Discussion

1114237 - Jones & Bartlett Learning ©

T

© ALMAGAMI/Shutterstock

CHAPTER 9

Assessing Risk and Preventing Harm in the Clinical Microsystem Paul Barach and Julie K. Johnson

To err is human, to cover up is unforgivable, and to fail to learn is inexcusable.

— Sir Liam Donaldson

he framework of health care delivery is shifting rapidly. Capital budgets and operational efficiency

1114237 - Jones & Bartlett Learning ©

are critical in this time of shrinking reimbursement, increasing share of risk, and evolving models of care delivery. Modern medical care is complex, expensive, and at times dangerous. Across the world, hospital patients are harmed 9.2% of the time, with death occurring in 7.4% of these events. Furthermore, it is estimated that 43.5% of these harm events are preventable (de Vries et al., 2008; Landrigan et al., 2010). As illustrated in FIGURE 9.1, medical error is now the third-leading cause of death, following heart disease and cancer (Makary & Daniel, 2016). The rates could be debated, as they depend on the methods used in the studies as well the levels of underreporting; however, what is clear is that we need to change and improve our health care systems dramatically. Most significantly, the study of patient safety has identified the need to design better systems—ones that make sense to providers and respect their work—to prevent errors from causing patient harm (Barach & Phelps, 2013). From all perspectives—patients’ and providers’ as well as the health care system’s—the current level of avoidable harm is unacceptable and financially ruinous.

1114237 - Jones & Bartlett Learning ©

FIGURE 9.1 Leading Causes of Death

Reproduced from Makary, M. A. & Michael, D. Medical error— the third leading cause of death in the US. BMJ, 353, i2139.

The goal of this chapter is to discuss how to develop reliability and resilience in health care systems while embedding a vision of zero avoidable harm into the culture of health care. Achieving this vision requires a risk management strategy that includes: (1) identifying risk—finding out what is going wrong; (2) analyzing risk —collecting data and using appropriate methods to understand what it means; and (3) controlling risk— devising and implementing strategies to better detect,

1114237 - Jones & Bartlett Learning ©

manage, and prevent the problems that produce harm (Dickson, 1995).

This chapter begins with the background and definitions of risk management. We discuss the universal ingredients of individual accidents, organizational accidents, and human error. We then discuss models of risk management, focusing on how to address the barriers to creating an effective culture of safety and engagement and present strategies for applying concepts to clinical microsystem settings. The chapter concludes with a discussion on the role of disclosure of adverse events as part of a comprehensive risk management strategy.

1114237 - Jones & Bartlett Learning ©

▶ Risk Management—Background and Definitions

Traditionally, risk has been seen as exposure to events that may threaten or damage the organization (Walshe, 2001). Essentially, risk is the chance of something happening that will have an impact on key elements. It can be measured in terms of consequences and likelihood. The task of risk management, therefore, becomes balancing the costs and consequences of risks against the costs of risk reduction. The goal in clinical risk management should be to improve care and protect patients from harm; however, a perverse incentive for the organization as well as for those who work within the organization is that risk management may become a financially driven exercise that is at odds with the clinical mission of the organization (Vincent, 1997). Clinical risk management is the culture, process, and structures that are directed toward the effective management of potential opportunities and adverse events. We measure risk in terms of the likelihood and consequences of something going wrong, which is in contrast to how we measure quality (i.e., the extent to which a service or product achieves a desired result or outcome). Quality is commonly measured by whether a product or service is safe, timely, effective, efficient, equitable, and patient-centered (IOM, 2001).

1114237 - Jones & Bartlett Learning ©

There is substantial evidence that the majority of harm caused by health care is avoidable and the result of systemic problems rather than poor performance or negligence by individual providers. The past several years have seen an increase in proposed or operative legislation, including a near-miss reporting system, changes in mandated reporting systems, and the creation of state agencies dedicated to the coordination of patient safety research and implementation of change. There is clear policy guidance and a compelling ethical basis for the disclosure of adverse events to patients and their families. Mechanisms have been developed to involve all stakeholders—the government, health care professionals, administrators and planners, providers, and consumers—in ongoing, effective consultation, communication, and cultural change. These cultural and regulatory devices have helped engender more trust, which is an essential element in developing and sustaining a culture of safety.

