Advance Care Planning in Nursing Home Patients With Dementia: A Qualitative Interview Study Among Family and Professional Caregivers

JAMDA 16 (2015) 979e989

JAMDA journal homepage: www.jamda.com

Original Study

Advance Care Planning in Nursing Home Patients With Dementia: A Qualitative Interview Study Among Family and Professional Caregivers Mirjam C. van Soest-Poortvliet MSc a, *, Jenny T. van der Steen PhD a, *, Giselka Gutschow MSc b, Luc Deliens PhD c, d, Bregje D. Onwuteaka-Philipsen PhD c, Henrica C.W. de Vet PhD e, Cees M.P.M. Hertogh MD, PhD a a Department of General Practice and Elderly Care Medicine, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands b VU University Medical Center, Amsterdam, The Netherlands c Department of Public and Occupational Health, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands d End-of-Life Care Research Group, Vrije Universiteit Brussel and Ghent University, Brussels, Belgium e Department of Epidemiology and Biostatistics, EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, The Netherlands

a b s t r a c t Keywords: Advance care planning dementia nursing home long-term care

Objective: The aim of this study was to describe the process of advance care planning (ACP) and to explore factors related to the timing and content of ACP in nursing home patients with dementia, as perceived by family, physicians, and nurses. Design: A qualitative descriptive study. Methods: A total of 65 in-depth qualitative interviews were held with families, on-staff elderly care physicians, and nurses of 26 patients with dementia who died in the Dutch End Of Life in Dementia (DEOLD) study. Interviews were coded and analyzed to find themes. Results: Family, nurses, and physicians of all patients indicated they had multiple contact moments during nursing home stay in which care goals and treatment decisions were discussed. Nearly all interviewees indicated that physicians took the initiative for these ACP discussions. Care goals discussed and established during nursing home stay and the terminology to describe care goals varied between facilities. Regardless of care goals and other factors, cardiopulmonary resuscitation (CPR) and hospitalization were always discussed in advance with family and commonly resulted in a do-not-resuscitate (DNR) and a do-not-hospitalize (DNH) order. The timing of care planning discussions about other specific treatments or conditions and the content of treatment decisions varied. The factors that emerged from the interviews as related to ACP were general strategies that guided physicians in initiating ACP discussions, patient’s condition, wishes expressed by patient or family, family’s willingness, family involvement, continuity of communication, consensus with or within family, and general nursing home policy. Two influential underlying strategies guided physicians in initiating ACP discussions: (1) wait for a reason to initiate discussions, such as a change in health condition and (2) take initiative to discuss possible treatments (actively, including describing scenarios). Conclusions: ACP is a multifactorial process, which may lean on professional caregivers’ guidance. The most acute decisions are covered in advance, but a responsive as well as a proactive style is seen with

The authors declare no conflicts of interest. This study was funded by a career award provided to J.T.v.d.S. by the Netherlands Organisation for Scientific Research (NWO, The Hague; Veni 916.66.073 and Vidi 917. 11.339), and by a grant from ZonMw, The Netherlands Organisation for Health Research and Development (Palliative Care in the Terminal Phase program, grant number 1151.0001), and by the Department of General Practice and Elderly Care Medicine, and the Department of Public and Occupational Health of the EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam. Preliminary results were presented at the 2011 International Psychogeriatric Association congress in The Hague. An abstract of this work was published as: van http://dx.doi.org/10.1016/j.jamda.2015.06.015 1525-8610/Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

Soest-Poortvliet MC. Barriers and facilitators of (timely) planning of palliative care in patients with dementia: A study in Dutch nursing homes (S07A.02). Part of Symposium 07: Palliative ways and means: Improving end-of-life care for people with dementia (2011). Symposia. International Psychogeriatrics, 23 (S1), p S29. * Address correspondence to Mirjam C. van Soest-Poortvliet, MSc, and Jenny T. van der Steen, PhD, Department of General Practice and Elderly Care Medicine, EMGO Institute for Health and Care Research, VU University Medical Center, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands. E-mail addresses: [email protected] (M.C. van Soest-Poortvliet), [email protected] (J.T. van der Steen).

