Routines

Routines are the primary substance of everyday organizational behavior (Cyert and March, 1963). Much of what happens in an organizational setting—from how work gets done to the decisions that are made—results from the ongoing execution of different routines. Routines are the habitual patterns of work in which individuals engage to accomplish activities and reach desired ends. They create efficiencies in organizational life (Cohen and Bacdayan, 1994). We recognize their value because routines also comprise the fabric of our own personal lives. Consider my normal workday: wake up, wake my son up, play toys or read a book with my son, shower, eat breakfast with my family, read the paper or smart phone, get into the car, leave my son at day care, grab coffee, go to work or meetings, work, eat lunch, work some more, go for a run or bike ride, pick my son up at day care, go home (or run a few errands first), eat dinner with the family, take a walk with the family, talk about the day with my spouse, maybe have a glass of wine, put my son to bed (sometimes after giving him a bath), sit with my wife and watch some TV, read before bed, and go to bed. With only slight variations on the overall theme, this is how most of my weekdays look.

What are we without our routines? How effective would we be in our daily lives, how dependable, if we did something new and different every day? Would we be successful people, parents, friends, or spouses? Could we do more with less because of the shortcuts achieved from creating more predictability in our day? How would others view us? Would we be seen as dependable and trusted? Would we have enough opportunity to correct the things we are doing wrong, if we recognize them as wrong? How personally satisfied would we end up without a predictable road map of what we should be doing and what ends to work toward?

Organizations are no different because they are made up of individuals. In this way, organizations use routines extensively to increase predictability that in turn enhances control. Routines promote speed and efficiency (Cohen and Bacdayan, 1994). Especially for managers and professionals entrusted to make the organization work, these are highly desired outcomes. The ubiquitous nature of routines in organizations also can enhance the quality of the products or services an organization produces. Consider a well-run production assembly line: key work processes standardized and encoded with the latest knowledge about how to do things right. The term assembly line has a negative connotation in the postmodern world. But think about how successful this means of organizing production has been in turning out high-quality products and services on a mass scale. A well-functioning assembly line is nothing more than individual routines stacked alongside each other, done every day the same way with an eye (everyone hopes) toward perfection. In daily life, it's the local Starbucks or McDonald's that shows us how effective routines are for meeting both organizational and customer ends. But in health care as well, routines are everywhere. Surgery, chronic disease management, the emergency room visit, immunization clinics, chemotherapy regimens, and the intensive care unit—almost everything we encounter in our health care system relies on routines to get things done.

The acknowledgment of routines as important organizational phenomena to appreciate and study has been around in organization theory for some time, particularly in discussions of how organizations learn and adapt (Cyert and March, 1963; March and Simon, 1958). This literature talks about how important routines are to enlarging the capacity of an organization to learn and perform (Argyris, 1999; March, 1991). In this way, routines are portrayed as important behavioral vehicles through which organizations may pursue change, creating safe havens in which leaders and workers can embrace failure and learning within the confines of how they “normally do things” (Feldman and Pentland, 2003; Hoff, Pohl, and Bartfield, 2006). However, this same literature also points out that routines function as potent barriers that enable the organization to resist learning, change, or adaptation (Weick and Roberts, 1993).

Thus, the dual nature of routines is revealed for us. They are the key to effective innovation, change, and learning in an organization because the organization and its workers already accept them as the right way to do things. They can provide the stealth under which new ways of doing things get incorporated, or failure is acknowledged and learned from. But routines also function as impenetrable walls against which change, innovation, and learning fall flat—the inertial forces acting on workers to convince them to keep things as they are. Routines create comfort within an organization, and this comfort often creates a risk-averse workforce. It is this paradoxical nature of routines that makes them so interesting, so powerful as organizational phenomena, and a big reason that health care researchers should place them squarely in their analytical sights.

Within the study of health care organizations and systems, routines merit a great deal of conceptual and empirical attention. First, they account for a great deal of what goes on in health care settings. Walk into any hospital or ambulatory care setting, and observe how things are done for a few hours. Almost all of what you will see are the same routines done over and over. Physicians, nurses, receptionists, and lab technicians are all trained to execute routines. They are trained to funnel their expert knowledge and discretion through different standardized ways of delivering health care services. They make decisions and exert their independence, but only to a point. When that point is arrived at, most often a preset routine takes over.

Patients are asked to participate in all of these routines, usually without question, with this lack of questioning becoming a routine unto itself. Hospitals, despite rhetoric to the contrary, largely look and do many things in the same manner as they did ten, twenty, even thirty years ago. Granted, the technology available is more advanced, and the clinical knowledge is better. But these things are simply raw materials that get loaded into generic routines that have existed for decades. One routine is that the doctor in a hospital makes all the key decisions. Another is that the nurse provides support to the doctor and does not question physicians' decisions. Another is that physicians visit patients at discrete points in the workday, usually early morning and early evening. And another is that family and friends rely on the patients themselves as the main source of information related to their conditions and the physician's assessment.

When we perceive an assembly line–like experience as we move through a health care service delivery experience, be it inpatient or outpatient, the reality is that the experience is defined by the execution of one routine after another. For example, a close relative of mine had quadruple bypass surgery performed on her at a world-renowned cardiac care hospital several years ago. What surprised family members was the manner in which this surgery was conducted. Several surgeons and cardiologists were employed throughout the surgery, each with a specific role and set of activities to perform. No single surgeon did everything. The surgery itself was organized in the same precise manner for my relative as for other patients there for the same thing. The postoperative care given to her was also standardized and prescribed to the letter in terms of staff, timing, activities, and evaluation. Wherever possible, the work was encoded into a routine. In this way, routines help organizations like hospitals serve a lot of patients simultaneously. They help physicians get better doing the complex work that patients require. They increase the level of certainty under which everyone working in health care operates, bolstering the collective level of practitioner confidence over medical work and allowing what can often be highly emotional, personal, and important activities get done with less anxiety, doubt, and delay. They also can improve the overall quality of a particular health care service, especially when there are multiple, complex steps in the work process—each of which can be done poorly if not practiced and perfected.

