Hospital Executives’ Perceptions of End-of-Life Care

Hospital executives are key stakeholders in the hospital setting. However, despite extensive medical and nursing literature on the importance of end-of-life (EOL) care in hospitals, little is known about hospital executives’ perceptions of the provision of EOL care in their facilities. The objective of this study was to capture hospital executives’ perceptions of the provision of EOL care in the hospital setting. This descriptive, naturalistic phenomenological, qualitative study utilized in-person interviews to explore executives’ opinions and beliefs. The sample consisted of 14 individuals in the roles of medical center directors, chiefs of staff, chief medical officers, hospital administrators, hospital risk managers, and regional counsel in Arkansas, Louisiana and Texas. An interview guide was developed and conducted utilizing a global question followed by probes concerning perceptions of EOL care provision. Hospital executives acknowledged that EOL care was a very important issue, and more attention should be paid to it in the hospital setting. Their comments and suggestions for improvement focused on (a) current EOL care, (b) barriers to changing EOL care, and (c) enhancing provision of EOL care in the hospital setting. The findings of this study suggest that hospital executives although key change agents, may have insufficient EOL information to implement steps toward cultural and infrastructural change and should therefore be included in any EOL discussions and education.

and describe phenomena critical to the understanding of lived experiences (Morse & Field, 1995). This qualitative method allowed us to explore executives' opinions and beliefs more fully than might have been possible with other approaches.

Overview and Setting
We recruited hospital executives, as they are the stakeholders who have control over the availability, accessibility, and quality-of-health services and provide the environment in which physicians, nurses, and other caregivers can practice EOL care effectively and efficiently from the Veterans Health Administration (VHA) and community medical centers in Arkansas, Louisiana, and Texas. University and VHA institutional review board approvals were obtained prior to the study. We obtained written consent and addressed privacy protection with participants prior to conducting the interviews.

Data Collection
An interview guide was developed based on the literature and consultation with a qualitative expert, the fourth author (J.C.M.), and a health services expert, the seventh author (J.E.K.). Questions were pre-tested with two executives and were found sufficient to evoke responses that included the executives' perceptions of EOL care in their institution. All interviews were conducted by the first author (K.K.G.) in participants' facilities. The interview began with a global question followed by probes (Morse & Field, 1995) concerning his or her perceptions of EOL care provision. Interviews lasted 0.75 to 1.5 hr and were audio recorded, transcribed verbatim, and verified for accuracy. Hospital executives not employed in the VHA system received a US$50 check for participation.

Data Analysis
Transcripts were entered into Atlas-ti (Version 5.2, Scientific Software Development, Berlin), a text-based data management program that allows data labeling and sorting by identified variables. The interview data were coded using content analysis, as described by Morse and Field (Morse & Field, 1995) in which a systematic, objective analysis of message characteristics are identified, with constant comparison, or continual analysis of data by comparing one interview to another, in an iterative process (Glaser & Strauss, 1967). The first author coded the initial three interviews, developed tentative definitions for each code, and prepared a preliminary codebook, and then the seventh author reviewed the codebook and transcripts. Emerging codes and coding differences were discussed until consensus was reached. The first author repeated this process with all interview narratives, comparing one to the next, and coded the remaining interviews. Narratives were aggregated by similarity, forming categories from the aggregates, and developing overarching themes of the perceptions of the executives. After 14 interviews, considerable repetition of data was identified, indicating data saturation (Morse & Field, 1995).
Several techniques were used to maximize the validity and reliability of the study, following guidelines for trustworthiness for qualitative studies (Lincoln & Guba, 1985). Dependability was enhanced through the use of one interviewer (K.K.G.) and one interview guide for all interviews. In addition, verbatim transcription and thoroughly reviewing the transcripts against the recorded interviews ensured validity. The first author enhanced credibility and confirmability by keeping a detailed audit trail of coding, theoretical decisions, and consensus with the health services expert. Unedited examples of narrative were used to validate findings. The separate coding by the two authors yielded strong agreement (86%).

Subjects
The sample consisted of 14 individuals in the roles of medical center directors, chiefs of staff, chief medical officers, hospital administrators, hospital risk managers, and regional counsel in Arkansas, Louisiana, and Texas. These executives were responsible for services provided by hospitals with a median bed number of 740, ranging from 277 to 1,299. They had a mean age of 60 years (range = 46-68), and 57.1% of the executives were male. There were 8 VHA (57.1%) and 6 community (42.9%) medical centers represented from metropolitan statistical areas, which ranged from 150,000 to well above 1,000,000. Out of concern for inadvertent disclosure of the executives' identity, no further information about the hospitals and locations is provided.
It was widely acknowledged by the executives that EOL care was a very important issue and more attention should be paid to it in the hospital setting. Their comments focused on (a) current EOL care, (b) barriers to changing EOL care, and (c) enhancing provision of EOL care in the hospital setting.