980

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

other treatment decisions. Further research is needed to increase understanding of whether and how the physicians’ strategies affect care processes and outcomes. Ó 2015 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

With the progression of dementia, care focused on comfort or quality of life, as in palliative care, may become more appropriate than curative care.1e5 Advance care planning (ACP) is a key component of quality of palliative care,6 also for nursing home (NH) patients with dementia.1,7,8 ACP is an anticipatory decision-making process among patients, family, and professional caregivers about prioritizing care goals and future care that will be considered appropriate when taking into consideration the patient’s preferences.9e11 It is different from traditional care planning, as it makes plans in anticipation of future deterioration in a person’s condition.12,13 The course of dementia is difficult to predict,14 which makes it difficult to determine the right time to start planning palliative care.15 ACP may allow patients and family more time to deal with the complex emotional and ethical decisions16 and may help family of NH patients to prepare for the death of their loved one.17 The mean or median length of stay of patients with dementia in longterm care facilities is often no more than a few years.18,19 For example, in the Dutch End Of Life in Dementia (DEOLD) study, a prospective observational study, the median of length of stay in the NH was approximately 2 years.20 Therefore, admission to a longterm care facility may be a good opportunity to initiate or reboot ACP discussions in which care goals can be discussed and care preferences identified. Interest in ACP is increasing, both as an element of high-quality palliative care and as a possibility for patients to retain control over their care and medical treatment in the event of incapacity.5,21e24 ACP for patients with dementia is important, as these patients will progressively lose decisional capacity with advancing illness. Also, family has an important role in ACP with dementia, as they will need to take over decision-making with advancing illness and will live on with the memories after the patient has died. ACP is used in NHs. However, family may first need to adjust to the patient’s admission to the NH; the terminal character of dementia is not clear to every family or professional caregiver.5,18 As a result, ACP is not always started at admission, and care planning is not always focused on future care. To improve ACP in NHs, it is important to know how the different actors (family and professional caregivers) experience the current ACP, including palliative care, in NHs and which factors are related to ACP as viewed from these different perspectives. Previous studies that identified factors that could affect ACP of patients with dementia reported mostly about family and professional caregivers separately and were not always performed in long-term care facilities such as an NH.20,25e30 The purpose of this qualitative study was to (1) describe the experiences with the process of ACP in Dutch NH patients with dementia, and (2) explore factors related to the timing and content of ACP as perceived by family, physicians, and nurses. Methods Design We used a qualitative descriptive methodology: explorative, qualitative in-depth interviews of family, physicians, and nurses using a topic list, and thematic analysis. Sample Selection In this qualitative interview study, we sampled from a larger project about care and outcomes in NH patients with dementia, the

