Advance care planning in Norwegian nursing homes – limited awareness of the residents’ preferences and values? A qualitative study

The objective of this study is to explore nurses and physicians’ aims and experiences of carrying out ACP in ordinary nursing homes. The present study is a first step in a larger research project, aiming to improve the autonomy and dignity of NHs residents suffering from life-threatening illnesses and dementia through better implementation of ACP [15].

The 850 nursing homes in Norway are part of the public health care system, and run by the municipalities. The population of Norway is 5.3 million people, and in 2018, 32,000 elderly persons permanently lived in NHs. Of persons aged 80 years and older, around 13% live in a NH, but this group makes up around 80% of the NHs population [16]. Mean age for persons who died in NHs in 2017 was 87,5 years old [17]. Women make up around 70% of the NHs population [18]. The low grade of institutionalization in Norway mirrors that independent living at home, as long as possible, is the ideological ideal, and influences health policy strategies. As a result, people are old and in need of extensive care and support when they move from home to a NH. Frailty, vulnerability, disability and multiple diagnoses [19, 20] characterize the NH population. Staff members are nurses, nurse assistants and physicians. There are large variations between municipalities and NHs when it comes to physician employment, ranging from full-time positions to GPs working 20% as medical supervisors [21].

While recruiting nursing homes for the study in 2014, we tried to identify NHs where health care professionals had some experience with ACP; where ACP had been practiced for at least 1 year. This turned out to be difficult, but in the end, we recruited eight NH wards (labelled as wards A, B, C…H). In these wards, we observed ACP conversations between NH’ staff, residents and/or next of kin [22]. Only in one of the conversations was the resident not present, due to severe dementia. The participating residents where assessed by the health care professionals to be competent. However, two of the residents had aphasia which no doubt made the conversations challenging. Next of kin were present in all conversations. The content of the observed ACP conversation was primarily related to questions concerning future medical treatment, hospital admission and DNR. After the conversations, we interviewed the resident and/or next of kin about their experiences of participating in ACP [23]. The interview guide was developed by all authors [24] (Additional file 1). The present study is based on seven group interviews with 20 NH staff (11 nurses, 8 physicians, 1 nursing assistant, and 1 nursing student) who had participated in the observed ACP conversations. For practical reasons, two to five staff members were interviewed together. Besides centering on what had taken place during the conversations, a semi-structured interview guide guided the interviews, covering issues like what staff members perceived as the aim of the conversation, and how they experienced the residents’ participation. The interviews were audio recorded and transcribed verbatim by the first author.

To analyse the entire data set, content analysis, as described by Schreier [25] was carried out. This approach, consisting of generating, defining and evaluating categories, is useful for identifying aspects that relate to the research questions, and for providing a good description of the material [25]. As a start, all authors read the transcribed interviews and discussed analytical ideas, thereby identifying the main categories. By main categories, we understand aspects of the material, which relate to or answer the study questions. The main categories were later systematically refined by the first author by generating subcategories which “specify what is said in the material with respect to these main categories” [25]. Creating subcategories means coding, that is, segments of individual staff members’ responses and answers during interviews have been ‘named’ and sorted. Generation of subcategories and naming of main categories have been discussed by all authors. Quotes illustrate themes in the results section, and are referred to by letter (NHs wards) and number.

Permission was granted from the head of the NH and from the Data Protection Official for the University of Oslo (NSD 37368). The study did not require approval from Regional committee for medical and health research ethics (2013/19937). Staff members were informed in writing and orally about the study, and asked if they wanted to participate. All agreed and gave their written consent.

Even though residents and next of kin were asked about future preferences, exploring resident preferences was not the main motivation for facilitating conversations concerning end-of-life care. Instead, nearly all participants described that the most important aim was to build alliances and come to some kind of agreement on future treatment with next of kin. The use of metaphors such as team, supporters and walking together (B44, E25, G18) underlines how much emphasis staff put on agreement with next of kin in order to avoid conflicts (G18), illuminated by the following quote: This is a conversation where we present how we see the resident’s state, and where we present how we think about medical treatment. I present some arguments about why we should change from active to palliative care…The aim is to find, or come to a common view about the resident, with the next of kin. To us, everything is logical and clear, but family members do not necessarily think the way we do. The conversation prevents misunderstandings and accusations between staff and next of kin when the time comes. We ask whether they agree with our understanding (C5, C9, C40).