Health care researchers must think about and study routines more. Regarding theory, we need to develop additional ideas around the concept of routines that help us explain organizational behavior more precisely. For example, one idea to explore and cultivate should center on the life cycle of routines within organizations and among workers. How does a specific routine become accepted as the dominant logic for organizing within the work setting? What factors, both situational and routine specific, shape how long a given routine remains ascendant as the chief organizing or behavioral logic within a health care setting? How do new routines in health care organizations rise up and compete with existing ones? How are entrenched routines toppled to make way for new ones? What are the developmental stages through which a routine passes on its way from something new to something normal to something rejected? These types of questions are important. They inform much of what occurs in a health care setting in terms of the inertia that arises and prevents change, as well as how often and to what degree change is possible.

The theories to apply to questions about routines include those centered on work, people, institutionalization, and change. For example, psychological and social psychological theories that address what motivates people to change, step outside their existing roles, and become less risk averse in their behavior have relevance here. Key concepts from these types of theories relevant to studying routines include trust, identity, motivation, social networks, autonomy, control, and schemas. These types of concepts speak to the drivers and barriers related to routine creation, maintenance, and destruction. In addition, macrosociological theories that address how organizations institutionalize norms and expectations from their external environment can inform microlevel processes such as the development and embedding of routines in everyday work. Studying the life cycle of routines in organizations benefits from the application of ideas associated with theories of organizational change, including the diffusion of innovations, organizational and occupational culture, the competing values framework, and resource dependence theory. In the study of routines as artifacts, concepts that span the micro-macro ideological divide are enormously helpful since they account more readily for how workers and their environments interact to shape the genesis, content, and transformation of routine-based work.

Those mentioned are not the only questions or ideas to fixate on in the study of routines in health care organizations. The key is recognizing that what we see as patients, workers, or researchers within any health care setting is typical and normal—not exotic or strange—at least to those living it on a daily basis. What we see when we look at routines is a world where people want to normalize as much as possible—perform their activities and interact in as easy and predictable a way as possible. The Chicago school of sociology, notably scholars such as Everett Hughes and C. Wright Mills, saw in the regularly occurring activities of workers the keys to understanding human behavior, social interaction, and why particular aspects of society, such as organizations, look the way they do (Hughes, 1958; Mills, 1959). It is not often the most glamorous research to conduct. Nor does it necessarily produce fashionable insights that burn bright momentarily but fade away on recognition of their limited applicability in the real world. But the increased study of routines in health care would help to explain additional variance in important health care outcomes like efficiency, quality, and patient satisfaction simply because most of what is done to affect these types of outcomes is based in work organized by routines.

Emotions

Emotions are not studied much by health care organizational researchers despite the highly personal and subjective nature of health care delivery. It is odd that something that plays an important role in outcomes such as service utilization, quality, safety, cost control, and patient satisfaction remains undervalued by those of us interested in knowing more about the ways in which the US health care system works. Why do we choose to understate the role of emotions in health care organizations? Part of it is that as researchers, we are encouraged to think in linear, rational ways about the everyday world, as if the most important phenomena can be reduced deductively to simple cause and effect. This view often portrays health care workers and organizations as instrumentalists, motivated to perform in predictable ways by tangible incentives such as reimbursement, regulation, and competition and by internal factors such as training, experience, and control. Within this perspective, the emotional content of work and people is secondary and marginalized.

A second reason for the absence of a focus on emotions in health services research is that all of us who are engaged in the health care system want to believe that health care delivery operates on a largely rational plane of behavior. Certainly in our roles as patients, we have faith that doctors and nurses, for example, interact with us and do their work in calculated, logical ways, perhaps because this makes us feel more confident in part that the care we receive is competently done. For the same reasons, we do not wish to believe that our auto mechanic might do a lousy job fixing our car because of his own feelings about how his employer is treating him, or his prior negative experiences with demanding car owners, or even ongoing problems in his home life; we also do not want to believe that decisions for our care might be driven by emotionally derived thought or behavior.

This is especially true in health care when we require complex interventions such as surgery or have a life-threatening disease. In these situations, we see stoicism on the provider's part as a symbol of confidence and expertise. In these situations, we are the emotional parties. Providers are the ones who think and see clearly for us. We often come to expect that our physician will view us directly in relation to the essence of our problem—the type of body part and condition for which we have been brought to that physician in the first place. As a result, we may find no fault when the physician treats us dispassionately and focuses not on our holistic being but rather on that which is making us sick—the cancer, the blocked artery, the diabetes. And as Michel Foucault (1973) notes, such a reality may be beneficial in the sense of advancing diagnosis and treatment and of making the individual patient better. In this way, though, the lack of attention given to emotions in health care also becomes partly a function of our own socially constructed identities as patients.

We also do not focus on emotions in health care research as much as we should because there are not a lot of ideas or theories that link the emotional aspects of care delivery to various health care outcomes. We might know full well, for instance, why being nicer or friendlier to a patient decreases the chances that a physician gets sued or that expressing such positive emotions leads to more satisfied patients (Levinson et al., 1997). However, beyond these smaller insights, our understanding of the full impact of emotional expression on medical care and decision making is limited. This is due significantly to the general lack of tie-in of emotional dynamics within the management and organizational literatures to tangible production or service-based outcomes. Instead, the trend within these literatures is to demonstrate the impact of emotional expression in shaping social interaction and in how people think about their work, roles, and the larger organization (Rafaeli and Sutton, 1987; Wharton and Erickson, 1993). In short, the connections get made between emotions and intermediate outcomes, with the assumption made (and often not proven) that the latter end up affecting production, quality, efficiency, and cost.