1114237 - Jones & Bartlett Learning ©

Accidents Research in managing risk has focused on the culture and structure of the organization. Perrow (1984) advanced the Normal Accidents Theory, which describes accidents as inevitable in complex, tightly coupled systems such as chemical plants and nuclear power plants. These accidents occur irrespective of the skill and intentions of the designers and operators; hence, they are normal and difficult to prevent. Perrow further argues that as the system gets more complex, it becomes opaque to its users, so that people are less likely to recognize and fear potential adverse occurrences. There are three universal ingredients of accidents:

1. All human beings, regardless of their skills, abilities, and specialist training, make fallible decisions and commit unsafe acts. This human propensity for committing errors and violating safety procedures can be moderated by selection, training, well-designed equipment, and good management, but it can never be entirely eliminated.

2. All man-made systems possess latent failures to some degree. This is true no matter how well designed, constructed, operated, and maintained a system may be. These unseen failures are analogous to resident pathogens in the human body that combine with local triggering factors (e.g., life stress, toxic chemicals) to overcome the immune system and produce disease. Adverse events such as

1114237 - Jones & Bartlett Learning ©

cancers and heart attacks in well-defended systems do not arise from single causes but from multifactorial hits that overwhelm the system’s natural defenses. The adverse conjunction of such factors, each necessary but insufficient alone to breach the defenses, are behind the majority of adverse events.

3. All human endeavors involve some measure of risk. In many cases, the local hazards are well understood and can be guarded against by a variety of technical or procedural countermeasures. No one, however, can foresee all the possible adverse scenarios, so there will always be defects in this protective armor.

These three ubiquitous accident ingredients reveal something important about the nature of making care safer (Bernstein, 1996). We can mitigate the risk of adverse events by process improvement, standardization, and an in-depth understanding of the safety degradations in systems, but we cannot prevent all risk. Embracing this uncertainty and implementing the training and curriculum to cope with it, as opposed to focusing solely on minimizing it, is essential in high- risk, highly coupled industries.

Outside health care, the most obvious impetus for the renewed interest in human error and impact of the organizational dynamics has been the growing concern over the terrible costs of human errors that led to disasters. Examples include the Tenerife runway collision in 1977 (540 fatalities), the Three Mile Island

1114237 - Jones & Bartlett Learning ©

nuclear disaster in 1979, the Bhopal methyl isocyanate tragedy in 1984, the Challenger space shuttle accident in 1986 (Vaughn, 1996), the Chernobyl nuclear disaster in 1986 (Reason, 1997), and the Deepwater Horizon disaster in 2010 (Ingersoll, Locke, & Reavis, 2012). There is nothing new about tragic accidents caused by human error, but in the past, the injurious consequences were usually confined to the immediate vicinity of the disaster. Today, the nature and scale of highly coupled, potentially hazardous technologies in society and hospitals means that human error can have adverse effects far beyond the confines of the individual provider and patient settings. For example, Pseudomonas aeruginosa contamination of a single damaged bronchoscope potentially infected dozens of patients (DiazGranados et al., 2009). In recent years, there has been a noticeable spirit of glasnost (openness) within the medical profession concerning the role played by human error in causing medical adverse events (Millenson, 1997).

