In keeping with the message from the Institute of Medicine that approaching health care as a system will be critical to improving quality of care and patient safety, tomorrow's physicians should recognize that high-quality health care is delivered by systems.1 This means that physicians must understand how optimal health care both goes beyond what happens between physicians and patients and depends on many other individual and health system stakeholders, including the clinical teams that help deliver care, the administrators and organizations that support the delivery of health care, the individuals and organizations that finance health care, and the individuals and organizations involved in developing and disseminating new treatment options. Moreover, physician educators have been charged with incorporating systems-based approaches into medical education and residency training,2,3 in part to help future physicians understand how their ability to provide high-quality health care depends on these other individual and organizational stakeholders with whom and, in some cases, for whom they work.
However, asking physicians to develop a systems-based approach to health care is, in effect, also asking physicians to accept that they are accountable to multiple systems' stakeholders in some fashion.4 In particular, accepting a systems approach forces physicians to consider whether they have ethical commitments to other stakeholders, beyond commitments that derive from legal and regulatory accountability, and, if so, how they ought to balance these against their ethical commitment to their patients. It also forces physicians to recognize that the ethical practice of medicine requires attending to matters beyond what happens in the context of the physician-patient relationship. For instance, when care for a patient relies on teams of individuals interacting in complex organizations, the physician should consider what ethical commitments she or he has to other members of the team (e.g., treating them with respect) as well as what ethical commitments she or he has to the organization and others who rely on the organization (e.g., taking care to practice in a manner that acknowledges the importance of organizational financial solvency) alongside providing high-quality care, an ethical commitment the physician has to the patient.
Some express legitimate concern that asking physicians to pay attention to a wide array of stakeholder relationships beyond the physician-patient relationship will move these doctors far away from focusing their attention on the interactions and activities associated with the physician-patient relationship and, therefore, direct too much attention away from their fundamental ethical obligations that derive from holding the patient's welfare as their primary ethical commitment.5 Although developing a systems-based approach will likely lead to a broadened perspective with regard to ethical commitments, it is not inevitable that physicians will lose their important ethical commitment to their patients' welfare. In fact, as we argue below, recognizing and harnessing these broader ethical commitments will ultimately allow physicians to better fulfill their ethical commitments to their patients. Indeed, systems theories and their applications in organizational management and business ethics support the notions that individuals can maintain primary professional ethical obligations while working within complex systems and that organizational systems can, and should, be constructed to support ethical practice on the part of professionals.6–9
Lessons learned in the management and business ethics realms suggest that paying explicit attention to system features that either enhance or impede the fulfillment of professional ethical commitments to patients may be important to professionals who work in complex systems.6,7,9 Similarly, physicians must understand how to identify, manage, reduce, or eliminate challenges to the ethical practice of medicine that arise because of system features.1,9 Indeed, as Susan Wolf points out,
A truly systemic view of …health care [in the United States] considers how this set of individuals, institutions and processes operates in a system involving a complex network of interrelationships, an array of individual and institutional actors with conflicting interests and goals, and a number of feedback loops…. Progress in analyzing [ethical issues] …can only be made with a full understanding of the systemic issues.10
To facilitate this progress, physicians would benefit from both a systems-informed mental model and a type of moral imagination developed with keeping a systems approach to the ethical practice of medicine in mind. Ideally, a systems-informed mental model, as well as moral imagination, should be introduced in medical school and reinforced throughout residency training and lifelong learning.11
In this article, we introduce a systems-informed mental model that we believe is simple and easy to understand yet comprehensive. We argue that our model provides with a cognitive framework that can help physicians understand what a system is and that could be helpful regardless of what the actual health care system looks like in the future. We show how this model provides physicians with a framework for not only identifying the ethical and professional challenges presented by working within complex systems but also for understanding how these systems challenges may affect the welfare of patients. We describe how our systems-informed mental model, together with moral imagination and the values associated with professionalism, provide physicians with a roadmap that will be critical to their ability to practice within, and to lead improvements to, the systems needed to support their ethical practice of medicine, no matter how these future systems look.
To illustrate our ideas, we use a case example of a practicing physician who considers introducing clinical research into her practice. This case captures practical and ethical dilemmas that highlight important elements of our systems-informed mental model. It also demonstrates how these dimensions critically affect a physician's sense of professional identity and integrity as well as the welfare of patients.12–14 We have used slightly different versions of this case and discussed its implications and challenges with medical students and residents. We find that it resonates with these learners. Many of them easily identify similar situations from recent clinical rotations, and many also believe that it represents a common challenge for the foreseeable future. We correlate our ideas with this case throughout the rest of this essay.
