Enriching Patient-Centered Care in Serious Illness: A Focus on Patients' Experiences of Agency

Abstract

Context: Following models of patient-centered care, including respecting patients' preferences and enabling patients (and their families) to participate in directing their own care, is especially difficult when providing care to cancer patients undergoing complex, aggressive therapy, when their capabilities are diminished. In this context, patient-centered care can be enriched when health care providers develop a deeper awareness of and appreciation for the multiple ways in which patients experience agency—defined as the capacity to make a difference—as both the initiators and the recipients of others' agency.

Methods: Qualitative data were collected between 2006 and 2008 in sixty-nine narrative interviews with both a panel of ten patients undergoing aggressive inpatient cancer therapy (autologous stem-cell therapy) and their lay caregivers. Data were analyzed using an interpretative approach and guided by a conceptual framework derived from recent sociological literature on agency.

Findings: We examined agency experiences through three distinct lenses: (1) the nature of the actors (both human and nonhuman), (2) the nature of the actions, and (3) the nature of the domains for action. The actors were individuals and teams, drugs and technologies, and the cancer itself. Actions could be initiated or received by the patients and included fighting and violating as well as comforting and healing. Domains were the patients' bodies, the health care system, and the patients' everyday worlds. In several instances, the same phenomena can be examined through more than one lens of agency, drawing attention to different aspects of the patients' experiences. Taken together, the lenses form a comprehensive picture of patients' experiences of agency.

Conclusions: This three-lens perspective can be a valuable template for health care professionals and policymakers to use in better understanding the experiences of patients in constrained circumstances. We propose that its use is a fruitful and promising contribution to recent conceptualizations of patient-centered care that emphasize the relational and facilitative aspects of care provision.

The Institute of Medicine's report Crossing the Quality Chasm: A New Health System for the 21st Century (IOM 2001) cited patient-centered care as one of six goals for improving health care quality, and since then, the concept has become increasingly prominent in health care delivery and policy. Each of the various definitions of patient-centered care builds on the fundamental tenet of respect for patients' values, preferences, and expressed needs, as initially proposed by Gerteis and colleagues (1993). The Institute of Medicine's articulation of this definition represents a focus on “the patient [as] the source of control” (IOM 2001, 304) in decisions about his or her care. That is, “patients should be given the necessary information and opportunity to exercise the degree of control they choose over health decisions that affect them. The system should be able to accommodate differences in patient preferences and encourage shared decision making” (IOM 2001, 3–4). These formulations of patient-centered care reflect the reverence for the patient's decision-making autonomy consistent with Beauchamp and Childress's definition of personal autonomy as “self-rule that is free from both controlling interferences by others and from certain limitations such as an inadequate understanding that prevents meaningful choice” (2009, 99).

Recently, scholars have widened the scope of patient-centered care beyond respect for the patient's independent or autonomous decision making. For example, Epstein and colleagues (Epstein et al. 2010; Epstein and Street 2007, 2011) argue for facilitating patients' involvement in their own care through enhanced relationships, including the thoughtful exchange of information and deliberation with clinicians, their patients, and the patients' families.

The growing amount of literature on implementing a patient-centered approach reflects this expanded conceptualization, which has three essential ingredients for patient-centered care: “an informed and involved patient and family, receptive and responsive health professionals …, and a well-coordinated health care environment that supports the efforts of patients, families, and their clinicians” (Epstein et al. 2010, 1492). For example, the Picker Institute (2004, 2010) provides extensive educational materials based on the premise that patients and their families should be enabled through knowledge, support, and confidence to direct their own care. Epstein and Street (2007) offer a framework for cancer communication that demonstrates how to give patients opportunities to participate in their care, highlighting the importance of an enhanced physician-patient relationship. Entwistle and colleagues (2008) use an example of communication about screening that illustrates how different types of interactions can be more, or less, compatible with enabling patients to participate in their own care. They further call for greater attention to patients' capability for autonomy as shaped by the social relationships and interactions that can enhance, or diminish, a patient's capability to “[be] one's own person and [shape] one's own life” (Entwistle et al. 2010, 742).