DEOLD study.20 In the prospective data collection of this DEOLD study, on-staff elderly care physicians and family of patients with a physician’s diagnosis of dementia who were newly admitted to the facility between January 2007 and July 2009 (n ¼ 372) completed questionnaires between January 2007 and July 2010 (survival was monitored until July 2011); 8 weeks after admission of the patient (baseline), semi-annually, and soon after death. In the homes that collected data retrospectively in the DEOLD study between November 2007 and March 2010, on-staff elderly care physicians and family of NH patients with dementia (n ¼ 119) completed questionnaires only after death. All questionnaires included items on discussion and evaluation of care goals. From the summer of 2009, in addition to the physician and family, the responsible nurse completed a questionnaire after the death of a patient. The contents of this questionnaire overlapped with the contents of the physician questionnaire and were used only to abstract the demographics of the nurses we interviewed. Twenty-seven of the 34 facilities that participated in the DEOLD study consented to participate in this qualitative interview study. We selected cases of deceased patients from the patients who died during the last year of data collection (July 2009 until July 2010) of the DEOLD study in 1 of the 27 facilities. Based on information known through the quantitative DEOLD data collection, we intended to include (1) patients with variable timing of establishing comfort care goals, (2) patients with families who indicated in the questionnaires that the timing of ACP discussions was right and not right (too early or too late), and (3) patients with different triggers for providing palliative care according to the physician (eg, deterioration in health condition or diagnosis of dementia). Variable timing of establishing comfort care goals included patients with a comfort care goal shortly after admission, at a later moment during the patient’s stay, and only at death or not at all according to the physician questionnaires. In the Netherlands, in the palliative care for people with dementia, a distinction is made between a palliative care goal and a symptomatic care goal.31 Palliative and symptomatic care goals both refer to comfort, quality of life, and well-being, but they differ as to whether prolongation of life is desirable. By a comfort care goal we mean both a palliative care goal and a symptomatic care goal. With this purposeful selection, we aimed to include many different situations in this qualitative interview study. The Medical Ethics Review Committee of the VU University Medical Center approved the study protocol. Data Collection In 2010, we contacted a family caregiver, the elderly care physician, and the responsible nurse of the selected patient to make an appointment for an individual interview. Interviews were held approximately 3 to 6 months after the patient’s death. We planned the interview with family first if possible, and maintained this order in each patient case. All interviewees gave oral consent. Based on saturation, a total of 65 in-depth interviews were held with 20 family caregivers, 21 physicians, and 24 nurses about 26 patients with dementia who died in 16 facilities. For 25 of these patients, data were collected prospectively in the DEOLD study, and for 1 patient, data were retrospectively collected. Perspectives of all 3 caregivers were available on 14 of the 26 patients; perspectives of 2 caregivers on 11 patients, and of a single caregiver on 1 patient. Six family caregivers of the 26 patients did not participate in the interview study: 3 family caregivers declined

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

(2 caregivers because they were not interested and 1 because of grief) and 3 family members were not reached. Four physicians did not participate: 3 physicians were no longer employed by the participating facility and 1 physician indicated a lack of familiarity with the particular patient because he was involved only in the final 2 weeks. One nurse did not participate because she was not reached. We could not include perspectives of 1 physician and 1 nurse, as no audiotape was available due to technical problems and no interview reports were available. Also, there was no audiotape for 5 of the 65 interviews used for analyses, but reports of these interviews were available. These interview reports were not used to explore new factors related to ACP, but only to confirm factors identified in other interviews. Confidentiality was maintained by not providing any information shared with the researchers to other interviewees about the same patient. Fifty-seven interviews were carried out by telephone; 8 interviews were held face-to-face. We carried out mainly telephone interviews for feasibility because the interviewees recruited via the national study were from across the country. We held 8 face-to-face interviews, including 2 with family, about 5 patients who died in nearby facilities, to examine if face-to-face interviews would yield additional information that could not be collected in telephone interviews. The physicians and nurses received a check of 25 Euros to compensate them for their time. Two female master-level researchers (MCS-P and GG) with backgrounds in health sciences and psychology conducted the interviews. Both researchers had received training in qualitative interview techniques and had been involved in dementia research in NHs for several years. Before the interview period, a topic list was drawn up together with the project group members with experience in qualitative research with topics that should be addressed during the interviews. During the interview period, the project group members supervised the interviewers and the interviewers regularly informed the project group members about the interview findings. The interviewers went into the interviews with an open attitude and all interviews started with an introductory question: “Could you tell me about the period that your family member/this patient lived in [name NH], and about how you experienced this period?” Subsequently, in a nonestablished order, but following up on the answers provided by interviewees, the following topics of the topic list were addressed: the perception of the contacts between family and professional caregivers, the timing of contacts and the reason for contact, the timing of treatment decisions and factors related to the timing of these decisions, the content of the treatment decisions, factors related to the content of these treatment decisions, definitions and communication of care goals, and the role of general practitioners (GPs) in planning end-of-life care. After all the topics were addressed, the interviewees were asked whether they wanted to mention any additional issues. We asked the interviewees to talk about the entire period that their family member/their patient lived in the NH, as we expected to identify more factors when we asked family more openly to recount the period that their family member/their patient lived in the NH. We did not explicitly refer to ACP in our questions but asked interviewees to talk about planning of end-of-life care, because different people may have different definitions of ACP. In addition, we talked about end-of-life care in our questions and not about palliative care, because palliative care is a form of end-of-life care and may be defined differently by different people. However, to gain insight into the various definitions, we asked physicians and nurses about their definition of a palliative care goal and a symptomatic care goal. We also asked family caregivers how physicians communicated about the care goals.