Case. A physician hears from her colleague that his clinic has recently started hosting clinical trials. He finds that enrolling some of his patients in clinical research offers his patients some alternatives to standard medical treatment; his clinic an additional source of revenue; and himself, his colleagues, office staff, and any patients who participate a sense of helping to advance medicine. She thinks about several of her patients for whom standard medical care is not effective. They have told her about a national clinical trial that asks physicians to enroll patients in their office so that patients do not have to travel far away. They have asked her as their physician to look into joining this clinical trial so that they can participate. Until now, she has not wanted to because she was uncomfortable with the idea, in part because of a vague notion that “doctors ought not experiment on their patients” and in part because she did not think that her medical training was sufficient to allow her to undertake an activity as seemingly foreign as medical research. However, after talking to her colleague, she wonders whether she should reconsider.
A mental model is a concept described by cognitive scientists and refers to deeply held or ingrained assumptions or generalizations about external reality which can take the form of patterns or images that, in turn, frame and focus how we understand the world.15 Individuals construct mental models on knowledge and experience gained in the past and use them to understand the present, which allows them to make decisions and take action.16 Mental models are socially learned and inevitably incomplete.17 Because they are socially learned, they can be changed. Mental models both shape and influence how individuals experience and learn, and, indeed, individuals cannot make sense of their own experience except through mental models.15 Because mental models represent the ways we constitute our experiences—the ways by which the brain stores, manages, and retrieves data selectively—they are important to the learning process.
Peter Senge,18 Ralph Stacey,19 and other organizational theorists use the concept of mental models to help explain the human component of organizational and systems functioning and of systems change, or lack thereof. Because mental models are socially learned, people can share them with one another, and when they share mental models their need to communicate before making decisions or taking action decreases. Furthermore, once mental models become embedded and implicit, one uses them rapidly and repeatedly. Ralph Stacey notes that this rapid, repeated use is a critical component in the process of becoming an “expert.”19
However, the very benefits of mental models, which allow the use of heuristics to speed mental processing, thereby facilitating learning and the development of expertise, also represent dangers. For, when mental models become deeply ingrained, people rely on them without question. Yet, the very assumptions or generalizations used to construct mental models may have been too narrow or too focused, or the situation or experience may have been framed incorrectly.15,17 Thus, individuals may make decisions or take actions based on incomplete information or an incomplete picture of what is actually occurring. Or, the situation may have changed since they originally constructed their mental models. In these cases, mental models may blind people into factoring out (or not factoring in) what may turn out to be important to a given situation. They also may limit the possibilities one can envision to improve what one does, thus inhibiting needed change. For example, our physician has not envisioned that enlarging her practice to include activities associated with clinical research might benefit her patients.
The traditional mental model of physicians
Physicians have traditionally viewed the practice of medicine primarily in terms of physician-patient relationships and the interactions, activities, and duties associated with this relationship. This is not surprising, because the relationship between patient and caregiver is fundamental to providing and receiving excellent care. The importance of the physician-patient relationship naturally leads to a mental model of the practice of medicine in which the interactions between physicians and patients figure largely, and interactions outside the physician-patient relationship are viewed as external to the practice of medicine per se—a mental model that medical education and residency training largely foster and reinforce. However, the importance of the interactions between physicians and patients should not preclude developing a mental model of the practice of medicine that is capable of including the complexities of the modern health care system.
Many current efforts at reforming health care are grounded in systems theories and urge a systems approach to organizational change.1,9 Yet, many physicians struggle with adopting a point of view that seems to present real or perceived conflicts to their ethical duties to prioritize their patients' welfare. Accepting that the system is important implies that one might begin to owe duties to the system, a possibility that may keep physicians from embracing the notion of a systems-based practice that seems to threaten the primacy of their relationship with patients.5
However, there is danger in a mental model that underemphasizes the other interactions necessitated by complex modern health care systems. The physician who sees an encounter with a patient as isolated from the systems that surround the encounter may be blind to the reality of what is happening to the patient. As Donald Berwick20 points out, in a systems-based practice, drawing tight boundaries within and among the roles of health care professionals and other important stakeholders might lead to a failure to cooperate or a failure to understand systemic implications, which could lead to a systems failure. Systems failures are not only wasteful, they also could harm the patient. From the viewpoint of the person served—the patient—the performance of a system “depends far more crucially on how elements work together than on how each element, in its role, performs separately.”20 For instance, if roles within a health care team are tightly constructed, a team member may not do something that should be done on behalf of a patient because that action does not fit comfortably within that team member's perceived role.
But, even in today's complex health care system, the physician-patient relationship remains important to patients and society. If this relationship is damaged by the systems of care in which physicians and patients interact, then the overall quality of health care suffers. If physicians want to protect this relationship, then they should learn how systems function as well as ways in which they can intervene appropriately to improve these systems. Moreover, if the ethical practice of medicine depends on the system in which patient and physician interact, then physicians should see that improving systems of care is an integral part of ethical practice. Accomplishing these two things requires that physicians change their mental models related to the ethical practice of medicine from that of a series of dyadic relationships to one that incorporates systems thinking.