The aim of our article is to contribute to this line of thought by introducing the perspective of agency, defined by Giddens (1984, 4) as the capacity to “make a difference” through a flow of conduct in which the person/actor could have acted differently. We argue that insights into the provision of patient-centered care can be deepened by considering patients' experiences of agency, not only as the persons who exercise agency (e.g., by participating in decisions about their care), but also as the recipients of others' agency (e.g., by being affected by others' actions during the course of treatment), and we offer a framework for doing so. In this sense, a focus on agency has the potential to incorporate both the capability and the relational concepts presented by others.

To situate our article conceptually, we begin with a brief review of recent sociological contributions to the theory of agency, drawing from those that emphasize human agency and those that introduce the possibility of nonhuman agency. Next we present an analysis of narrative interviews with patients and their lay caregivers (typically a close family member), which we examine through different lenses of agency. We conclude with observations about the importance of considering patients' experiences of agency as a valuable complement to patient-centered care models, and we encourage this approach, particularly during aggressive treatment for seriously ill patients.

Theoretical Background

For our purposes, Giddens's definition is straightforward and useful: Agency is the capacity to “make a difference” through a flow of conduct in which the person (or actor) could have acted differently (1984, 4). This is also referred to as transformative capacity. In addition, Bandura (2006) notes that self-efficacy, the subjective belief in one's capability to act, is a key mechanism of agency.

Erimbayer and Mische's definition adds that agency can vary situationally and temporally: “[Agency is] the temporally constructed engagement by actors of different structural environments—the temporal-relational contexts of action—which, through the interplay of habit, imagination, and judgment, both reproduces and transforms those structures in interactive response to the problems posed by changing historical situations” (Erimbayer and Mische 1998, 970).

In our context, the changing historical situations pertain to an individual's experience with illness and encounters with the health care system. This includes the patient's participation in decisions about treatment and in the subsequent experience of the treatment itself.

Actor-network theory (ANT) maintains that the social world is made up of heterogeneous networks of both human and nonhuman actors (Latour 2005; Law 1992). While nonhuman actors such as machines and technologies may not have intention in the way that humans do, they do have the capacity to make a difference in a continuous flow of conduct. In this sense, machines and technologies serve as tools that assist or serve as proxies for human actions, and humans often attribute agency to these actors (Pouloudi and Whitley 2000; Rose and Jones 2005).

An additional insight from ANT, which is relevant to our study, is the observation that actors frequently act together in unified networks, almost as one. These networks reflect complex interactions among actors working in coordination to accomplish a task.

To simplify the complexity of such networks, people tend to adopt a strategy of “punctualization” (Law 1992, 385), a process of packing networks of human and machine agents into a single composite entity of action, sometimes referred to as “they.” In many cases, “they” are considered more powerful than the individual elements making up the network. Nevertheless, elements in the network sometimes function at cross-purposes, interfering with accomplishing tasks, despite being seen as a unified actor. This possibility makes it especially challenging to identify the source of difficulties in interactions with networks or teams.

We draw on these streams of theory to form a richer appreciation of individuals' encounters with the health care system during intensive treatment for serious illness. We use data from a qualitative study to illustrate the complexities of agency as they are experienced by the patients themselves and, occasionally, by their lay caregivers.

Research Setting and Methods

The study was designed to investigate the experiences of patients undergoing autologous stem-cell transplantation (ASCT) to treat relapsed or refractory lymphoma. ASCT is an extremely challenging treatment for patients and is recommended and undertaken only after previous cycles of chemotherapy, sometimes with radiotherapy, either have been unsuccessful or have been followed by a relapse of the lymphoma. The possibility of ASCT is raised by clinicians in discussions with patients and their families early in the course of the fight against the disease. Patients undergoing ASCT face prolonged hospitalization, often in isolation, as well as recurrent and invasive medical procedures and a range of severe and toxic side effects, which themselves can be life threatening.