981

Data Analysis All interviews were transcribed verbatim, checked for accuracy, and then coded. Initial coding of the interviews started early during data collection and allowed exploration of new issues in subsequent interviews. Two researchers (MCS-P and GG) independently coded the transcripts of the first 8 interviews and compared codes, discussing differences until consensus was reached on a uniform coding system. The next interview transcripts were coded by the first researcher (MCS-P). Further, selected interviews (eg, when in doubt how to code it or when the interviewee, mostly family, was hard to follow) were coded by 2 or 3 researchers (MCS-P, GG, and CMPMH). This process was supported by the software program Atlas.ti 5.2 (Scientific software development gmbh, Berlin, Germany). After all the interviews were coded, we made a list of codes per type of interviewee. Two researchers (MCS-P and JTS) independently categorized the physicians’ codes into themes and we discussed differences until consensus was reached. The first researcher (MCS-P) then categorized the family and nurse codes in the same themes if possible, and added additional themes when necessary. The other researcher (JTS) checked this categorization. Appendix 1 lists all themes that emerged from the interviews and the research purpose for which it provided relevant information. Within the themes, we looked at findings relevant to the two research purposes (to describe the experiences with the process of ACP and to explore factors related to timing and content of ACP). After identifying the findings relevant to the research purposes, for illustration, we selected 3 information-rich patient cases with multiple caregiver perspectives on 1 patient that covered several findings relevant to the 2 research purposes. Results Twenty-six patient cases were discussed in the interviews. The quantitative DEOLD study showed that for these 26 patients the mean age at death was 83 years; 17 of them were women, 6 patients had severe dementia on admission, 12 patients had a comfort care goal within 8 weeks after admission, and, at death, 23 patients had a comfort goal (Table 1). According to the physician after-death questionnaire, in 9 patient cases, the most important trigger to provide comfort care was an obviously poor prognosis. The mean age of the 20 family caregivers was 61; 9 of them were women and 11 of them were the patient’s child. Fourteen of the 21 physicians and 22 of the 24 nurses were women, with a mean experience of 15 years for physicians and 20 years for nurses. The average duration of interviews was 30 minutes (range, 15 to 90 minutes). In the interviews, the interviewees described the period that their family member (the patient) lived in the NH, and how they experienced contacts between family and professional caregivers and decisions that were made about care goals and treatments. In the shortest interviews, interviewees had little to report in relation to the topics from the topic list. In the following sections, we describe interview results that relate to the 2 research purposes of our study: description of the ACP process and factors related to it. In addition, to illustrate the experiences with ACP of family, physician, and nurse regarding 1 patient, we present 3 information-rich example patient cases. Research Purpose 1: Description of the Process of ACP Multiple contact moments between family and professional caregivers In general, family members, physicians, and nurses indicated in the interviews that there were multiple contact moments during the

982

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

Table 1 Characteristics of the 26 Patients With Dementia and the Interviewees: Their Family Caregivers, Physicians, and Nurses

Mean age (range) Female gender, n Dementia severity at baseline (BANS-S),z mean (range) Severe dementia at baseline,z n Mean length of stay in years (range) Comfort care goal at baseline,x n Comfort care goal at death,x n Families’ opinion about the timing of the first ACP discussion in relation to the patient’s health, n Too early Just right Too late Don’t know Had no ACP discussions Relation to the patient, n Partner Son/daughter Son/daughter in law Brother/sister Friend Physician felt the most important “trigger” for providing palliative or symptomatic (comfort) care was, n An obviously poor prognosis Deteriorating condition Intercurrent disease Patient wish before diagnosis of dementia: Do not want to live this way No comfort care provided Missing Years of experience in care (range)