Case correlation. In our example, a mental model of physician practice embedded primarily in relationships with patients led the physician to see her reservations based strictly on her professional and ethical obligations to her own patients and on her own professional identity and competence. But, in the era of evidence-based medicine (EBM), the ability of a physician to fulfill professional and ethical obligations to her patients relies heavily on having appropriate evidence on which to base clinical recommendations. And, having such evidence depends in part on clinical research that is conducted by physicians in real-world settings, with medically ill individuals as research participants. Although not every physician and patient must participate in clinical research, some must, or else EBM becomes a farce.
As the clinical research enterprise gears up to respond to the challenge of producing more relevant and useful information, practicing physicians will increasingly face the dilemma that our physician faces. And we argue that it is not sufficient to avoid confronting the ethical and practical complexities associated with deciding whether and how to participate in the clinical research enterprise by relying only on the notion that it is not part of one's existing mental model of what it means to be a physician. Rather, physicians need to start thinking systemically in terms of “what would it mean to participate, or not participate, in this?” This would alert them to a number of systemic issues with practical and ethical implications, among them possible changes to practice as a result of adding clinical research, as well as issues concerning conflicts of interest and commitment as discussed in more depth below. The former might include conflicts with patient insurers or conflicts that arise if enrolling patients in clinical trials pays the physician more than providing the patients with clinical care. The latter might include the physician's conflict between providing medical care to patients and attending to those in trials or allocating finite financial resources to either clinical research or clinical care.
Towards a systems-informed mental model for health care: Understanding the characteristics of a system
A systems-informed mental model for health care relies on knowing and understanding what a system is, how systems work, and what happens when multiple systems interact. This entails understanding systems from a basic and functional perspective, rather than simply recognizing the various stakeholders (e.g., patients, families, other community members, physicians, nurses, other patient-care team members, managers and administrators, payers, pharmaceutical and device industries, hospitals, medical schools, governments, lawyers, professional organizations) in the current health care system and their potential effects on the ethical practice of medicine.
Leading thinkers have used various models of systems and have highlighted different aspects of systems theories in describing organizations, groups of organizations, and organizational processes.18–25 We draw from these models their basic concepts and elements so that the mental model we introduce below is simple but comprehensive enough to lay the groundwork for a systems-informed mental model for physicians.
A system has been defined as “a complex of interacting components together with the relationships among them that permit the identification of a boundary-maintaining entity or process.”25 The key characteristics of our systems-informed mental model for health care are (1) purpose(s) or goal(s), (2) boundaries, (3) resources, (4) interactions, (5) outcomes, (6) effectiveness or efficiency in achieving outcomes, and (7) ability to evolve. Below, we briefly describe these characteristics and why they are important to a systems-informed mental model in health care.
Purpose(s) or goal(s).
The scope of a system may be micro (small, with few interactions and intersections, such as a health care treatment team), mezzo (intermediate interactions and intersections, such as a hospital) or macro (large, complex, consisting of a large number of interactions and intersections, such as the health care delivery system which includes hospitals, clinics, payers, and clinical teams). But systems, whether micro, mezzo, or macro, are intended to fulfill one or more purposes or goals. The intended purposes or goals of a system help define its normative dimensions,26 allowing judgments about whether the intended purposes or goals are appropriate and whether the system is fulfilling them.
When people work together, they can form a system (as opposed to a collection of individuals) in which the whole is greater than the sum of the individual efforts. This only occurs, though, if they have a common goal and a vision of how it is to be achieved.18,19,22 It is particularly important to be clear about the primary purpose or goal of a system because systems as well as individuals often have more than one purpose or goal, and these can conflict, producing confusion and/or affecting desired outcomes. Consider that health care organizations often embrace the twin goals of excellent care and cost-effective care. In practice, these two goals can and do conflict, resulting in confusion among professionals who are trying to deliver excellent care but are told that they are not as cost-effective as the organization would prefer. It is not difficult to imagine the ire, frustration, and even anger among professionals who are rebuffed as they seek to achieve what they imagine is the primary goal of the institution. Further, because the primary goal of the organization is unclear, there will be inconsistency among stakeholders as to how to judge the organization. Stakeholders will judge the organization according to their own priorities which might be, and often are, very different. Vendors, for example, may judge the organization well, in that the organization might be paying them on time, whereas the community might reach an entirely different judgment on the question of whether the organization is fulfilling its goal to provide care for the underinsured.