1114237 - Jones & Bartlett Learning ©

The Organizational Accident Certain systems have been designed and redesigned with a wide variety of technical and procedural safeguards (e.g., operating rooms), yet they are still subject to accidents and adverse events (e.g., fire in the operating room, wrong-patient procedure, drug errors). These types of accidents and adverse events have been termed organizational accidents (Perrow, 1984; Reason, 1997). Organizational accidents are mishaps that arise not from single errors or isolated component breakdowns, but from the insidious culture change due to an accumulation of failures occurring mainly within the managerial and organizational spheres. Such latent failures may subsequently combine with active failures and local triggering factors by errant but well-meaning providers to penetrate or bypass the system defenses. In his classic study of surgical teams, Forgive and remember, Bosk (1979) demonstrates how the underlying culture acts to acculturate newcomers, encourage normalized deviance, and suppress external views for change while enabling substandard practices to go unchecked. Systemic flaws set up good people to fail (Vaughn, 1999). People often find ways of getting around processes that seem to be unnecessary or that impede the workflow. This accumulated and excepted acceptance of cutting corners or making work-arounds poses a great danger to health care over time. By deviance, we mean organizational behaviors that deviate from normative standards or professional expectations (e.g., low handwashing compliance before patient contact, minimal consultant oversight of hospital care on weekends, suppressing information about poor care,

1114237 - Jones & Bartlett Learning ©

and so forth). Once a professional group normalizes a deviant organizational practice, it is no longer viewed as an aberrant act that elicits an exceptional response; instead, it becomes a routine activity that is commonly anticipated and frequently used (Weick, 2009). A permissive ethical climate, an emphasis on financial goals at all costs and an opportunity to act amorally or immorally, all can contribute to managerial and clinician decisions to initiate deviance (Barach & Phelps, 2013).

The etiology of an organizational accident can be divided into five phases (Perrow, 1984):

1. Organizational processes giving rise to latent failures;

2. Conditions that produce errors and violations within workplaces (e.g., operating room, pharmacy, intensive care unit);

3. The commission of errors and violations by “sharp end” individuals (the “sharp end” refers to the personnel or parts of the health care system in direct contact with patients. In contrast, the “blunt end” refers to the many layers of the health care system that affect the individuals at the sharp end. These colloquial terms are more formally known as “active errors,” which occur at the point of contact between a human and some aspect of a larger system, and “latent errors,” which are the less apparent failures of the organization);

4. The breaching of defenses or safeguards; 5. Outcomes that vary from a “free lesson” to a

catastrophe.

1114237 - Jones & Bartlett Learning ©

Viewed from this perspective, the unsafe acts of those in direct contact with the patient are the end result of a long chain of events that originate higher up in the organization. One of the basic principles of error management is that the transitory mental states associated with error production—momentary inattention, distraction, preoccupation, forgetting—are the least-manageable links in the error chain because they are both unintended and largely unpredictable (Hollnagel, Woods, & Leveson, 2006). These errors have their cognitive origins in highly adaptive mental processes (Sagan, 1993). Such states can strike health care providers at any time, leading to slips, forgetfulness, and inattention. The system must be designed with in-depth defenses to mitigate the impact of these momentary lapses in awareness and to maintain the integrity of patient care.

1114237 - Jones & Bartlett Learning ©

Human Error and Performance Limitations Although there was virtually no research in the field of safety in medicine and health care delivery until the mid- 1980s, in other fields, such as aviation, road and rail travel, nuclear power, and chemical processing, safety science and human factor principles have been applied to understand, prevent, and mitigate the associated adverse events (e.g., crashes, spills, and contaminations). The study of human error and of the organizational culture and climate, as well as intensive crash investigations, have been well developed for several decades in these arenas (Rasmussen, 1990; Sagan, 1993). The presumption of the non-linear and increasing risk in health care, associated with an apparent rise in the rate of litigation in the 1980s and attributable to the media’s amplified attention to the subject, has brought medical adverse events to the attention of both clinicians and the general public (Kasperson et al., 1988).