The essence of health care services as highly personal and subjective means that emotions probably do matter a lot and so should be studied in greater detail in health care. We see so much failure and variability in health care that it makes sense to consider all possible culprits. For instance, we feel ill, so we go right to the physician. The physician is then supposed to work through a differential diagnosis, carefully ruling out certain illnesses while ultimately arriving, with the aid of testing and experience, at the correct diagnosis. Armed with the correct diagnosis, the doctor and other supporting structures such as hospitals then determine and provide the correct level of service to cure us and alleviate our symptoms. Insurance companies recognize the value of physician and hospital expertise in this regard and pay for these stakeholders to perform diagnosis and treatment. We get better. The system works.

Only it often does not work, and the rational process and its implicit assumptions described often do not hold true, in part because of the emotional nature of health care work. For example, we do know that physicians' fear of being sued drives a portion of their clinical decision making, which in turn drives the use of services, most often in an upward direction (Weisman et al., 1989). Thus, the same patient with a severe headache visiting two different physicians may get two different levels of service. The physician recently sued or deposed in a colleague's malpractice trial might end up ordering expensive magnetic resonance imaging to rule out a brain tumor or some other low-probability diagnosis. The physician with no history of being sued might instead send the patient home with a prescription for ibuprofen and instructions to rest for a few days. In this case, it is the emotional component that accounts for differences in how the physicians act. The first physician has let his or her own fear drive decision making, producing an additional test, with additional inconveniences, costs, and risks for the patient. Thus, rational behavior does not explain the variance observed in the diagnostic process. Rather, it is the subjective at work.

Physicians who fear their patients more because of having been sued may come to view them with greater ambivalence and less empathy for their conditions. This may affect how they choose to treat those conditions. Patients may pick up on physicians' body language in this regard and become more dissatisfied with their encounters and treatment regimens, even if the care provided is appropriate–or, quite the opposite, they may be happier because the physician is compliant to their wishes for another test or prescription or an additional visit to the specialist, but this increases use and decreases quality because unnecessary clinical care may be provided. Continuing the malpractice example further, research does support the notion that how doctors communicate with their patients often influences whether patients feel that malpractice has been committed and whether a malpractice claim gets filed (Levinson et al., 1997). These findings imply that patients who feel better about how doctors express themselves toward them are less likely to sue, even if actual mistakes have been made on them. Once again, emotions drive behavior.

There are many other examples in health care where emotion is shown to matter. Patients who fear or do not like going to doctors may ignore negative signals concerning their health, making their conditions harder to diagnose, treat, and cure once they enter the system. The effectiveness of teamwork in health care may depend on the level of emotional influences at work in a given work situation. For instance, tensions between doctors and nurses on a clinical team may influence their effectiveness across a range of work-related outcomes. Frustrated nurses who feel compelled to resist physician authority may end up engaging in shirking or underperformance of their own duties as a retaliatory action to gain some degree of control within the team. Physicians bent on establishing their roles as chiefs of the team may be driven by anger or resentment to act more authoritatively within the team, lessening the opportunities for group learning and dialogue. In fact, when one sees the equivocal findings in the health care literature regarding the effectiveness of teams (Bosch et al., 2009), it is likely that the emotional baggage endemic to health care teams and the professionals working within them plays a role in reducing their effectiveness. This baggage could range from how power relations are structured within the team and the feelings such relations evoke, to strong personal preferences among different team members concerning how their work and the work of others should be coordinated and valued.

Other examples of emotive effects in health care include surgeons who get bullied by nervous, aggressive, or worried patients into doing surgeries that are not imperative or could wait longer given the trajectory of the illness or healing process; patients who leave a physician's office dissatisfied because of how they have been treated by staff, even when the clinical services provided are both appropriate and effective; hospitals and physicians who cover up errors and safety issues because of a collective fear that someone will find out and make them look bad, hurting their reputations and businesses; and insurance companies forced to cover a specific service like infertility or a promising new cancer treatment drug not because they want to or because it makes good business sense but because of legislation enacted in response to public outcry about the evil nature of any insurer that does not want to pay for these services that could produce or save lives. In all of these cases, a heavy emotional component drives the behavior of stakeholders.

How might we begin to recognize and study the emotional component of the health care workplace more closely? First, we need to describe and classify the various ways in which different emotions come into play across different health care situations. We need to develop propositions and ideas that provide a unifying thread to how emotions work on health care behavior and outcomes, in the process creating an applied research agenda. As part of this initiative, taxonomies can be articulated of relevant emotional dynamics that help to shape health care production processes and outcomes. For example, in considering the outcome of service utilization, we should ask how in different situations the emotional component might produce overuse or underuse of care and which types of emotional dynamics are relevant. Do emotions such as frustration and fear work similarly on clinicians to influence their decision making? How does an emotion like fear shape the manner in which clinicians think about their work and patients? Do physicians and nurses experience and react to an emotion like fear in a similar manner? These are the types of research questions that push toward a comparative frame through which to see the general patterns of how emotions shape health care delivery.

A second direction to take in bringing theory and research on emotions into health care is to apply ideas from the management literature that relate to emotion management (Hochschild, 1979) and test their validity for health professionals and patients. The assumption that organizational effectiveness in service industries relies in part on the emotions that employees express and how those emotions are managed (Rafaeli and Sutton, 1987) makes emotions an important management focus. It motivates the organization, in both positive and negative ways, to monitor and control how its workers present themselves to customers. It also raises general questions of how emotions among workers are treated, whether it is the worker or organization that proactively attempts to manage those emotions, and the outcomes that result from strategies used to shape the emotional presentation of self in everyday work life (Goffman, 1959).