Further, a single person, team, or organization could work across several systems. If these systems have different purposes or goals and the individuals commit to these purposes or goals, they will likely experience conflicts of commitment. Furthermore, different systems use different incentives to promote various activities according to their own norms. Individuals working across multiple systems may not recognize the conflicts of interest that these varying incentives introduce, especially if they have never faced them before and do not understand the norms that make them okay within a particular system.
Case correlation. The primary purpose of the health care delivery system, in which our example physician practices, is to provide appropriate health care to patients. The primary purpose of the clinical research enterprise is to provide generalizable knowledge so that future patients may get appropriate care. These primary purposes carry different normative dimensions. Appropriate health care is intended to benefit the patient and is individualized as much as possible to each patient's preferences and circumstances. Research aimed to produce generalizable knowledge requires substantial standardization as dictated by scientific methodology and does not necessarily intend to benefit the research participant. Our physician, considering adding clinical research to her practice of medicine, must understand that medical care and clinical research are activities with different primary purposes, goals, and normative dimensions.13 She also must recognize that because these activities serve different fundamental purposes, they will present different conflicts of commitment and potential conflicts of interest than those she experiences in her current practice of medicine. She must determine whether she can accommodate both of these activities and their itinerant properties within her practice. Understanding that these activities rely on separate, though overlapping, systems will be an important first step in deciding where conflicts of interest or commitment arise and what to do with them when they are identified.
Although defining a system's boundaries is important to identifying it, this is not as straightforward as it may seem.19,26 A system may be relatively autonomous, functioning mostly independently of other systems, and, hence, it may have boundaries that are relatively easy to identify. More commonly, however, a system intersects with other systems—both horizontally and vertically, such that its boundaries can be difficult to discern. For instance, a billing system in a hospital intersects horizontally with hospital charts on which physicians record some of the information that payers use. On occasion, however, this billing system might intersect vertically with a collection agency. Nevertheless, clarifying boundaries allows identification of relationships with other systems, and this clarification becomes important to understanding how the systems work independently and together. For instance, physicians are used to thinking of the cardiovascular system as a defined system. The fact that the nervous system interacts with it functionally fits into the mental model of these systems as separate but related, as does the fact that both of these systems are parts of a system that we identify as a human being. The mental model of nonbiological systems should also include this dynamic type of understanding.
Case correlation. Different systems underlie the provision of clinical care and the conduct of clinical research. When both activities occur in a single physical location and are carried out by the same individuals, they do not become the same. There are still important distinctions between activities conducted for the purposes of clinical research and those conducted for the clinical care of the patient. The boundaries between these systems may dictate important practical and ethical features of these activities. For example, our physician may need to file results of laboratory tests and clinical rating scales separately in a participant's research charts, rather than in that individual's medical charts, enforcing the boundary between clinical care and research. This filing is largely practical (e.g., research records contain very specific information at specific time points), but it also serves an ethical function: it maintains confidentiality related to an individual's (perhaps private) decision to participate in research. At the same time, this separation would render the research results unavailable for medical decision making, so our example physician would have to make provisions for clinically significant results to be available for clinical decision making, thus allowing interactions between these fundamentally separate systems. The boundaries between these systems have other important implications, as described further below.
Resources include technology, infrastructure, and people with different knowledge and skills. This is what Donabedian21 refers to as “structure.” A system cannot achieve its purposes or goals if its resources are insufficient or inappropriately aligned. Organizations and individuals allocate resources in a manner that supports their understanding of the primary purpose or goal of the system. When people either do not understand what the primary purpose is or disagree with a goal, they can hold back on resources—devoting inadequate time, attention, or energy to an activity. Conflicting understandings about purposes and goals may also lead to misalignment of resources. Or, two competing commitments may both lack adequate resources. Further, some individuals may selfishly keep resources for themselves and, thus, contribute to a systems failure. For instance, an emergency department (ED) needs certain equipment and/or technology and the requisite number of people possessing certain skills. If these are missing or inappropriate, the ED will not achieve its goals of saving lives, helping accident victims, alleviating pain, or treating immediate medical emergencies. Similarly, the ED will fail to meet goals if key individuals do not allocate their time, energy, or other resources in a manner that aligns with others in the system, or if individuals do not contribute as anticipated.
Case correlation. The physician will have to decide whether she has the appropriate resources in her practice to achieve the goals of both clinical care and medical research. Does she need more resources? If so, what kinds of resources? Will she need people with different skills? What skills should they have? Can she retrain her current staff, or will she need new staff? What type of further training will she herself need? Will she need different equipment? Resource issues can also affect patients. For example, are her patients covered by their insurers when participating in a trial? What happens when the trial ends or if adverse events occur?