Our study's participants were ten patients who were being treated with ASCT, plus the lay caregivers of nine of them. We recruited them in the bone-marrow transplant unit of a large teaching hospital in Sydney, Australia. The transplant unit is part of the hospital's hematology and cancer services department. The unit has a clinical director, a laboratory and scientific director, seven transplant doctors, two specialty transplant nurses, thirty nursing staff, one part-time clinical psychologist, one part-time psychooncologist, and one full-time bone-marrow transplant social worker.

We used a convenience sample in which all eligible, newly admitted inpatients were invited to participate until ten were recruited; no patients who were approached refused to participate in the study. Consenting patients were invited to nominate a lay caregiver (e.g., a spouse, parent, or other close family member or friend) to participate in the study; one nominated caregiver declined to participate because his English was poor. The participants ranged in age from their late twenties to their sixties; their employment situations included telemarketer, homemaker, allied medical professional, businessman, company director, and retiree.

The patients and caregivers consented to be interviewed first after the patients had agreed to the therapy but immediately before undergoing ASCT, three months after completing the ASCT, and then every six months, for up to six interviews. If a patient in the study died, the caregiver remained eligible to continue if he or she so chose to do. In all, we completed sixty-nine semistructured interviews over the course of two years, with the data collection culminating in 2008. (Additional details are available from the authors.) The data constitute part of a larger study with a longitudinal aim. Although this article is not intended to provide comparisons over time, the study design does allow for reflection from the patients and their caregivers on their experiences during the acute phase of treatment and its aftermath, providing useful insights into our analysis.

We interviewed the patients and caregivers separately and at a private location of their preference, which was typically in the hospital or the participant's home. The interviews were conducted by members of the research team, experienced in collecting data of this nature. The first interview commenced with the prompt “Tell me the story of your illness from when it first started and bringing us up to now.” Later, the interviewers asked the participants to update what had happened in the months since the last interview. The interviews ranged in length from about twenty minutes to several hours. The interviews were tape-recorded, with the interviewees' permission, and were transcribed immediately after each interview by a professional research transcriptionist. All the participants' names here are pseudonyms.

Analysis Procedures

The data were initially coded for described processes (Charmaz 2006), keeping as close as possible to the participants' own words and using gerunds for the codes (e.g., “fighting the cancer” and “doing all we can”). We employed an interpretive approach, working between the qualitative data and the relevant conceptual literature on agency to create appropriate codes, themes, and categories. We developed tentative understandings of the categories and repeatedly returned to the data to confirm or rework the categories. The authorship team jointly and cumulatively reached the final analysis and its interpretation reported here.

Findings

Following Bolman and Deal's (2008) notion of reframing, we present our findings as distinct perspectives or lenses through which we examined agency as it was expressed and experienced by the study participants. As Bolman and Deal suggest, reframing begins with analyzing each lens separately and then combining insights from each lens into a comprehensive picture of the concept under investigation. For our study, we identified three lenses: The first lens focuses on the nature of the actors; the second lens focuses on the nature of the actions; and the third lens focuses on the nature of the domain in which the actions take place. As we show, the same action sometimes can be viewed through more than one lens, enabling a richer understanding of the patients' experiences of agency.

The Nature of the Actors

The first lens focuses on the actors that emerged in the narratives as the most salient to the patients and their caregivers. Both human and nonhuman actors were characterized as having independent agency, capable of “making a difference.” For example, the patients, the clinicians, the lay caregivers, and other family members (individually or as members of a team) were all described as being engaged in the fight. In addition, other types of actors—the patients' bodies, the cancer cells, the stem cells, and the chemotherapy—were described as engaged at various levels in the fight against the disease.

Not unexpectedly, the lymphoma itself was regarded as a dominant actor, as the following quotation from a caregiver shows:

Louise: The lymphoma was just too aggressive. The transplant, stem cells, worked beautifully, but they just weren't strong enough yet … the lymphoma was just too quick growing … for the stem cells to sort of mature enough to work.