Patient Cases, n ¼ 26

Family Caregivers, n ¼ 20

Physicians, n ¼ 21

Nurses, n ¼ 24

83 (63e95) 17 13 (8e20) 6 1.6 (0.07e2.81) 12 23 d

61 (46e86) 9 d

44 (31e59)* 14 d

44 (23e59)y 22 d

d d d

d d d d

d d d d

d

d

1 8 1 4 6 d 6 11 1 1 1 d

d

d

d 9 4 2 1 4 1 15 (3.8e30)y

20 (5e41)z

*Age and years’ experience was missing for 3 physicians. y Age and years’ experience was missing for 14 nurses because we asked nurses to also complete questionnaires only in the last phase of data collection. z Measured with the Bedford Alzheimer Nursing Severity-Scale (BANS-S). Rating form 7e28; higher score means more severe dementia. A score of 17 or higher refers to “severe dementia”. x Comfort care goal ¼ palliative care goal or symptomatic care goal. (Volicer L, Hurley AC, Lathi DE, et al. Measurement of severity in advanced Alzheimer’s disease. J Gerontol 1994;49: M223eM226; Bellelli G, Frisoni GB, Bianchetti A, et al. The Bedford Alzheimer Nursing Severity Scale for the severely demented: Validation study. Alzheimers Dis Assoc Disord 1997;11:71e77; van der Steen JT, Volicer L, Gerritsen DL, et al. Defining severe dementia with the Minimum Data Set. Int J Geriatr Psychiatry 2006;21:1099e1106.).

NH stay in which care goals and treatment decisions were discussed. Families indicated that most contacts about daily care were with nurses, but indicated that care goals and treatment decisions were discussed with the physician and mostly in the presence of the responsible nurse. Almost all interviewees identified that the initiative for care-planning discussions came from the physician. Only 2 family members in the interview said that they took the initiative to bring up care goals and treatment decisions. Physicians said that the role of GPs in planning of end-of-life care is mostly limited to the transfer of the patient’s chart. The chart often only refers to the availability of an advance directive or a euthanasia statement (see more detail in Table 2, c, wishes expressed by patient or family). In some NHs, there were fixed contact moments between family and professional caregivers, whereas in other NHs, contact moments were planned as needed. Several physicians said that planning of end-oflife care is a process and that the contact frequency with family increased as death approached. Not all patients have the same care goal When asked about the care goals discussed and established during patients’ NH stay, physicians and nurses described a great variety in care goals: “palliative goal,” “symptomatic goal,” “comfort goal,” “treat in conformance with the NH policy,” “treat within the possibilities of the NH,” “conservative policy,” “proportionally active policy,” “active policy in NH,” and sometimes “a curative goal.” Some of the care goals are comparable with a palliative care goal and other goals referred to curing intercurrent conditions. Physicians indicated that in general a curative care goal is established only shortly after admission and a care goal never changes into a curative care goal later during NH stay.

In their definition of a palliative care goal and symptomatic care goal, physicians frequently used the terms “quality of life,” “comfort,” “symptom control,” and “life-prolonging treatments are not desired.” Nurses were more familiar with the term symptomatic care than with palliative care and for physicians it was easier to give a definition of palliative and symptomatic care than for nurses. Physicians’ and nurses’ perceptions of palliative care and symptomatic care goals differed. Some physicians and nurses thought a palliative care goal was relevant, as dementia is a terminal disease, whereas some physicians and nurses thought a palliative care goal was relevant to all NH patients. Other physicians and nurses thought a palliative care goal was only relevant when the patient’s condition was life-threatening or the patient’s quality of life was poor. Most physicians indicated that they actively explained and described care goals to avoid ambiguity. According to the physicians, the terms of the care goal as stated in patient charts were sometimes mentioned in physician-family communication. Some physicians also indicated they purposefully explained care goals in patient charts, to clarify them for all persons involved. A few family members indicated that they were familiar with the terms palliative and symptomatic care, as they were mentioned in the discussions or in the NH forms. Research Purpose 2: Factors Related to ACP The interviews revealed reasons for and content of discussions about cardiopulmonary resuscitation (CPR) and hospitalization; timing and content of these discussions are not very variable. In addition, multiple factors related to the timing and content of other treatment decisions.