Interactions occur between components within a system as well as between and among systems, and they are a function of how these systems are designed. Interactions between components can advance or detract from a system's goals, so interactions are important to consider when determining whether a system can fulfill its purposes or goals.19,23,24
Interactions take place on a spectrum ranging from rigid to entirely flexible. If one desires the same outcome again and again and the same interactions are required to produce it, one can and probably should design a more rigid system. Such a system's interactions will be preordained and mechanical to produce the same outcome consistently. For instance, a hospital or practice will probably design its billing system, so its internal interactions are mechanical. Moreover, because a billing system has very specific goals and, as such, will probably produce very specific outcomes, it will most likely interact with other systems in a relatively rigid fashion.
But, in some instances, achieving the purposes or goals of a system requires greater flexibility.22 For instance, interactions in an emergency room may have multiple goals and multiple desirable outcomes, some of which cannot be anticipated a priori (e.g., patient preferences), and, thus, they should probably be allowed to be flexible. Flexibility allows for innovative ways to address new problems as well as recurring problems. Moreover, because there are a multiplicity of goals and desired outcomes, interactions with other systems should probably be more flexible. For instance, discharging a patient from the ED may be straightforward, or it may require involving different systems such as nursing homes, child welfare agencies, or the police. How a system is designed—that is, how interactions take place—is fundamental to whether a system is able to achieve its purposes or goals.
Case correlation. The systems required to provide high-quality health care and the systems required to produce high-quality research will inevitably interact when they are colocated in a physician's practice. Ideally, these interactions could enhance both systems. For example, practicing physicians and patients know best which clinical decision-making questions would benefit from systematic study. Clinicians and patients who participate in clinical trials of a new intervention learn early the ways in which it provides improvements over older ones (if, indeed, it does) as well as some of the side effects and how to manage them. And, this is done with the extra support and resources that are frequently available in the context of a clinical study that may not be generally available in clinical care. Lessons learned suggest that clinicians who are involved in showing that an intervention works or is better than another are quicker to incorporate these findings into their clinical practices.13 In these ways, interactions between the clinical research enterprise and the clinical care delivery system can enhance both.
At the same time, negative effects can result from the conduct of clinical research in a physician's office alongside the provision of clinical care. For example, patients may become confused about the goals of clinical research, mistakenly believing that a study is intended to provide direct benefit to them and that their physician believes that participating in the study is their best option for treating their illness, simply because their trusted physician discusses the trial with them and suggests that the study may be an option for them. This phenomenon, called “therapeutic misconception,” can lead to a sense of betrayal and profound distrust down the road.
The physician in our example must also be aware of how interactions of these two systems or of their components will affect her practice on a practical and ethical level. For example, interactions among the physician and her staff with patients participating in clinical trials will differ from interactions that take place in routine patient care. The physician will have to think carefully about how she and other members of her staff recruit patients into clinical research and their interactions with these patients. What information will she need to disclose? How will she disclose this information? Does she feel that only patients involved in a clinical trial need to know that she is also conducting clinical research in the practice, or does she want all patients to know this information?
Moreover, the physician must monitor how sponsors of clinical trials interact with her patients and her clinical practice and what effects these interactions will have on the practice and how they might enhance (or not) the purposes and goals of her practice. For example, how will she respond if a research sponsor requests to review the patient records of anyone who might meet a study's entry criteria and also offers to reimburse her for the staff time required to conduct this review? In making these decisions, she will have to think through practical and ethical dimensions related to how she will allow the systems of clinical research and clinical care to interact in her practice, as well as how these types of interactions will affect her staff, her patients, and, indeed, herself, as time goes on.
The outcomes produced by a system are difficult to predict and, in some cases, even to identify, because they may be unintended. This is particularly true for systems designed to be flexible.22 For instance, we doubt that the goal of managed care was to cause widespread mistrust in and suspicion of the health care delivery system. And unintended outcomes can cause a system to change. Widespread mistrust and suspicion caused many managed care organizations to change their delivery structures and to rethink and, in some cases, reprioritize, their goals.27 Once one identifies the outcomes associated with a system, one can use them to evaluate whether a system is fulfilling its purposes or goals. Moreover, by measuring outcomes against goals or purposes, resources, and interactions, one can determine a system's efficiency and effectiveness.
Case correlation. Once the physician has made up her mind to conduct clinical research in addition to providing medical care in her practice, she will continually need to assess whether her practice is achieving appropriate outcomes from both a practical and a ethical standpoint with regard to (1) her patients (e.g., providing appropriate care to patients, monitoring the safety of patients participating in research), (2) her business (e.g., achieving minimal financial stability to ensure continuity of her services, ensuring that staff members feel competent to carry out their duties and have the opportunity for growth and development), and (3) her commitments (e.g., to patient confidentiality, to producing good data for research studies, to truth telling and transparency to all stakeholders), and she can ask whether the means of achieving her goals are effective and efficient.
Effectiveness and efficiency.