One idea worth exploring in depth is that emotion in the health care workplace is governed heavily by social rules—existing expectations, norms, and meaning systems that provide constraints and opportunities for workers to express themselves in particular ways (Goffman, 1959; Hochschild, 1979). In this way, the manner in which health care workers such as physicians express themselves in their jobs is subject to both external manipulation and social control. This sets up the potential for conflict and tension between the health care worker and the settings in which he or she works. Studying this conflict and how it is resolved or negotiated may inform issues such as patient satisfaction, service quality, professional autonomy, and health care quality. The malpractice fear example I already mentioned might be evidence of one type of external manipulation and how it acts on physician behavior. Another physician-related example is seen in how particular reimbursement structures or incentives shape how doctors present themselves to their patients during clinical encounters. Primary care doctors, for example, have been shown to be both more emotionally detached and paternalistic with their patients because they are forced to move through workdays with large numbers of face-to-face visits offering low reimbursement and limited time frames (Hoff, 2010).

Health care workers also take advantage of opportunities in their everyday environments to define how they wish to interact emotionally with patients. This agency focus is characterized by a heavier emphasis on self-reflection and choice regarding worker presentations of their emotional selves. It suggests a number of questions:

• How do professionals like physicians and nurses process patient emotions in constructing their roles as caregivers?
• Do health care providers form their own assessments about how patients wish to view them, and does this influence their thinking about how to express themselves as professionals?
• Do surgeons feel normative pressure, for example, to act dispassionately and present themselves as unemotional technicians because they think that is what patients expect of them and because they know at some level that how they present their role has implications for the levels of emotion felt by the patient for his or her own situation?
• Because of this ingrained pressure, do surgeons tend to resist more often signals from their patients that call for a different or more overt type of emotional presentation of their role?
• Does this resistance create forms of dissonance that make surgeons more frustrated or ambivalent in their work?

These types of research questions illuminate both psychological and social-psychological mechanisms of emotion management that focus on how individuals in the workplace willingly negotiate external perceptions of their roles.

Roles, Careers, and Identities

Roles, careers, and identities are also key sociological artifacts to study in health care. Much organization theory views these as heavily influenced by social processes occurring within organizational workplaces (Ben-David, 1958; Pratt et al., 2006; Rizzo, House, and Lirtzman, 1970). For health care professionals, the influence of social construction on roles, careers, and identities is particularly intense. For example, physicians and nurses are “experts,” trained and socialized to behave in uniform ways that convey confidence, certainty, and compassion. The social manufacturing of a physician's identity begins the first day in medical school, symbolized through a white coat ceremony where the traditional garb of practicing physicians is bestowed on the student, signaling the start of a several-years-long transformation into a competent clinician and the subversion of key aspects of his or her own persona that do not align with the clinical role (Becker et al., 1961). A primary goal of any type of professional training is to imprint the collective identity of that profession on the individual participant who is seeking to become a part of it. Medical students are socialized over four years through culture, experience, and a rigid structure of testing and evaluation to see themselves as doctors in a particular way.

But identities alone are the not sole artifact of interest. Rather, identities shape role definition and content; in turn, the enactment of roles influences the career trajectories of individuals in the workplace. Pratt et al. (2006) allude to this in their emphasis on how the specific work in which professionals engage changes the substance of their professional identities over time. In short, a worker's identity functions as the proverbial piece of clay that is molded through ongoing role performances. Over time, these performances coalesce into an observable career trajectory that interacts with identity to forge a dynamic understanding of the latter concept. This reality places significant empirical emphasis on phenomena such as identity construction, maintenance, and destruction, which serves as a window into seeing the links between the work individuals do, their career choices, and how they ultimately think about who they are (Pratt et al., 2006). In this way, the interconnection between identity, work roles, and careers is illuminated.

This interconnection helps researchers better understand why things look the way they do in health care delivery. Consider another example with physicians. Young physicians graduate from residency socialized to think and act as competent clinicians. Yet during residency, they are denied full freedom in the everyday manifestation of their professional identities as they work alongside older, more experienced clinicians and take the majority of their clues around enacting identity from these colleagues. In this way, the clinical identity gets encoded within them, whether as primary care physician, orthopedist, or cardiologist—depending on their specialty choice. But it is only once they find themselves burdened with a full visit schedule of their own patients or a full day of surgeries they must perform as the lead surgeon, without the referent of a superior on which to transfer some role accountability, that they have an opportunity to enact their identities fully—that is, the capacity to construct and deconstruct their identities further—to maintain or change them through their social interactions and work duties.

For instance, young primary care physicians encounter a role in which they must see twenty to thirty patients each day within their practice, spend fifteen to twenty minutes on average with each, work in their spare time as patient advocates with insurance companies and specialists, perform as team leaders guiding and overseeing practice staff, and engage in medical work that contains a fair amount of routine, low-level decision making (Hoff, 2010). Over time, working in such a busy, high-pressure setting, these young doctors may become less satisfied with the primary care work role and may seek to shift their clinical identities toward another clinical specialty such as emergency medicine or a new organizational work role such as manager or medical director.