Some narrators spoke about the different elements of the ASCT treatment process (though not always technically accurately) as actors at a microlevel, while others focused on elements of the health care system as actors at a more macrolevel. The following quotation from a patient is typical of the microlevel:

Abby: At the moment, my cells in my body, I guess, are friendly cells, and they're accepting the cancer … they're not fighting it. What they have done by taking the stem cells out is froze them; … well, stem cells make cells, like white and red cells … and they're gonna take every single bad cell out of me and then re-put my stem cells back in, and hopefully they'll [the stem cells] produce angry cells … to fight the cancer.

In contrast, the next quotation describes the caregiver's expectation of agency from the health care system, including networks of human actors (“the oncology”) and nonhuman actors (“machine and those kinds of things”):

Francis: The oncologists announce that they can't do anything about it, she will die. … I thought [this] is a big hospital, is supposed to have a more, like, machine and those kind of things, you know, and do something about it. Basically, it's like hopeless … they don't have anything they can do, they can't do anything to save her. … I told him, I looked at him, and I know there's another machine there.

Both these quotations exhibit what actor-network theorists call punctualization, in which health care delivery is treated as a single unified actor. For Francis, the network—the medical professionals, the hospital, and the technology—was one entity whose agency (i.e., the capability to make a difference in the battle against her disease) he sought to act on behalf of his wife. For Abby, the health professionals (“they”) and the chemotherapy were a single actor engaged in the fight against her cancer cells.

These examples of punctualization demonstrate the challenges that patients and their caregivers confront when trying to make sense of a frightening disease, a highly complex treatment regime, and a daunting health care delivery system. Resorting to the nonspecific “they” to describe the various actors that exert agency on behalf of—or in threat to—the patient helps simplify an otherwise untenable situation that involves forces of substantial power. Nevertheless, as pointed out earlier, “they” may not always work as a coordinated team or a singular entity, which may lead to additional distress for patients.

The Nature of the Actions

The second lens shifts the perspective from salient actors to actions by focusing on how patients experienced the actions surrounding their care. To a great extent, these actions placed patients in the position of being the recipients of others' agency. Their acceptance of “being acted on” depended in part on whether they had authorized the action, in keeping with the tenets of patient-centered care that encourage patients' participation in directing their care. The acceptability also depended in part on the patients' belief that others' actions were necessary for their return to health. In the context of aggressive treatment for serious illness, discerning these two conditions requires a high degree of awareness and sensitivity from the health professionals.

For example, the following quotation illustrates a situation in which the patient initially agreed to a procedure that caused her substantial discomfort, and she became angry to discover that it was not an essential part of her treatment regime:

Norah: Before I got it [the nasogastric tube] removed, I was like, “This is very painful and I can't stand this” … and they said, “Do you want it removed?” and I said, “Are you telling me I don't need this?” and they said, “Oh, no not really, it is just a precaution that we put it in,” and I was going, “You know, you didn't tell me that, you made it sound like I had to have it in”; … I was furious.

Although some treatment procedures, such as having a nasogastric tube, presented opportunities for patients to direct their own care, in many instances opportunities for participating in directing their care were greatly constrained. Their illness was severe; the treatment options were limited; and once the treatment had commenced, discontinuing it could be fatal. The next quotation reveals the patient's recognition that the disease and treatment had reduced her to being the object of actions performed by others, with few opportunities to actively participate in the process:

Ingrid: Once you get a cancer like I've got, your body's not your own anymore, you've got to just give over, and I've had to do that … you are pushed and needles and everything, your body's not yours … you're on the tables, you're on the X-rays, you're this, you're that … what went on there this morning wasn't even easy, thing put in there and “put your arms back up” and then they inject this dye into you and you burn up all over, and it feels horrible.

Being acted on—becoming an object of another's agency—was not always perceived as a violation, however, and could be experienced as facilitating the restoration of one's own agency:

Colin: I was laying in bed; I couldn't move; [my son] would come three days a week, and he would bathe me. Literally pick me out of bed, put me in the bath, run the bath water, and give me a good wash. Take me out, massage my body, put me back in bed. … And I just started bouncing back, and that is why I am on my legs. I can walk around and so do things.