Obtaining shared views could, however, take some time. Next of kin were described as needing time to digest information (F25), and as ‘ready’ if they have seen the same as us, have the same thoughts (C23). However, some family members never accept that their loved ones are dying: you may have told them, but the message is not received…it is difficult if your view is not shared with close family members. Some need a long time even if the time is limited (D 44).

The staff also found disagreement among next of kin difficult to handle: you are talking to one family member today, and another tomorrow, and he or she might think quite differently. That is a challenge. It is essential to talk to the right person (E26). According to one physician, next of kin could also have problems understanding their role and responsibility in the decision-making process: …and we talk to them (next of kin) face to face, and explain what their role in the decision-making hierarchy is and who makes the final decision. We find that some become somewhat unsure about their role; they become angry and we often have many discussions with them. This might lead to a difficult terminal phase; it can become a nightmare (B40).

Our study indicates that the most prominent motive for raising the issue of end-of-life care with residents and their next of kin was to prevent future misunderstanding, disagreement and conflicts between staff and family members. If it was possible to establish a common view about the resident’s situation and best interests concerning life-prolonging treatment, misunderstandings and conflicts could be prevented. There may be good reasons for this way of thinking. Conflicts with next of kin, particularly during the last phase of the resident’s life, are uncomfortable [26] and ethically challenging [27, 28]. That health care professionals, and in particular the physicians, found it demanding and hard to discuss end-of-life care, may also explain why they focused mainly on medical issues. From the perspectives of family members, mutual understanding between next of kin and health care professionals improves their experiences when someone close to them is dying [29]. However, although ACP aims to improve residents’ involvement in health care, in this study, less attention is actually paid to what matters to the NH residents; to their interests and wishes, and their abilities to participate in ACP. As we see it, there are good reasons for considering new ways of involving frail, older persons living in NHs.

Giving clear answers of future treatment at the end of life may be challenging. ACP is one way to avoid not talking about these issues. However, a too strong focus on directness and medical information may contribute to more hesitant and unclear answers, and less, rather than more patient centeredness [22].

Tensions concerning the aim and content of ACP may be related to how these conversations are organized and facilitated. From observing ACP conversations, we know that the physicians were the most active party, while the participating nurses were more passive [22]. The nurses did not take an active part in the exchange about future illness or dying in any way. In fact, one nurse claimed that she was happy to leave the difficult questions concerning end-of-life decision-making to the physicians. This has been described and discussed elsewhere [22]. This pattern was also present in the interviews, with physicians being quite active in describing and reflecting on how they carried out ACP, while the nurses contributed less. However, co-operation between health care professionals is considered important to improve resident involvement and increase the quality of ACP [30, 31]. Also, if we want ACP to cover broader aspects of the resident’s life; if we want to know more about the resident’s overarching philosophies and priorities in life [32] and the resident to become more involved in decision-making, other staff members should be involved in ACP. NH residents have daily, lengthy and often close contact with nurse assistants, unskilled staff, as well as the cleaning personnel. In this contact, ACP-relevant aspects may have been talked about and important relations may have developed. The possibility of involving these staff members should be explored. They may know what matters to the residents and can bring these insights into the ACP-process, contributing to a richer picture, also regarding planning for the future [33]. They may also understand more of what residents’ hesitance when it comes to expressing their views, actually is about, e.g. whether it is caused by the sensitive character of the issues or by cognitive impairment. This is particularly important when we know that many of the residents have dementia and reduced capacity, and that trust in staff members is important for successful ACP [33, 34]. Trusted staff and next of kin that knows the resident well may contribute to increase the resident decision making capacity and supported decision-making, rather than substituted decision-making. To improve the implementation and use of ACP in NHs, Gilissen et al. [14] suggests that all nursing home staff as well as volunteers receive training to help them recognize “triggers for ACP”, and how to engage in spontaneous ACP-related conversations. The importance of involving the whole ward in ACP is also described by Sævareid et al. [35].

End-of-life issues are sensitive [36], and there is a lack of educational programs teaching how to carry out such conversations. None of the participants in our study had been trained in end-of-life communication, advance care planning or how to involve residents or next of kin in ACP. The legal requirement and moral obligation to secure patient participation in health care require communication training programs, as well as the development of tools and meeting places, in order to build cultures where communication and decision-making processes are implemented and adapted locally [37]. Communication training may improve professionals’ self-efficacy and knowledge [38, 39], and should include reflections on the staff members’ own emotions [40]. Knowledge and availability of helpful pathways can contribute to reducing the burden on the staff [33, 40]. Training and guidelines may make it less stressful for residents and next of kin to participate, for example through emphasizing ACP as a process over time and that may include also more spontaneous conversation [15].