From one perspective, effectiveness and efficiency can be considered evaluative criteria and categorized as one of the dimensions of “outcomes” or “interactions.” But we are talking about systems that are designed by people and rely on the interactions of people to fulfill goals, and people have the ability, to a greater or lesser degree, to change systems through their actions and interactions.19,22 Moreover, systems must be at least minimally effective and efficient in achieving their intended purposes. That is, once functioning, they must achieve at least some of their purposes or goals. If they do not, then they either will not survive or will change into something not originally envisioned, potentially with very different purposes or goals. For instance, the health care delivery system is changing (and, perhaps, changing its purposes and goals) in response to the perception that its outcomes are not currently produced efficiently or effectively.1 Because effectiveness and efficiency are critical to a system's sustainability as its components interact to achieve its purposes and goals, we include effectiveness and efficiency as necessary characteristics to keep in mind with regard to systems that rely on people.
Case correlation. If the physician chooses to host clinical trials in her practice, she will be spending less time on patient-care activities. She needs to think about the impact that adding clinical research activities will have on her efficiency and effectiveness in clinical care activities as well as that of her staff and indeed of her overall practice. She also must think about what it will take for her practice to achieve and maintain efficiency and effectiveness in clinical research, both practically and ethically, and whether she can access the resources necessary to do so. In addition, she needs to be aware that optimizing the efficiency and effectiveness of one pursuit may adversely affect the other, though this need not always be the case.
Ability to evolve.
Another key characteristic of human systems is that they can and most likely will change over time, for better or worse.18,22 Human systems can evolve in a positive direction based on the development of either a creative new way to achieve existing goals or a reassessment of purpose and realignment of resources towards a new purpose. This is true of both flexible and rigid systems. On the face of it, flexible systems have more potential to evolve and change in positive ways because one expects interactions to vary according to circumstance, and, for many human systems, flexibility and the potential to evolve in positive directions are very important to fulfilling the goals of the system.
Nevertheless, the desire for predictability and control can lead to designing systems more rigidly. However, even in a rigidly designed system, one cannot absolutely control human behavior or beliefs. This inability to control human behavior and beliefs gives human systems the ability to evolve in a negative manner as well. A person may simply decide not to do an expected task and so throw the system into disarray. If a person believes that the goal of a system is unattainable, the interactions of that person—not completing assigned tasks in an appropriate time frame or failing to communicate necessary information—may reflect that belief. In these circumstances, the outcomes produced by the system will probably not be consistent with the intended goals. If this happens over and over, the system will have effectively changed or evolved into something not earlier envisioned and, most likely, in a manner that is undesired. The fact that systems can change highlights the need for the other characteristics of a system to be aligned toward its purpose and goals, as well as for the need to monitor how a system is evolving.
Case correlation. There is nothing to prevent the physician from evolving and enlarging the goals of her practice, and, indeed, such a change may allow her to better fulfill her commitment to her patients. On the other hand, if bringing in clinical research is not done in a carefully considered way with attention to systemic elements, the introduction of clinical research could derail the practice, make staff uncomfortable with their duties, and create a situation in which patients ultimately suffer. Thus, the physician must also remain aware that adding this layer of complexity into her practice brings further ability for unexpected change and requires attention from her and her staff to ensure that the systems characteristics of her practice remain aligned as the changes unfold over time.
In summary, developing a systems-informed mental model means knowing these seven characteristics of a health care system— its (1) purpose(s) or goal(s), (2) boundaries, (3) resources, (4) interactions, (5) outcomes, (6) effectiveness and efficiency, and (7) ability to evolve—as each characteristic is important to the successful functioning of the system. A mental model that considers these characteristics can be helpful to physicians as it can help them locate themselves and their patients within the relevant system(s). Moreover, it will inform them of the location of others in the system(s). This is a mental model that physicians can use to advantage in any system, or overlap of systems, in which they find themselves. For example, in our case, it is critical that the physician understand that evolving her practice to include participation in the clinical research enterprise changes the systems that compose her practice and fundamentally affects her, her staff, and her patients.
Although understanding how a system works and locating the important stakeholders within the system are both important, improving a system requires additional steps. It requires recognizing that the system needs improvement, formulating a solution or alternative, and taking action either to implement or to advocate for the implementation of the solution or alternative.
Professionalism: Recognizing that the system needs improvement
Physicians do, or should, ground their purpose and goals in their understanding of professionalism and ethical practice. By professionalism, we mean the ethical underpinnings of professional practice and professional integrity and the ethical commitments implied, rather than the sociological concept of professionalism.28–30 The specifics of these commitments have changed with time and are in the midst of significant reevaluation. Nevertheless, they generally include obligations to patients as well as obligations, frequently less well defined, to other individuals (other professionals, trainees, community members, etc.) and to larger groups (the community, the public or society at large, the profession).28–30 Professionalism informs and is informed by the ethical practice of medicine, and it concerns physicians' behavior and the virtues for which they should strive, including honesty, compassion, and trustworthiness. Professionalism gives physicians normative criteria, such as “the good of the patient,” by which to recognize an ethical conflict or dilemma in the practice of medicine, and criteria with which to judge how a system affects a patient. Moreover, physicians can also use the normative criteria supplied by professionalism to determine how a system should affect a patient.