Colin appreciated being the object of his son's agency because this might restore his own capacity for action. Norah, in contrast, was angry that the health care team placed her (albeit with her agreement) in the recipient role without any apparent justification for doing so. Ingrid was upset about being the object of others' actions, partly because of the nature of the actions (which were painful or uncomfortable) but perhaps more so because she could not realistically refuse them.

Most revealing is the yearning by patients and caregivers for the network of health care professionals to be able to undertake actions—and to make a difference—in their fight against cancer, as well as for the belief that they could succeed in the fight.

Delia: To me, it felt like the doctors had given up, you know … it wasn't even long after the stem cell that they said, “Sorry, we can't do anything more for you” … when we left the hospital, we felt like the doctors had just given up on us, “go home and die” … so that upset me quite a bit.

This quotation and the earlier one from Francis, another caregiver, demonstrate the sense of disillusionment and despair when the health professionals offered no further actions in the fight against cancer. This shows that Bandura's concept of self-efficacy is central to agency, especially when the capacity to make a difference was the paramount purpose of the ASCT treatment.

The Nature of the Domain

The third lens draws attention to the structural environments in which patients experienced their own and others' agency, recognizing that agency operated differently in different domains. Three domains emerged in the narratives as especially meaningful: the patient's body, the health care system, and the everyday world.

The Body

As already shown in the first lens, patients spoke of their bodies as microworlds in which stem cells, technologies, and drugs were engaged in a battle to fight the lymphoma:

Louise: [The lymphoma] was, you know, an aggressive thing, and [the doctor] said, “Even though you have [had] the stems [and] the chemo, the lymphoma can be floating in the chemo, but the core inside is protected, it's got little pumps, pumping away the chemo from it, and it's keeping that little core inside protected … so you've got this lymphoma constantly pumping away getting rid of the chemo.”

Patients envisioned this domain as a distinct environment, which they hosted but in which they had little capability to actively or willingly participate. Thus, even though they were closely attentive as onlookers to this microworld, patients regarded the health care system and their everyday world as the domains in which they had more realistic opportunities and the capability to actively participate in their health care.

The Health Care System

Networks of human and nonhuman actors were the most evident in the domain of the health care system. Medical professionals, technologies, and hospitals were commonly described as unified networks, spoken of as “they,” and patients viewed these networks as having a substantial capacity to act. Nevertheless, even within the constraints of serious illness, patients spoke vividly of efforts not to “give up”—not to relinquish their capacity to make a difference. Some patients also encouraged other patients not to give up but instead to be more actively involved in their care:

Tony: But here was a lady [a fellow patient] … my only advice to her was that … she really didn't understand what was going on, and I said to her, “Well, start a diary, y'know, find out what you are getting, what they're doing to you, y'know, ask them what the tablets are for,” because … they were changing her tablets, and she didn't know what it was all about. … I said, “Ask em, annoy the shit out of them.” … I said, “It is your body.” So I gave her an exercise book I had, I just gave her a pencil, and then she went and got her son to bring her in a diary, and her outlook changed so dramatically … she took an interest in what was happening.

Thus we can see how patients found creative ways to increase their participation in their own care (providing a book and pencil, encouraging asking and recording), even while experiencing the limitations caused by the illness and treatment in the health care system and the hospital environment.

The Everyday World

In contrast to the first two domains, in which the patients' ability to sustain a capacity to act and to make a difference was compromised because of the disease and aggressive treatment, the parallel everyday world represented a centrally important domain because it allowed patients and caregivers to remove themselves, symbolically if not always physically, from the immediacy of the treatment. In this domain, patients and their caregivers frequently spoke of “doing everything we can.” This set of actions, beginning in the hospital and intensifying after discharge, included asking questions, doing research, consulting alternative medical practitioners, engaging in strict diet and exercise regimes, and vigilantly watching for signs of relapse. It was also in this domain that caregivers spoke of having their greatest potential for action on behalf of their loved one. In the following quotation, for example, a parent describes her efforts on behalf of her daughter following ASCT treatment:

Bronnie: I keep saying, well, you've got to remember that you're at a higher risk for everything than everybody else … and I said you can't forget that, which I think she does sometimes, which is good too, you don't want to dwell on it, you don't want to be thinking about it all the time, that's my job [laughs] … she should be out having a good time and, you know, enjoying herself with the most of her life and all of that.