Alternatively, they may allow the reality of their everyday work situations to shape their clinical identity in a way that opens up this identity to direct competition from other social identities such as that of spouse or parent. Career trajectories may then be affected by this identity-based competition as greater commitment to these other social identities develops and, for example, makes some primary care physicians want to work less, spend more time with family, and pursue nonwork interests (Hoff, 2010). Increasing ambivalence with the existing clinical identity may also occur, further weakening the centrality of that identity within the young doctor's mind-set. The important point is that all of these potential outcomes have implications for how these doctors approach primary care work and their patients and for the quality of their primary care service delivery. For example, primary care doctors who become less engrossed in their clinical careers or to a full-time professional identity may have less commitment to any given patient or may come to view different parts of their everyday work as necessary evils in which to engage but with less dedication or passion (Hoff, 2010). This in turn might have an impact on how patients perceive their clinical encounters and the overall quality of the primary care product put forth by a practice.

At the intersection of roles, careers, and identities sits meaningful questions related to workplace dynamics involving commitment, role tensions and conflict, worker agency and adaptation, the development and management of cognitive schemas, and organizational socialization processes. Exploring these issues in greater depth illuminates, as the example above does, the linkages between worker psychology, work content and context, and the manner in which health care organizations function. Given that health care is a service industry in which service providers are the key production inputs, this empirical focus makes perfect sense and should be pursued to a greater extent than it is at present. Examining more the interplay between worker psychology and specific features of the workplace illuminates the social structural aspects of well-worn concepts such as professionalism, occupational culture, and bureaucratic-professional conflict, which breathes new life into these concepts by opening up new empirical questions to pursue.

For instance, commitment represents the general feelings workers have toward the different identities engrained within them, with the idea being that the intensities and distribution of such feelings influence everyday loyalties and behavior in specific directions (Wallace, 1995). A long-standing but relevant area for commitment research is examining how professionals enact their feelings of commitment when working for competing interests. For example, social structural views emphasize that physicians who leave clinical roles and go into full-time management positions maintain allegiance to both their professional ideals and the interests of the organization (Hoff, 1999).

In theory, this dual commitment should make physician executives more effective managers than nonphysicians trained and socialized to place greater emphasis on the organization's goals in health care delivery. Being dually committed should produce organizational decision making in health care that gives equivalent weight to individual patient interests and the larger corporate interests of profit and efficiency. Thus, studying the ways in which health care workers, especially those in key decision-making positions, come to ally themselves with one or more sets of interests helps us to predict the kind of organizational cultures we should expect and the types of outcomes most sought after by the organization in doing its work. But first we must acknowledge, through the study of sociological artifacts such as identity, that phenomena like dual commitment are possible.

The interconnection of roles, careers, and identities also raises the conceptual significance of two contrasting phenomena: human agency and organizational socialization. Both have something to say about how these three artifacts interrelate for a given worker. Agency refers to the ability of individuals within organizations to act voluntarily on and shape their environments and to make conscious choices about how to approach and perform their work (Hodson, 1995). An agency-based view of health care workers is a logical outgrowth of emphasizing the relevance of a role- and identity-based perspective because it places empirical emphasis on the internal motives and external incentives that drive health care worker job choices and role performances.

In health care, an agency view is appropriate given the higher degrees of autonomy and discretion many groups of workers possess in doing their work. Studying the hows, whens, and whys of this work independence can help us better predict the types of variations and outcomes in health services delivery that rely in greater depth on worker discretion to produce the service in question. It also enables us to consider the opportunities and challenges associated with implementing health policies that depend substantially on the human capital component to make them work effectively, such as the patient-centered medical home model that is being counted on to transform US primary care.

Alternatively, roles, careers, and identities are also shaped through organizational socialization processes. Generally such processes function to teach workers how to perform in a given work role within the context of larger expectations established within the organization for how such a role should be fulfilled. The organization may be an institution with a fixed geographical location, such as a hospital, or it may be boundaryless, as in the case of a profession to which the worker belongs, such as medicine or nursing. Regardless of the contours of the given organizational milieu, the key point is that the organization in question communicates to workers acceptable ways of doing their work and enacting their role. This in turn puts pressure on these individuals to manifest their identities in specific ways consistent with the organizationally approved work role enactment. In this case, the human agency—the freedom to act—is secondary to the normative pressures of the institutions to which the worker belongs.

Studying how an organization socializes its workforce in ways that shape health care work and outcomes should be an important goal for health care researchers. In part, this is because the freedoms and autonomy afforded different groups of health care workers may not always produce consistent work role enactments across settings, leading to unwanted and excessive variation in how medicine is practiced, patients are served, quality is achieved, and costs are controlled. Understanding better how health care organizations influence workers' role performances while molding their identities in ways that are aligned better with desired outcomes could inform more effective policy and management strategies for improving health care systems. It would also place the study of organizational socialization processes squarely into an applied framework that heightens the importance of contextual phenomena such as workplace culture, teamwork, organizational learning, leadership, and work flow design.

As socially constructed artifacts, roles, careers, and identities should occupy a place of prominence when considering how health care organizations do and should perform. Each of the three reflects on a large array of social and psychological dynamics that characterize health care workers and how they behave. They enable us to get inside the heads of workers who possess a high degree of freedom in how they do their jobs, which allows us to study how this freedom shapes important health outcomes such as quality and patient satisfaction. They also allow us to incorporate two contrasting perspectives that the field of organization theory uses to explain worker behavior: the role of agency, on the one hand, and organizational socialization, on the other. Finally, these artifacts reveal the health care worker as a dynamic entity whose allegiances, role play, and career experiences are subject to change over time, exhibiting a chameleon-like tendency that provides value to the individual and, in many cases, the organization.

Work

We spend less time than we should in health care research studying what actually occurs in health care settings. If we spent more time, we would question to a greater extent a number of taken-for-granted assumptions related to how health care delivery works. Of course, there have been many informative studies of health care work in areas such as organizational learning, patient safety, culture, and quality of care (Edmondson, Bohmer, and Pisano, 2001; Hoff et al., 2006; Nembhard et al., 2012). However, this inductive research arguably represents more of an exception than the rule. It also is research that tends to focus disproportionately on the issue of culture within settings and organizations, which limits its ability to speak to other health care dynamics and outcomes that are not as closely associated with the shared meaning systems existing within the everyday workplace.