Current EOL Care
All the participants described EOL care as an important concern for hospital executives; yet they gave different reasons for its importance. For some, a major concern was the lack of preparation that patients and families had regarding their prospective needs for EOL care. The executives noted that very serious and chronically ill patients arrived at the hospital without clearly defined goals of care and with potentially unrealistic expectations for what the hospital was going to be able to do for them. One executive said, I mean you get patients here who are very sick and who may have a low chance of survival to begin with, you get patients who are here thinking they're coming to the "Mecca" and they're finally going to get the you know [the] "savior" that they've not been able to get somewhere else . . . and quite frankly that can be more than somewhat of a challenge . . .. (Interview 8) Several executives expressed concern about the current practice of avoiding or shifting EOL decision making to patients and families who had little preparation and often low literacy or personal barriers to making these difficult and complex decisions. Another hospital executive said, Physicians shift a lot of burden to families . . . especially when the physician doesn't want to make a decision. And my basic thought on that is that it's the physician's job, that's not generally left to lay people. (Interview 12) Because of current organizational and system processes, health care providers frequently have little to no long-term relationships with patients or their caregiver/families. Another executive noted, [The lack of a long-term relationship] is exacerbated by hospitalists and intensivists and all these different roles, which make a lot of sense . . . to have critical care people at the bedside . . . What it now means though is when you're at your sickest you're probably with somebody you know you've never met, your family's never met, you know . . .. (Interview 2)

Barriers to Change
These hospital executives pointed to several difficulties in providing quality EOL care in the hospital setting. Personnel shortages in all areas of health care were frequently mentioned as barriers to effective EOL care. Lack of time to adequately discuss EOL care with patients and families was also frequently cited. The executives noted the significant amount of time such discussions take and the increasing need for these discussions along with ongoing shortages of health care providers. In addition, many of the executives felt that health care providers, even if they had adequate staffing and were willing to engage in an EOL care discussions, did not have the training to do so. One executive said, We don't have enough physicians to do that with the elderly . . . I mean golly there's so few physicians who are trained in that area or have an interest in it, we have one physician that is very good at these discussions . . . but it takes so much of her time and she's patient and willing, but we don't have many physicians like that. (Interview 7) Another executive with medical training noted that it was most difficult to have discussions with patients at the EOL who expected treatment but for whom current medical treatments would not be effective. He explained, You know . . . I grew up in an environment where I had to learn, learn, learn, but I never learned how to talk to somebody where there's no hope . . . and I'm not equipped to have that kind of conversation. (Interview 10) In addition, executives noted that financial and care management restrictions result in a fragmented process of care, with shorter inpatient stays and rapid transfers to lower levels of care or outpatient management. One hospital executive said, Case management has been used so heavily to keep length of stay short, make sure people get to the right level of care and discharged timely. But how do you get to that other whole population [patients at EOL] resourced early with a plan? I don't think case management . . . has really shown us that they have expanded their net in this arena yet. (Interview 10)