Case correlation. As the push continues to increase the relevance and utility of research for practical clinical decision making by conducting research in real-life settings, practicing physicians will increasingly face the option of incorporating clinical research activities into their everyday practice of medicine.13 These opportunities to participate in the clinical research enterprise, which also offer patients the chance to enroll in research, will include studies sponsored by the National Institutes of Health and others sponsored by industry. Researchers will design some studies to contribute to the evidence base supporting clinical decisions, for example, through head-to-head trials of competing available medications; others will design studies for the approval of the U.S. Food and Drug Administration (FDA) to market a “me-too” medication that is fundamentally another version of an existing medication with the goal to win market share, rather than aiming to truly advance treatment options.
As we have demonstrated, any incorporation of clinical research into a physician's practice will present ethical challenges that arise from the interactions between systems designed to provide health care and those designed to conduct clinical research and will have the potential to profoundly change the nature of the physician's relationship with her patients. The goals, purposes, and desired outcomes of the health care delivery system writ large (which includes individual and organization providers, payers, professional associations, the government, suppliers, and others) differ from those of the clinical research system writ large (which includes individuals and organizations such as academic medical centers, pharmaceutical and medical device companies, the government, including the FDA, and others), even though some of the components of these two systems are the same (e.g., physicians, individuals with medical illness, clinical care settings). Thus, when our physician plays a role in both of these two systems (clinical care and research), she will experience some conflicts of commitment and potential conflicts of interest because she is participating in two systems. Similarly, the resources required to produce effective and efficient outcomes in these two systems differ, which the physician will also experience in her practice. These differences suggest that understanding the boundaries that naturally exist between these systems, the interactions created when they are juxtaposed within a single practice, and the potential for evolution resulting from the various interactions are important to preserving the physician's integrity as she navigates the conflicts of interests and commitment that arise.12–14 A systems-informed mental model helps the physician anticipate these ethical challenges and, together with the values associated with professionalism, provides a framework for thinking them through.
However, recognizing that there is a problem or ethical conflict within a system or created by the interaction of systems, whether at the micro, mezzo, or macro level, is only one step toward resolving the conflict. The next step entails using moral imagination, a concept we describe in detail below.
Moral Imagination: Formulating a Solution or Alternative to the Problem or Ethical Conflict
Building on Mark Johnson's31,32 work, Patricia Werhane15 defines moral imagination as
the ability in particular circumstances to discover and evaluate possibilities not merely determined by that circumstance, or limited by its operative mental models, or merely framed by a set of rules or rule-governed concerns.
Moral imagination differs from imagination in that moral imagination is framed by normative considerations that provide ethical guidance for recognizing a problem, grasping the problem from another perspective, considering alternatives, and exploring new avenues before formulating and evaluating a solution. For example, imagination might allow our physician to recognize that she could expand her practice by interacting with the clinical research enterprise. Moral imagination, however, constrains imagination by providing ethical guidelines as to how this expansion and subsequent interactions should take place.31 Thus, the use of moral imagination itself entails several stages.15 First, moral imagination is initiated by recognizing that an experience, situation, event, or outcome presents a moral problem. Professionalism can ignite moral imagination by providing the physician normative criteria by which to evaluate an experience, situation, event, or outcome and its operative mental models. Our physician is alerted to the dangers of expanding her practice because she continually assesses what is good for her patients and is unsure whether the conflicts of interest and commitment that will almost certainly arise from a clinical research will benefit them or, perhaps, expose them to unacceptable risks of harm.
Moral imagination also facilitates reasoning by helping individuals out of a specific mental model and helping them to redeploy their attention and criticize, revise, and create other mental models. Physicians can employ moral imagination and professionalism so that normative criteria govern their attention, criticisms, revisions, and creations. In this way, physicians do not revise an ethically problematic situation with an equally ethically problematic situation. In our example, the physician begins to ask how a relationship with the clinical research system might benefit her patients and might give her patients, her staff, and herself the opportunity to help advance medical knowledge. These activities lead to the development of more creative normative perspectives that enable one to find or devise solutions or alternatives that may be novel, morally justifiable, and economically viable. At this point, our physician understands that new processes and procedures might be necessary in her practice if her patients are to possibly benefit from, rather than be harmed by, her relationship with the clinical research establishment.