In parallel with these developments, researchers from several disciplines have developed increasingly sophisticated methods for analyzing all incidents (Turner & Pidgeon, 1997). Theories of error and accident causation have evolved and are applicable across many human activities, including health care (Reason, 1997). These developments have led to a much broader understanding of accident causation, with less focus on the individual who commits an error at the “sharp end” of the incident and more on preexisting organizational and system factors that provide the context in which errors and patient harm can occur. An

1114237 - Jones & Bartlett Learning ©

important consequence of this has been the realization that rigorous and innovative techniques of accident analysis may reveal deep-rooted, latent, and unsafe features of organizations. James Reason’s “Swiss cheese” model captures these relationships very well (Reason, 1990). Understanding and predicting performance in complex settings requires a detailed understanding of the setting, the regulatory environment, and the human factors and organizational culture that shapes and influence this performance.

1114237 - Jones & Bartlett Learning ©

▶ Models of Risk Management Risk management deals with the fact that adverse events, however rare, do continue to occur. Furthermore, risk management as a concept means that we should never ignore the possibility or significance of rare, unpredictable events. Using a creative analogy, Nassim Nicholas Taleb (2010) argues that before Europeans discovered Australia, there was no reason to believe that swans could be any color but white. However, when they discovered Australia, they found black swans. Black swans remind us that things do occur that we cannot possibly predict. Taleb continues the analogy to say that his “black swan” is an event with three properties: (1) the probability is low based on past knowledge, (2) the impact of the occurrence is massive, and (3) it is subject to hindsight bias; that is, people do not see it coming, but after its occurrence, everyone recognized that it was likely to happen.

In general, risk management models take into consideration the probability of an event occurring, which is then multiplied by the potential impact of the event. FIGURE 9.2 illustrates a simple risk management model that considers the probability of an event (low, medium, or high) and the impact of the consequences (limited/minor, moderate, or significant). Assigning an event in one of the cells is not an exact science, but the matrix offers a guideline to an appropriate

1114237 - Jones & Bartlett Learning ©

organizational response to the risk. The individual unit or organization would need to determine how to translate each cell into action (e.g., what is required for “extensive management” of events with a significant negative impact and high likelihood of occurring).

FIGURE 9.2 Risk Management Model

1114237 - Jones & Bartlett Learning ©

▶ Engineering a Culture of Safety How do we create an organizational climate in health care that fosters safety? What are the ingredients and key values of a safety culture? While national cultures arise largely out of shared norms and values, an organizational culture is shaped mainly by shared practices. And practices can be shaped by the implementation and enforcement of rules. Culture can be defined as the collection of individual and group values, attitudes, and practices that guide the behavior of group members (Helmreich & Merritt, 2001). Acquiring a safety culture is a process of organizational learning that recognizes the inevitability of error and proactively seeks to identify latent threats. Characteristics of a strong safety culture include (Pronovost et al., 2003):

1. A commitment of the leadership to discuss and learn from errors

2. Communications founded on mutual trust and respect

3. Shared perceptions of the importance of safety 4. Encouragement and practice of teamwork 5. Incorporation of nonpunitive systems for

reporting and analyzing adverse events

Striving for a safety culture is a process of collective learning. When the usual reaction to an adverse incident

1114237 - Jones & Bartlett Learning ©

is to write another procedure and to provide more training, the system will not become more resistant to future organizational accidents. In fact, these actions may deflect the blame from the organization as a whole. There is a long tradition in medicine of examining past practices to understand how things might have been done differently. However, morbidity and mortality conferences, grand rounds, and peer reviews share many of the same shortcomings, such as a lack of human factors and systems thinking, a narrow focus on individual performance that excludes analysis of the contributory team factors and larger social issues, retrospective bias (a tendency to search for errors as opposed to the myriad system causes of error induction), and a lack of multidisciplinary integration into the organization-wide culture (Cassin & Barach, 2017).