Discussion

Our goal in this article is to offer, using a conceptual framework of agency, deeper insights into patients' experiences during aggressive cancer therapy and to suggest how these insights can be applied to models of patient-centered care. Guided by recent theoretical directions in agency, briefly defined as the “capacity to make a difference,” we identified three lenses that constitute a novel framework through which to understand patients' experiences of agency, as both actors themselves and the recipients of others' actions. Although we discuss the three as distinct lenses, in fact each considers the same actors and actions from a fresh perspective through a process of reframing, which, in turn, can enrich the interpretive potential of this approach.

The first lens focuses on the nature of the actors, recognizing that actors can engage individually and as nondistinct members of a network, operating and perceived as a single entity, referred to as “they.” The data show that the narrators attributed agency to both human and nonhuman actors from the micro- to the macrolevel and that the patients and their caregivers were especially focused on the highly agentic lymphoma.

The second lens focuses on the nature of the actions. The patients' narrations highlighted the common experience of being the recipients of others' actions. This lens, in particular, revealed the difficulties that patients face during a period of intensive and risky therapy that places substantial limits on their own capacity to participate and act. Patients must rely on others, including health care teams and lay caregivers, in order to manage their treatment.

The third lens focuses on the nature of the domains for action, representing different environments even within the same illness. This lens allowed us to analyze the narrators' partitioning of different levels of concurrent action. Thus, patients described the fight going on within their bodies against the lymphoma while simultaneously experiencing interactions in the social settings of the hospital and their everyday world. The patients' opportunities to actively participate in their care varied substantially, depending on the domain, and when they felt constrained in one domain, they sought opportunities in another.

Implications for a Patient-Centered Care Approach to Health Care Delivery

Implementation of a patient-centered care approach needs to take into account not just respect for patient's preferences and facilitative efforts to enable patients and their families to have greater involvement in their care. This approach also must appreciate the patients' experiences of agency, both their own capacity to act and the actions they receive from others, especially for patients undergoing aggressive treatment for serious illness. This goal, in particular, underscores the patient-centered ingredient of “receptive and responsive health professionals” called for by Epstein and colleagues (2010, 1492).

While there has been substantial attention to providing patient-centered care to patients in the context of primary care, acute illness, and chronic disease, a focus on this particular context is rare. Even in the detailed typology of population segmentation by health status (Lynn et al. 2007), which includes suggestions for patient-centered care for each of eight segments, there is no obvious category that would be relevant to the patients in our study.

Several themes emerged from our analysis that could enhance patient-centered care in this context. First, although opportunities to be involved and to be able to act really mattered to them, the patients' capacity to act could vary substantially. Indeed, the narrators recognized that the cancer itself was an aggressive actor they could not surmount without the agency of others. The patients understood that they could neither battle the cancer nor undertake the therapy on their own. This could be seen as a condition of multiple vulnerabilities, first to the cancer and then to the therapy and to those administering it. We suggest that health professionals who maintain a heightened awareness of this situation for their patients will better appreciate their own role in their patients' experiences of agency. They may also be better able to recognize that being caught in a condition of multiple vulnerabilities can be a source of significant distress that patients may not be able to articulate. Such awareness should help guide members of the health care team in their discussions with the patients and families, not just before treatment has begun, but also during and after the treatment.

Second, as noted earlier, the process of the aggressive treatment itself places patients in the unenviable role of being the object of others' agency, with little guidance about how to maneuver in this role. In a typical patient-centered care model, patients and their families are expected and encouraged to engage with these other actors more interactively, by asking questions, expressing preferences about their treatment, and actively participating in their care during the treatment. But this context may inhibit opportunities for patients to participate meaningfully in the direction of their own care, especially once the treatment has begun. Highly specialized medicine such as ASCT is commonly delivered within a tightly structured, institutionalized framework, drawing on protocols designed for maximum treatment effectiveness and patient survival. This context thus places additional demands on the health professionals to discern and accommodate differences in how patients express their agency.