Through the use of moral imagination, physicians can begin critiquing the situation and understanding its dominant mental models and the alternatives that present themselves. During this process, they step out of the situation and their own mental models, working toward a solution that was not evident when the process began. Often, a solution will require a different mental model or worldview heretofore only latently available in the specific situation and context. The ethical values associated with medical professionalism are needed to guide a physician's use of moral imagination. These values help physicians recognize a problem or dilemma; they guide in critiquing, revising, and creating alternatives; and, finally, they help physicians evaluate solutions. A systems-informed mental model helps in the stages of formulation and evaluation of solutions because it gives the physician directions in which to look for new possibilities, and it allows a physician to recognize where the systems in which she interacts might need to be changed. For example, once our physician's practice has been expanded, if she encounters a problem, such as incomplete or misleading information being given to patients under the guise of an informed consent process, her knowledge of the characteristics of a system allows her to probe into the causes and origins of the problem.10 A systems-informed mental model encourages the physician to probe areas not commonly perceived as contributing to a conflict or offering a solution or alternative. Thus, combining a systems-informed mental model, professionalism, and moral imagination might give physicians a tool they can use to understand, work with, evaluate, and, ultimately, improve the systems of care that they rely on in their practice of medicine and that critically affect the welfare of their patients. But, there is a caveat.
Moral imagination requires the ability to disengage from the context in which one finds oneself. But no one can ever disengage completely from an event, conflict, or dilemma—largely because of the existence of mental models. Even as individuals revise, critique, create, and evaluate, they are still context bound. A person's context may be determined by historical circumstances, culture, surrounding political and social pressures, and values perspectives. Nevertheless, employing moral imagination, a systems-informed mental model, and the values associated with professionalism to critique and evaluate alternatives and solutions requires the individual at least to attempt disengagement.
Case correlation. Using moral imagination and a systems-informed mental model, together with the values associated with professionalism, our example physician can begin to probe where challenges to the ethical practice of medicine might arise in a reconfigured practice, and how she can solve them. For instance, to avoid confusion and misunderstanding as to the goals of her practice and her own ethical commitments, she can work to ensure that she hosts only studies that she believes are designed to advance clinical decision making or to provide novel treatment alternatives to patients. She can work to ensure that resources are available to supply information that is timely and transparent for patients and potential participants in clinical trials and for her practice to conduct both activities with utmost care.
Conclusion: The Physician's Role in Improving the Health Care System
We believe that in tomorrow's complex health care system, physicians should, as a matter of their professional ethical commitments, pay attention to and help to improve the systems that affect their ability to provide high-quality care to their patients. However, unlike the physician in our case example, practicing physicians often lack the knowledge, skills, and authority to change the systems in which they work. So, asking physicians to improve the systems of care in which they interact has the potential to create a certain amount of cynicism and frustration within the profession and about the profession. Nevertheless, preserving the ethical practice of medicine, protecting the physician-patient relationship, and preventing inadvertent harm to patients may require realigning or rearranging a given system. This may involve acting or advocating for a potential realignment or rearrangement of a system such that the systems of care that surround a patient reflect the ethical values associated with medical professionalism.
At the same time, we are not suggesting that physicians should shift their primary focus from the welfare of their patients, which is their chief ethical responsibility. But, in today's complex health care system, focusing only on the physician-patient relationship and the interactions and activities associated with it is not sufficient to ensure the welfare of patients. Thus, to improve systems of care that affect the welfare of patients, we suggest that current and future physicians
1. continue to pledge themselves to ethical values and goals associated with professionalism;
2. acquire a broadened perspective that incorporates an understanding of systems—a systems-informed mental model for health care;
3. cultivate a moral imagination that employs an understanding of systems; and
4. act or advocate as necessary for change in the systems that surround patient care.
Our suggestions may be met with skepticism; some may argue that physicians are already overstressed and that asking them to acquire a new perspective is wasteful, especially because many physicians do not have the authority to improve the systems within which they work. Our response is that if we want to improve the systems of care, if we want these systems to be consistent with our understanding of the ethical practice of medicine, then who else, other than physicians and physician-educators (who are themselves components of and resources within the system), can take the lead in reaching this goal? And how else can physicians become key players and, eventually, leaders, in continually improving these systems unless they develop a systems-informed mental model of the practice of medicine early in their professional identity formation and in their understanding of professionalism and continually practice applying moral imagination to systems-level problems? In this way, physicians can work to bring the systems they rely on to provide high-quality care for their patients in line with ethical standards to which, ideally, they also are holding themselves.
All authors contributed equally to the preparation of this manuscript. We are grateful for the comments of Edward Spencer, Brad Worrall, Margaret Mohrmann, Anil Shukla, and other students who have taken our courses, Elizabeth Karlin, and two anonymous reviewers in the preparation of this manuscript.