If clinicians at the sharp end are not empowered by managerial leadership to be honest and reflective on their practice, rules and regulations will have a limited impact on enabling safer outcomes. Health care administrators need to understand the fundamental dynamics that lead to adverse events. Employing tools such as root cause analysis and failure mode and effects analysis can help clinicians and others better understand how adverse events occur, but only if done in an open and safe manner (Dekker, 2004). Collecting and learning from near-misses is essential. Definitions vary somewhat, but we define a near-miss as any error that had the potential to cause an adverse outcome but did not result in patient harm (Barach & Small, 2000). It is

1114237 - Jones & Bartlett Learning ©

indistinguishable from a full-fledged adverse event in all but outcome (March, Sproull, & Tamuz, 2003). Near- misses offer powerful reminders of system hazards and retard the process of forgetting to be afraid of adverse events. A comprehensive effort to record and analyze near-miss data in health care has, however, lagged behind other industries (e.g., aviation, nuclear power industry) (Small & Barach, 2002). Vital but underappreciated studies suggest that near-misses are quite common. In a classic study, Heinreich (1941) estimated there are approximately 100 near-misses for every adverse event resulting in patient harm.

A focus on learning from near-misses offers several advantages (Barach & Small, 2000):

1. Near-misses occur 3–300 times more often than adverse events, enabling quantitative analysis.

2. There are fewer barriers to data collection, allowing analysis of interrelations of small failures.

3. Recovery strategies can be studied to enhance proactive interventions and to deemphasize the culture of blame.

4. Hindsight bias is more effectively reduced.

If near-misses and adverse events are inevitable, how do we learn from them once they occur? The next section discusses an approach for learning from events using a microsystem framework.

1114237 - Jones & Bartlett Learning ©

1114237 - Jones & Bartlett Learning ©

▶ Applying Risk Management Concepts to Improving Quality and Safety Within the Clinical Microsystem

The clinical microsystem as a unit of research, analysis, and practice is an important level at which to focus patient safety interventions. A clinical microsystem has been defined as a small team of providers and staff providing care for a defined population of patients (Mohr, 2000; Mohr & Batalden, 2002; Mohr, Batalden, & Barach, 2004). Most patients and caregivers meet and work at this system level, and it is here that real changes in patient care can (and must) be made. Errors and failure occur within the microsystem, and ultimately it is the well-functioning microsystem that can prevent or mitigate errors and failure to avoid causing patient harm. Safety is a property of the clinical microsystem that can be achieved only through a systematic application of a broad array of process, equipment, organization, supervision, training, simulation, and teamwork changes. The scenario included in EXHIBIT 9.1 illustrates a patient safety event in an academic clinical microsystem and how the resulting analysis enables a microsystem to learn from the event. Throughout the story, as told from the perspective of a

1114237 - Jones & Bartlett Learning ©

senior resident physician in pediatrics, there are many process and system failures.

1114237 - Jones & Bartlett Learning ©

EXHIBIT 9.1 Patient Safety Scenario—Interview with a

Third-Year Pediatrics Resident

Resident: I had a patient who was very ill. He was 12 years old and was completely healthy up until 3 months ago, and since then has been in our hospital and two other hospitals pretty much the entire time. He has been in respiratory failure, he’s had mechanical ventilation (including oscillation), he’s been in renal failure, he’s had a number of ministrokes, and when I came on service he was having diarrhea—3 to 5 L/day—and we still didn’t know what was going on with him. We thought that an abdominal CT would be helpful, and it needed to be infused.

He was a very anxious child. Understandably, it’s hard for the nurses, and for me, and for his mother to deal with. He thought of it as pain, but it was anxiety, and it responded well to anxiolytics.

When I came in that morning, it hadn’t been passed along to nursing that he was supposed to go to CT that morning. I heard the charge nurse getting the report from the night nurse. I said, “You know that he is supposed to go for a CT today.” She was already upset because they were very short staffed. She heard me and then said that she was not only the charge nurse, but also taking care of two patients, and one had to go to CT. She went off to the main unit to talk to someone. Then she paged me and said, “If you want this child to have a scan, you have to go with him.” I said, “OK.” Nurses are the ones

1114237 - Jones & Bartlett Learning ©

who usually go. But it didn’t seem to be beyond my abilities . . . at the time.