Most of the time, we study health care phenomena deductively, using prespecified hypotheses and, often, secondary data in ways that understate the social dynamics of health care work and how those dynamics ebb and flow over time. We also use economic theory extensively to test universal assumptions about how professionals such as physicians will think and act in relation to their organizations, incentives, patients, and other workers, regardless of the nature of the settings in which they carry out their work.

This is not a criticism of the field but a lament. As a sociologist interested in how people experience their work lives and how work is organized, I am drawn to focus on the nature of the work itself: what activities and duties constitute a job, the levels of complexity and autonomy within that job, how job duties and work change over time, and how the performer of that job thinks and feels about it. My experiences as both a health care administrator and a qualitative researcher have convinced me that interesting things reveal themselves when we open ourselves up to this conceptual frame. Of course, a focus on work and its meaning comprises an entire field of sociology (the sociology of work) and part of another (organization theory and behavior). The study of work across various industries has a long, cherished history. Every type of worker, from managers to elite professionals to factory workers, has been studied in relation to what they do, how they do it, and how they feel about it. In short, a wealth of ideas and prior empiricism exists to draw on from across a range of industries and workers when studying health care work.

In health care, we have studied the work of doctors (Freidson, 1970; Hoff, 2010) and nurses (Chaska, 2001; Gordon, 2005) extensively. We have also spent a good deal of time examining how these professionals relate to both their work and their employing organizations. But this research remains largely outside the mainstream of health services research. Much of it also focuses heavily on a single phenomenon such as professional autonomy that oversimplifies the plight of health care workers and risk, turning them into one-dimensional social creatures (Freidson, 1970). If work is artifactual and thus socially constructed, there are no easy oversimplifications that can be made. There can be no universal justifications for focusing on a single value like autonomy at the expense of the full range of values, beliefs, and norms that may typify individuals in a given workplace (Hoff, 1999).

Anyone who has ever worked can appreciate the complexity of a typical everyday work life. The work we do and our experiences doing it are variable. There are good days and bad days. There is work that at one moment seems interesting and the next boring. Different employing organizations and managers can make the same work tolerable or unbearable. Things happen in a job that make easy work suddenly hard, hard work easier, some work disappear, and new work appear. A surrounding context affects how work is conceived and performed—things like time and production pressures, staffing patterns, and types of customers—and all workers are affected by these contextual elements. In one context, the desire for greater control may drive our thoughts and actions. At other times, control may be secondary to security, comfort, compensation, or simply getting home on time. Work is done one way at one moment, and at a later moment, the same work may be done differently. All workers relate to their work and jobs in ways that incorporate their personal desires, experiences, internal values, and overall talents.

These peculiarities of work comprise a rich and diverse set of phenomena that illuminate larger organizational dynamics such as culture, structure, human resource management, organizational change, organizational learning, power and politics, and strategy. Studying what motivates workers to perform well in their jobs, for example, can reveal the dominant meaning systems within a particular organization, such as what is valued, the manner in which incentives are structured by its managers, and who is in charge, such as the physicians or the administrators. Examining how workers experience constraint or opportunity as a result of performing their jobs within a specific set of contextual circumstances informs our understanding of how organizations shape production through proactive strategies and serendipitous events that are less predictable. Through these examples, we see how the microworld of work can inform the macroworld of organizational functioning.

What does it mean to study health care work in greater depth as a sociological artifact? First, it involves looking closely at how individuals perform and relate to the tasks they are paid to do, because this is where much of the social construction that typifies work occurs. To do this, a researcher must take the time to watch and listen; one learns much about how surgeons approach and conduct the work of surgery by watching actual surgeries performed and speaking with the surgeons who did them. Watching surgeons work and listening to what they said about their performances revealed insights into how they thought about and behaved with respect to important outcomes like patient safety and health care quality. For example, it is instructive to witness firsthand how some surgeons downplay mistakes they or others make and to hear them talk about mistakes as a natural feature of their work. This indicates that they think about safety differently from the rest of us (Hoff et al., 2005). Others have found similar dynamics (Bosk, 1979). Watching and listening to these professionals conveyed the sense that a certain number of errors were acceptable within a field of work that for them possessed constant uncertainty, no matter how simple or complex the operation. They acted around safety in a more casual manner than we might expect, but in part it was due to their own perceived need to remain confident that they could perform all of their operations successfully while not being overwhelmed by negative events that might occur.

In short, it often appeared from close observation of these professionals that they accepted a trade-off in their work of occasional errors without dwelling on them for too long to prevent what they felt might be an even larger number of mistakes resulting from heightened anxiety and doubt associated with the constant self-reflection around their work that a greater emphasis on safety might provoke (Hoff et al., 2006). These types of findings, gleaned as a result of examining directly how individual surgeons think about and enact their work roles, tell us much about the potential for organizations that employ surgeons to improve quality and the strategies needed to do so. It reveals that there may be limits to how much we can improve patient safety in places such as hospitals simply because of how the key production workers within that setting believe they need to think and act.

A second way to carry out the study of work as a sociological artifact is to examine how a given work or production process is organized because that organization provides clues as to who is in control, who has power, what work is most valued, and the types of outcomes to expect from how the work is done. Some of this focus includes an examination of routines as described above. For instance, work in health care that is standardized through, for instance, a clinical care guideline at the expense of a process left primarily up to worker discretion assumes that everything important that needs to be known about producing that product or service is indeed known. That knowledge can be encoded within written procedures and a set routine. This sends a clear message about how the organization views the type of work that is standardized, and it conveys the notion that bureaucratic rather than professional control ultimately carries more value in achieving the desired outcome.