For example, some patients may wish to assert strong preferences and seek every opportunity to exercise their own agency. If those preferences are at variance with the rigorous treatment protocols, the health professionals will be called on to provide compelling explanations to justify the recommended protocol or to be prepared to modify the protocol to accommodate the patients' wishes. Health professionals must distinguish for themselves between essential protocols, deviations from which could place the patients at high risk, and protocols in which acceptable variations could reduce the patients' experience of being the object of others' agency. Norah's experience with the nasogastric tube is an example of the latter.

In contrast, some patients may be unable or unwilling to express their preferences or to engage in actions supporting their care, especially during the debilitating stages of treatment. The risk here is that the patients and their family may be unrealistically expected to make care decisions about the course of treatment when, because of the illness, they have lost their objective capacity to act, along with a subjective belief in their capacity. Health professionals need to recognize when they must assume a greater facilitative role for certain patients than might be necessary for others and to encourage that role from families as well. One way to do this is by recognizing that agency is experienced differently over time and in different domains. For example, opportunities for patients to engage as agents in the everyday world appear to compensate somewhat for the loss of agency in other domains, as did opportunities for patients' families to serve as facilitators, such as through information gathering.

Even in the hospital, the health care team could encourage patients' involvement, such as that described by Tony in his effort to be engaged and to help his fellow patients remain vigilant as well. The health care team could facilitate and show appreciation for the patients' families' involvement in caregiving, such as bathing. Again, however, the health care team must be sensitive to the patients' capabilities and preferences as their strength ebbs and flows during the course of the treatment, ensuring that they do not have unrealistic expectations for debilitated patients while at the same time encouraging others to be involved as much as possible.

In summary, an agency approach can augment efforts to implement patient-centered care by offering health care professionals an analytical template that uses these three lenses, in order to understand from different perspectives their patients' experiences throughout their serious illness and aggressive treatment. Indeed, the course of treatment is a process, throughout which the preferences and capabilities of patients and their families may shift. Revisiting the lenses of agency at various stages in the process can help health care professionals remain alert to their patients' needs and preferences throughout the experience. We thus suggest that insights from this approach can be fruitfully incorporated into implementation guidelines for both clinical and lay providers of health care services, and to that end, we have provided several examples of ways in which this can be accomplished.

Limitations and Further Research

As with any qualitative study, the strengths of our data—a rich set of narratives during a period when patients' own agency was threatened—require that the interpretation of the findings be accompanied by attention to the context and situation of the study. We employed a small convenience sample, selected from one organizational setting. We recognize that the narratives are undoubtedly influenced by the personal characteristics of the narrators, their relationships, and the particular setting and that a study with a different set of narrators and from a different setting might yield different results. Nevertheless, the series of interviews with a panel of patients and their lay caregivers over a two-year period offers rare insights into their experiences that would not otherwise be possible.

Our current study offers a framework on which future studies can be designed to explore an agency approach using the three lenses of agency experiences. For example, a longitudinal study of health care teams, before and after exposure to the conceptual framework developed here, may highlight the feelings of professionals working in a high-intensity treatment environment about the best way to adhere to a patient-centered care approach. A parallel study from the perspective of patients could reveal additional ways of applying this model.

In addition, although we have restricted our discussion of the agency perspective to high-intensity treatment, future studies could explore its application in less-intensive environments. That is, the three-lens perspective for reframing patients' experiences of agency need not be restricted to serious illness but could also be relevant to a range of health care, from ambulatory and primary care to chronic and acute care.

In conclusion, we believe that there is substantial promise in the framework offered here to contribute to how health care is provided, by offering new ways to understand and apply the relational and facilitative imperatives of patient-centered care to today's health care environment.