So, I took the child for his CT and his mom came with us. We gave him extra ativan on the way there, because whenever he had a procedure he was extra anxious. When we got there, they weren’t ready. We had lost our spot from the morning. My patient got more and more anxious and was actually yelling at the techs, “Hurry up!” We went into the room. He was about 5 or 6 hours late for his study, and we had given him contrast enterally. The techs were concerned that he didn’t have enough anymore and wanted to give him more through his G-tube. I said, “That sounds fine.” And they mixed it up and gave it to me to give through his G-tube. I went to his side and—not registering that it was his central line—I unhooked his central line, not only taking off the cap but unhooking something, and I pushed 70 cc of the gastrografin in. As soon as I had finished the second syringe I realized I was using the wrong tube. I said, “Oh no!” Mom was right there and said, “What?” I said, “I put the stuff in the wrong tube. He looks OK. I’ll be right back, I have to call somebody.”

I clamped off his intravenous line and I called my attending and the radiologist. My attending said that he was on his way down. The radiologist was over by the time I had hung up the phone. My patient was stable the whole time. We figured out what was in the gastrografin that could potentially cause harm. We decided to cancel the study. . . . I sent the gastrografin—the extra stuff in the tubes—for a

1114237 - Jones & Bartlett Learning ©

culture just in case he grew some kind of infection and then we would be able to treat it and match it with what I had pushed into the line. I filled out an incident report. I called my chiefs and told them. . . . They said, “It’s OK. He’s fine, right?” I said, “Yes.” They came up later in the evening just to be supportive. They said, “It’s OK. It’s OK to make a mistake.”

Interviewer: What was your attending’s response?

Resident: The attending that I had called when I made the mistake said, “I’m sorry that you were in that situation. You shouldn’t have been put in that situation.” Another attending the next day was telling people, “Well, you know what happened yesterday,” as if it were the only thing going on for this patient.

I thought it was embarrassing that he was just passing on this little tidbit of information as if it would explain everything that was going on. As opposed to saying, “Yes, an error was made, it is something that we are taking into account.” And he told me to pay more attention to the patient. Yes, I made the mistake, but hands-down, I still—and always did—know that patient better than he did. I just thought that was mean and not fair. And the only other thing I thought was not good was the next morning when I was prerounding, some of the nurses were whispering, and I just assumed that was what they were whispering about. I walked up to them and said, “I’m the one who did it. I made a

1114237 - Jones & Bartlett Learning ©

mistake. How is he doing?” I tried to answer any questions they had and move on.

Interviewer: How did the nurses respond when you said that you made a mistake?

Resident: The nurse that had sent me down with him told me, “It’s OK, don’t worry about it.” The others just listened politely and didn’t say anything.

Interviewer: How did the mother respond to you the next day?

Resident: The next day, I felt really bad. I felt very incompetent. I was feeling very awkward being the leader of this child’s care—because I am still at a loss for his diagnosis. And after the event, when the grandma found out—she was very angry. I apologized to the mom, and I thought it would be overdoing it to keep saying, “I am so sorry.” So, the next day, I went into the room and said to the mom, “You need to have confidence in the person taking care of your son. If my mistake undermines that at all, you don’t have to have me as your son’s doctor, and I can arrange it so that you can have whoever you want.” She said, “No. No, it’s fine. We want you as his doctor.” Then we just moved on with the care plan. That felt good. And that felt appropriate. I couldn’t just walk into the room and act like nothing had happened. I needed her to give me the power to be their doctor. So, I just went and asked for it.