The way work is organized also makes clear the specific work-related processes and outcomes that are important and rewarded within the organization. For instance, performing work through a standardized clinical guideline encourages work flows to be structured in ways that emphasize efficiency and speed, since the activities themselves are known in advance and accepted by all involved in the work flow, making the work somewhat predictable. Thus, when one examines a work flow associated with the use of a standardized guideline, it often looks substantively different from one at the mercy of an individual physician's discretion where individual preferences, work style, competence, and demeanor become important inputs that shape care provision differently across patient encounters. Clinical work flows organized by standardized guidelines also will likely be associated with pay-for-performance incentive programs that reward workers on the basis of how well they implement the guideline, as opposed to how well they perform work that falls outside the guideline.

A third way to study work as an artifact is to study how, when, and why work changes. This focus provides a view into how events within the surrounding environment may dictate how work is organized and conducted. This grounds the study of work firmly in the study of organizations. For example, studying transformations in work as a result of innovations such as the introduction of new technologies, shifts in the organization's mission, new customer preferences, or the goal of achieving some outcome more effectively both informs and draws from topics such as organizational change and adaptation, institutional survival, strategic implementation, and organizational learning. Taking as an empirical focus the moments when specific types of work processes are pressured to modify in response to external stimuli provides an opportunity for health care researchers to bring in theories and concepts from a wider array of organizational literatures. It also maps more precisely onto events within a current US health care industry that is characterized by a high frequency of small disruptive changes due to technology and innovation enhancements and a potentially impending revolutionary change such as that typified by US health reform.

A Word about Context

In this emphasis on the health care workplace as a sociological artifact, I emphasize that context is to be included rather than controlled for in research. As we study the socially constructed elements of the health care workplace, a naturalistic approach is warranted that considers how artifacts are bound up with their surroundings. I have articulated this general point already, but it merits additional treatment. Giving context a central place in the study of health care organizations is challenging because context is the element many researchers want to control away the most. Much of our work controls for it through the application of quantitative analytical techniques. Organizational researchers often isolate the effects of one or a few independent variables on a dependent variable, examining the variance accounted for after the effects of other variables have been removed statistically. In this sense, many health care organizational researchers aspire to function like hard scientists—determining proven cause and effect within a controlled setting to help create general understandings of how our health system works.

But accounting for context is not a tidy process. Workplaces are bubbling cauldrons of contextual phenomena interacting with each other constantly in unpredictable ways. To separate out the influence of contextual variables, even for a moment, is to create a potentially artificial relationship between the remaining variables that may not hold up in the real world. For example, we may know that financial incentives work universally on some level in shaping physicians' clinical behavior. But what we often get wrong is estimating the precise magnitude of an incentive's impact in changing that behavior, identifying the full range of intended and unintended outcomes that come from using incentives to modify physician behavior, and how other organizational factors work on physicians simultaneously alongside financial incentives to minimize, distort, or enhance the latter's impact. In these instances, understanding how the surrounding work context influences and complicates the physician-incentive relationship is critical to understanding a wider range of actual situations in which the relationship plays out. Otherwise we risk gaining an incomplete picture.

In one sense, attempting to control for context within an organization in order to see the true variance associated with one variable acting on another is akin to going to the zoo and assuming that what one sees is a true picture of how lions, seals, or giraffes act in the wild. Watching a lion in a cage or artificial enclosure obscures the fact that in its natural setting, a host of factors, some predictable and others unpredictable, determine the lion's demeanor, eating and reproductive habits, longevity and survival, and general behavior. In a zoo, we may get a clear picture of how a lion acts, but it is action observed outside its normal setting and represents a partial glimpse of how it really thinks and behaves. It may be accurate insofar as how lions living in zoos act, but it arguably lacks authenticity.

Similarly, it is important to examine organizational phenomena like the workplace artifacts discussed in this chapter as they occur naturally within their everyday settings—precisely because they are created and sustained by social processes that derive from everyday organizational life and involve interaction, relationships, and belief systems. These social processes cannot be disentangled from the circumstances that give rise to them. This means that by statistically controlling for confounders in quantitative research, as contextual dynamics are traditionally labeled, we create an artificial organizational world that probably does not exist in everyday reality—a world where it is presumed that one variable acts independently on another, free from any influence by other variables also operating within that particular work setting.

Statistically, we employ techniques such as interaction analysis, which assesses the moderating effects of additional variables on the statistical relationship between two other variables, to bring more contextual features (i.e., confounders) into view when developing causal models. But interaction terms are crude proxies for context because they must be specified precisely and thus on their own capture extremely small segments of reality. This approach is less successful for studying multidimensional, hard-to-see phenomena such as routines, emotions, roles, and work. One problem is that it is the researcher who often determines what variable interactions to examine. We take responsibility for defining the context of relevance despite what is our limited (at best) knowledge of how any workplace actually works on a daily basis. This approach depends on our own ability to understand all the relevant social dynamics and whether we have captured the appropriate contextual features. Given the limits of human cognition and experience, this approach, even if aided by theory, is unrealistic.

A second problem is that most interaction analysis ends up wholly insufficient for understanding how many different organizational features might be working both together and separately on the phenomena being studied because such analysis often tends to include only a handful of variables that are presumed relevant. Yet we know that context is more than a handful of variables. It is the people in an organization and what they are doing and thinking. It is management and their strategies for getting things done. It is the workplace culture and what is valued and believed. It is how workers feel about and interact with each other. It is the design of the organization in terms of physical features, policies and procedures, and organizational charts. It is the use of power and politics to get things done. It is the makeup of the production process and the technologies employed. It is the world outside the organization that compels it to behave in certain ways, shaped by regulations, competitors, and customers. It is the attitudes of staff and leaders. It is the pace and type of change occurring at a given moment. It is the fiscal situation. It is the type of product or service, its life cycle stage, and its complexity to understand and produce. It is the level of resource availability.

