incident pathways and collected data both retrospectively and prospectively on critical incidents, which they defined as
… a human error or equipment failure that could have led (if not discovered or corrected in time) or did lead to an undesirable outcome ranging from length of hospital stay to death.
Each event was categorized by its primary cause: human error, equipment failure, disconnection (a special type of equipment failure), or other. For human errors, a distinction was made between technical errors in performing appropriate actions and judgment errors in which the actions occurred as planned but were inappropriate. In addition to these categorizations, the authors collected information on a variety of “circumstances that conceivably could have contributed to the occurrence of an error or to a failure to promptly detect an error;” these were termed associated factors .
Table 1-1 shows the distribution of the 25 most frequent critical incidents reported in these studies. Note that the true incidence of such events is unknown because the denominator of total cases from which these events were taken is unknown. Although the distribution of events may have changed somewhat since the original studies in the late 1970s and mid-1980s, critical incident studies have been repeated in many settings and countries during the ensuing 3 decades, with similar results.