Many methods are available to explore the causal system at work (Dekker, 2002; Reason, 1995; Vincent,

1114237 - Jones & Bartlett Learning ©

2003; Vincent, Taylor-Adams, & Stanhope, 1998), and they all suggest the importance of holding the entire causal system in our analytic frame, not just seeking a “root” cause. One method that we have found to be useful for systematically looking at patient safety events builds on William Haddon’s (1972) overarching framework on injury epidemiology (Mohr et al., 2003). Haddon, as the first Director of the National Highway Safety Bureau (1966–1969), is credited with a paradigm change in the prevention of road traffic deaths and injuries moving from an “accidental” or “pre-scientific” approach to an etiologic one. Haddon was interested in the broader issues of injury and system failures that results from the transfer of energy in such ways that inanimate or animate objects are damaged. The clinical microsystem offers a setting in which this injury can be studied. According to Haddon (1970), there are a number of strategies for reducing losses:

■   Prevent the marshaling of the energy. ■   Reduce the amount of energy marshaled. ■   Prevent the release of the energy. ■   Modify the rate or spatial distribution of release of

the energy. ■   Separate in time and space the energy being

released and the susceptible structure. ■   Use a physical barrier to separate the energy and the

susceptible structure. ■   Modify the contact surface or structure with which

people can come in contact. ■   Strengthen the structure that might be damaged by

the energy transfer.

1114237 - Jones & Bartlett Learning ©

■   When injury does occur, rapidly detect it and counter its continuation and extension.

■   When injury does occur, take all necessary reparative and rehabilitative steps.

All these strategies have a logical sequence that is related to the three essential phases of injury control relating to preinjury, injury, and postinjury.

Haddon developed a 3 × 3 matrix with factors related to an auto injury (human, vehicle, and environment) heading the columns and phases of the event (preinjury, injury, and postinjury) heading the rows. FIGURE 9.3 demonstrates how the Haddon Matrix can be applied to analyze an auto accident (Haddon, 1972). The use of the matrix focuses the analysis on the interrelationship between the factors and phases. A mix of countermeasures derived from Haddon’s strategies is necessary to minimize injury and loss. Furthermore, the countermeasures can be designed for each phase—pre- event, event, and postevent—similar to designing mechanisms to preventing patient harm. This approach confirms what we know about adverse events in complex health care environments—it takes a variety of strategies to prevent and/or mitigate patient harm. Understanding injury in its larger context helps us recognize the basic nature of “unsafe” systems and the important work of humans to mitigate the inherent hazards (Dekker, 2002).

1114237 - Jones & Bartlett Learning ©

FIGURE 9.3 Haddon Matrix Used to Analyze Auto Accident

Haddon, W. A. (1972). Logical framework for categorizing highway safety phenomena and activity. J Trauma, 12(2), 197.

We can also use the Haddon Matrix to guide the analysis of patient safety adverse events. Adapting this tool from injury epidemiology to patient safety, we have revised the matrix to include phases labeled pre-event, event, and postevent instead of preinjury, injury, and postinjury. The revised factors, patient–family, health care professional, system, and environment, replace human, vehicle, and environment. Note that we have added a fourth factor, system, to refer to the processes and systems that are in place for the microsystem. This addition recognizes the significant contribution that systems and teams make toward harm and error in the microsystem.

“Environment” refers to the context (enablers and barriers) that the microsystem exists within. FIGURE 9.4

1114237 - Jones & Bartlett Learning ©

shows a completed matrix using the pediatric case. The next step in learning from errors and adverse events is to develop and execute countermeasures to address the issues in each cell of the matrix. FIGURE 9.5 provides a list of tools that would be appropriate to assess each “cell” of the matrix as part of a microsystems risk assessment framework.

FIGURE 9.4 Completed Patient Safety Matrix

1114237 - Jones & Bartlett Learning ©

FIGURE 9.5 Patient Safety Matrix with Tools for Assessing Risk and Analyzing Events

1114237 - Jones & Bartlett Learning ©

Microsystems and Macrosystems Health care organizations are composed of multiple, differentiated, variably autonomous microsystems. These interdependent small systems exhibit loose and tight coupling (Weick & Sutcliffe, 2001). Several assumptions are made about the relationship between these microsystems and the macrosystem (Nelson et al., 2002):