All of these things are constantly in motion, affecting organizational behavior and outcomes in ways not easily modeled. Presuming that a given hospital setting, for example, does not have multiple structural, human capital, and cultural realities operating simultaneously on outcomes like safety or quality sets up our research for conclusions that have limited practical value and may explain only small amounts of unexplained variance. Indeed, part of the reason that the field of health services research has been struggling over the past two decades to help us understand how to improve patient safety and reduce errors within hospital settings (Hoff et al., 2004) may be found in our own willingness to oversimplify hospital work and organizational contexts to the point where our hypotheses no longer mirror the everyday reality of the contexts themselves.

Health care researchers should study context in a more organic way if they wish to view health workplaces as sociological artifacts. Organic means gaining rich description over conducting hypothesis testing and increasing the level of qualitative inquiry to illuminate the holistic picture of a typical health care organization. There is no other choice. There has to be greater appreciation for how health care settings operate in all their contextual glory, because it is the context that breathes life into the sociological artifacts we observe. Accepting this assertion inevitably directs the researcher to specific research designs or methods that, when used, are more effective at revealing the full range of contextual influences that are occurring in the workplace. Idiographic analysis involving heavy doses of primary data collection must become a primary goal.

Conclusion

Studying health care organizations through an artifactual approach is not easy. It is time-consuming and messy. Ideally it will involve a significant amount of primary data collection. To understand social construction fully, a researcher must see that construction in action or gain firsthand understanding from those who are involved in it. Observations, interviews, focus groups, and archival analysis of organizational documents such as meeting minutes—these are the empirical grist by which we gain understanding of which artifacts matter, and how. Primary data collection is not emphasized as much as it could be in health services research (Hoff, 2011), although its value and use have been more recognized by the field. However, governments, think tanks, and foundations still favor the compilation of large data sets filled with hundreds of variables measured in precise, cross-sectional ways. This creates an incentive for organizational researchers not to collect their own data or question what else might be missing for which we require additional information.

The overavailability of secondary data in health services research shapes both the questions we ask and out analytical focus. The questions we ask with secondary data can move us away from gaining insight into the social organization of the health care workplace and how it shapes different health care outcomes, in part because such data do not drill down deeply into the specifics of workplace action. As a result, we end up knowing a little about a lot rather than a lot about a little.

But there are efficiencies gained from gathering primary data in health care workplaces that make an artifactual approach feasible. For example, some researchers (including me) establish, through ongoing primary data collection, trusting relationships with organizations that last for many years and provide a reliable, ongoing opportunity to collect firsthand qualitative data on a variety of different workplace dynamics. Establishing an initial research relationship with one or two hospitals or large ambulatory care practices might be time-consuming, but continuing that relationship through additional research studies could be a relatively simple task. In this sense, mutual trust and positive experiences, as well as the researcher yielding some value for the organization, often by showing something that is relevant to those who work there, produce economies of scale in gathering artifactual data over time.

A second efficiency arising from ongoing primary data collection in health care workplaces is the enhanced ability as a researcher to identify the relevant social dynamics that help to shape how organizations function. Researchers get better at knowing what to look at artifactually in a given instance the more artifactual research they do. Within this chapter, I have identified only a handful of artifacts that seem to merit more empiricism and theory development in health services research. But they are also artifacts whose importance I have become better at understanding and seeing as I have done my own health care research over the years. The more one does artifact-focused empirical work, the more one sees the importance of a full range of social and psychological dynamics such as culture, conflict, trust, power, politics, schemas, and sense making. After all, other than theory and previous research telling us it is so, how else are we to appreciate fully the manner in which an intangible factor like trust matters between physicians and patients, and how it ebbs and flows through ongoing interaction and situations, unless we also watch and hear firsthand about this interpersonal relationship as it is enacted in everyday life?

Taking an artifactual approach to studying health care phenomena can be frustrating, especially if the researcher exhibits too much dependence on established theory. An overabundance of theory and ideas applied a priori or deductively tends to constrict what Mills (1959) aptly called “the sociological imagination.” Even when the research and data gathering are inductive rather than deductive, our training tells us that we must anchor our empiricism in well-articulated understandings of how organizations should, do, or could work—never straying far from what we already think we know or from existing research.

But too much emphasis on theory or existing research stifles the researcher's ability to engage in creative, original thinking. New understandings in organizational research often come about serendipitously and are not always directly descended from an existing concept or finding (Eisenhardt and Graebner, 2007). Since studying directly the social artifacts of health care often requires researchers to wait for the setting to reveal itself to them, rather than the other way around, the probability for such serendipity increases greatly. Thus, our methods matter for understanding health care organizations as artifactual in nature. Some methods reveal artifacts and their connection to important organizational processes and outcomes, while other methods do not.

This chapter has sought to make a case for greater study of the health care workplace as a collection of sociological artifacts. The artifactual approach has value because health care is a service- and people-focused industry. As such, social dynamics shape much of the thought and action, both collectively and at an individual level. Studying health workplace artifacts more closely and frequently is a necessity given the need to know more about how health care organizations work and outcomes are achieved in the areas of service use, quality of care, and efficiency. Until we acknowledge the full importance of the social structure of health care organizations and how this social structure influences everything, we remain somewhat blinded to how best to improve the system as a whole. These are blind spots we cannot afford, given the enormity of the challenges in health care delivery.

Key Terms

1. Agency
2. Careers
3. Context
4. Organizational socialization
5. Social construction
6. Sociological artifacts