Changing EOL Care
The executives described many strategies to improve EOL care in the hospital setting. Many indicated the need for a champion at each facility. One executive said, So if you have one or two who are leaders . . . either officially or unofficially . . . on the medical staff who will say "no . . . this is a valuable service . . . this is a valuable program . . . we need to support it . . . we need to be open to it." That, I think is essential. (Interview 7) They also recognized the importance of their own role in the process. The executives noted that EOL care, although critically important for patients and for families, is essentially an unappreciated endeavor. However, many recognized a need for sustained system changes, and they proposed the development of outcome measures that recognize the provision of EOL care.
Who's out there saying today, you know, isn't it great that Dr. "so and so" had four patients die comfortably in the proper way with their family at their side? But you hear when so and so finished a procedure in twelve minutes . . . I mean what do we really recognize people for in medicine? (Interview 8) In addition to outcome measures, the executives suggested the development of a hospital infrastructure and systems to support EOL care: I would say that we need to acknowledge our limitations today and look for ways to strategically begin to look for statistical projections to be able to develop a plan that addresses infrastructure development of services over time. (Interview 1) The executives suggested that this infrastructure should include education; EOL care should not only be an essential component of current curricula but a part of continuing education of the current health care force and the general public: I think a lot of it comes back to education and I don't know what we're doing even in physician education to make people in training aware of the needs that are out there and the programs like palliative medicine that can be of benefit. I don't know, I'm almost certain in our master's degree program in health administration we're not teaching that. (Interview 7) The executives also indicated a critical need to educate the general public.
And a lot of that then you know the final thing is educating the general public because so much of this is American culture. You know this I want everything, every last thing done to me. I find it interesting because most people, many people . . . with more of a clinical background are like, uh, uh, I don't want that . . . I don't want to be put through some of the things that you know we put patients through. I don't think a lot of patients and their families have those conversations, understand that, etc. and so we get ourselves into those kinds of positions as well. (Interview 14) Discussion This is one of the first qualitative studies to examine a range of hospital executives' perceptions on and experiences with the provision of EOL care in acute care settings. It is clear from the participants' accounts that a complex range of barriers exist in the provision of EOL care. The responses of hospital executives indicated that there is a need for focusing efforts on EOL care, and they identified some potential strategies to overcome barriers that were identified.
Hospital executives characterized current EOL care as care in which the burden of decision making is commonly shifted to families when patients become incapacitated at or near the EOL. Often, health care providers demonstrated discomfort with discussing palliative care with patients and their families, which can delay or limit discussions. As noted by Braun (Braun & McCullough, 2011), this can have dire consequences because these are very difficult decisions, with an emotional and socioeconomic overlay that most families are poorly prepared to make. Our results concur with other literature that has found that with little or no guidance from the patient, the family may have difficulty identifying and expressing the patient's preferences rather than their own (Braun & McCullough, 2011;Garner, Henager, Kirchner, & Sullivan, 2011;Quill, 2000). In parallel with family uncertainty and increasing dependence on the provider, others have also suggested that providers may be paternalistic by guiding the families toward their preferences, instead of the patients (Breier-Mackie, 2001;Clark, 2012;Marik, Varon, Lisbon, & Reich, 1999). The sometimes paternalistic view of the provider was not found to be a salient theme identified by hospital executives in this study. A prominent barrier identified in this study was the lack of adequate time and training to prepare patients and their families for the end of life, limit unnecessary treatment, and focus care on achieving wishes and improving quality of life. Hospital executives also noted that a fragmented care process contributed to poor communication and barriers to effective EOL care. While lack of time and training have been shown to be barriers to advanced care planning in primary care, our findings, as well as others (Meier, 2011), confirm that these constraints also affect acute care providers. This may be additional area for case managers, who are often on the front lines of communicating with and arranging discharges and transitions in care based on these goals (Ahalt et al., 2012).
Some of the hospital executives in this study suggested changes in communication about preferences at EOL. They suggested that EOL care could be enhanced by linking case management with interdisciplinary care to assist health care providers in proactively establishing goals of care. Recent research identified improved communication as a high priority, specifically regarding preferences of care at EOL (Sadler et al., 2014). This research concluded that inadequate communication predicted family dissatisfaction with care. Case managers could initiate palliative care assessments and conversations about setting goals of care, even if only for the current hospitalization. Identifying and communicating these preferences with families and health care providers early in care, before a patient becomes incapacitated, could enhance the ability of health care providers to achieve a patient's wishes, even if only for the duration of the current hospitalization. These assessments rarely occur in the community, and as has been found by Meier (Meier et al., 2004), combining palliative care with case management is a "logical, feasible and effective strategy to improve the care" of patients with a lifelimiting illness. By initiating the assessment and discussions, case managers may be able to facilitate subsequent communication and planning with primary care providers and providers in other settings such as long-term care facilities.
Hospital executives indicated that improving education about EOL care (Cooper et al., 2010;Lamba, Pound, Rella, & Compton, 2012;Smith, O'Sullivan, Lo, & Chen, 2013) and options for care is a critical need (Coulourides, Brumley, Wilber, & Enguidanos, 2012;Morrison, 2013). However, they noted that hospital stays are relatively short and frequently not the best time to have extensive discussions about such an important topic. Recently, Fink and colleagues (Fink, Oman, Youngwerth, & Bryant, 2013) surveyed administrators about palliative care and similarly concluded that inadequate knowledge by practitioners should be improved, especially in rural hospitals where resources are scarce. While they acknowledged that hospitals have an important role to play in education, they thought that established organizations with education as a primary mission (e.g., the American Heart Association, American Lung Association, etc.), local community or faith-based organizations and state or local health departments could address advance care planning with individuals and families and address specific disease concerns or cultural or spiritual concerns (Field & Cassel, 2011).
Executives indicated that there is a great need to use administrative and clinical tools to benchmark and track the quality of EOL care. Key performance indicators (Twaddle et al., 2007), quality outcome measures (Casarett et al., 2008;Morrison, 2013), and evidence-based practices (Qaseem et al., 2008) have been developed for use in palliative care settings. These measures could be adapted and used to improve hospital-wide care by recognizing health care providers and systems that achieve/exceed indicator measures and by developing compliance plans for those that do not. As Twaddle (Twaddle et al., 2007) has reported, hospital systems need to be in place to identify patients with potential life-limiting illnesses and address the physical, psychological, social, and spiritual needs of these patients and their families.
In considering how to improve EOL care in hospital settings, the executives acknowledged their own leadership role but, at the same time, suggested that large-scale cultural change was needed to move providers from "cure" to "care" and toward an appreciation for the provision of EOL services. The hospital executives in this study noted that such change could be achieved through infrastructure that supports the development of champions, as well as ongoing education of providers, patients, and the general public. The findings of this study suggest that hospital executives, although key change agents, may have insufficient EOL information to implement steps toward cultural and infrastructural change and should therefore be included in any EOL discussions and education.
It is important to consider the limitations of this study. Because of its qualitative nature and small sample, further research is needed to investigate the prevalence of the barriers and potential solutions noted by these executives. In addition, this study was conducted in the Southern "Bible belt," which has been associated with more intensive lifeprolonging medical care near death (Phelps et al., 2012); thus, the executives' perceptions of patient expectations and tendency of providers to shift responsibility to patients and families may be a function of geographically prevalent beliefs and coping styles.