1. Bigger systems (macrosystems) are made of smaller systems.

2. These smaller systems (microsystems) produce quality, safety, and cost outcomes at the front line of care.

3. Ultimately, the outcomes from macrosystems can be no better than the microsystems of which they are formed.

These assumptions suggest that it is necessary to intervene within each microsystem in the organization if the organization as a whole wants to improve. A microsystem cannot function independently from the other microsystems it regularly works with or its macrosystem. From the macrosystem perspective, senior leaders can enable an overall patient safety focus with clear, visible values, expectations, and recognition of “deeds well done.” They can set direction by clearly expecting that each microsystem will align its mission, vision, and strategies with the organization’s mission, vision, and strategies. Senior leadership can offer each microsystem the flexibility needed to achieve its mission and ensure the creation of strategies, systems, and methods for achieving excellence in health care,

1114237 - Jones & Bartlett Learning ©

thereby stimulating innovation and building knowledge and capabilities. Finally, senior leaders can pay careful attention to the questions they ask as they nurture meaningful work and hold the microsystem’s leadership accountable to achieve the strategic mission of providing safer care.

TABLE 9.1 builds on the research of high-performing microsystems (Mohr, 2000; Mohr & Batalden, 2002; Mohr, Batalden, & Barach, 2004) and provides specific actions that can be further explored to apply patient safety concepts to understanding the impact and performance of clinical microsystems. It also provides linkages to the macrosystem’s ongoing organization- centered and issue-centered quality efforts, which can either support or conflict with this approach, as discussed in greater detail in Chapter 13. BOX 9.1 provides a set of accountability questions that senior leaders should ask as they work to improve the safety and quality of the organization.

1114237 - Jones & Bartlett Learning ©

TABLE 9.1 Linkage of Microsystem Characteristics to Patient Safety

Microsystem Characteristics

What This Means for Patient Safety

1. Leadership ■   Define the safety vision of the

organization

■   Identify the existing constraints within the organization

■   Allocate resources for plan development, implementation, and ongoing monitoring and evaluation

■   Build in microsystems participation and input to plan development

■   Align organizational quality and safety goals

■   Engage the board of trustees in ongoing conversations about the organizational progress toward achieving safety goals

■   Recognition for prompt truth-telling about errors or hazards

■   Certification of helpful changes to improve safety

2. Organizational support

■   Work with clinical microsystems to identify patient safety issues and make relevant local changes

■   Put the necessary resources and tools in the hands of individuals

1114237 - Jones & Bartlett Learning ©

3. Staff focus ■   Assess current safety culture

■   Identify the gap between current culture and safety vision

■   Plan cultural interventions

■   Conduct periodic assessments of culture

■   Celebrate examples of desired behavior, for example, acknowledgment of an error

4. Education and training

■   Develop patient safety curriculum

■   Provide training and education of key clinical and management leadership

■   Develop a core of people with patient safety skills who can work across microsystems as a resource

5. Interdependence of the care team

■   Build PDSA* cycles into debriefings

■   Use daily huddles to debrief and to celebrate identifying errors *PDSA: Plan, Do, Study, Act

6. Patient focus ■   Establish patient and family

partnerships

■   Support disclosure and truth around medical error

7. Community and market focus

■   Analyze safety issues in community, and partner with external groups to reduce risk to population

1114237 - Jones & Bartlett Learning ©

8. Performance results

■   Develop key safety measures

■   Create feedback mechanisms to share results with microsystems

9. Process improvement

■   Identify patient safety priorities based on assessment of key safety measures

■   Address the work that will be required at the microsystem level

10. Information and information technology

■   Enhance error-reporting systems

■   Build safety concepts into information flow (e.g., checklists, reminder systems)

1114237 - Jones & Bartlett Learning ©

BOX 9.1 Questions Senior Leaders Could Ask About

Patient Safety

What information do we have about errors and patient harm?

What is the patient safety plan?

How will the plan be implemented at the organizational level and at the microsystem level?

What type of infrastructure is needed to support implementation?

What is the best way to communicate the plan to the individual microsystems?

How can we foster reporting—telling the tru