Table 2 Factors Identified from the Interviews as Related to the Timing and Contents of ACP Discussions and Decisions of NH Patients with Dementia Description

Identified in Interview(s) With

Quote

a. Physician’s strategy

We found 2 different general strategies that guided physicians in their planning of end-of-life care: (1) discuss and make decisions about CPR and hospitalization of the patient, and subsequently wait for a reason to initiate discussions about other treatment discussions, such as a change in condition/condition occurs and (2) in addition to decisions about CPR and hospitalization, physician takes initiative to discuss other possible treatments (actively, including describing scenarios).

Family Physician Nurse

b. Patient’s condition

The condition of the patient includes difficulty to predict survival, severity of dementia, change in condition, quality of life, age, and the situation before admission. As indicated in the strategies, for some physicians, the change in the condition of the patient is the trigger to take the initiative to discuss specific treatment decisions. However, a decline in condition is also a reason for physicians who discuss treatment decisions in advance to discuss and evaluate advance treatment decisions once more. In case of health decline, treatment decisions mostly focus on comfort.

Family Physician Nurse

c. Wishes expressed by patient or family

The physician’s knowledge of an existing advance directive facilitated initiating the discussions about end of life by the physician. Additionally, previous or recent expressions of patient wishes by patients or family make it easier to make treatment decisions. For example, the patient’s wish of not to pursue pointless extension of life or the opposite wish of to live as long as possible. The role of GPs in planning end-of-life care is mostly limited to the transfer of the patient’s chart. The chart often only refers to the presence of an advance directive or a euthanasia statement. Some physicians said that the role of GPs should be extended.

Family Physician Nurse

Strategy 1 “You have discussions at the moment you need to. . you talk about it at the moment it comes up. There is little point in presenting an entire trajectory about something that may or may not happen. In that sense the answer is very simple: the timing was right. At the moment it was necessary.” (Family about Mrs. E.) “It is my experience that you still have time to discuss the situation that has developed at that moment with the family and to start a palliative policy at relatively short notice.” (Physician, about Mrs. D.) “Many things never happen. If you discuss the whole 9 yards every time, people forget. . .They are given so much information that it won’t stick.” (Physician about Mr. S.) Strategy 2 “The physician indicated that we needed to plan our first discussion a month after admission . I liked the idea, you know, what can we expect and it made you feel that things will be alright when a problem does occur.” (Family about Mrs. G.) “Well, the purpose of discussing policy at an early stage is that when a person starts to deteriorate, you know what to do, you don’t first have to educate and inform the family, but you can fall back on a conversation you have already held.” (Physician about Mrs. W.) “It is very important that the family can think about this decision of treatment or nontreatment. If you ask people at the last moment, they often choose the lifeprolonging treatment. In my experience they tend to change their minds later on.” (Physician about Mrs. H.) “My father developed full-blown dementia in a time span of 3 months, from living on his own to institutionalization . In view of the speed at which everything had happened at that point, I was very happy that he [the physician] brought it up fairly quickly.” (Family about Mr. A.) “That was a positive conversation emphasizing that she enjoyed life, even now on the psychogeriatric [dementia] ward. And that there was therefore no reason to decide to a conservative policy, except for a DNR policy. We would start the discussion when she deteriorated.” (Physician about Mrs. N.) “At the moment that a person deteriorates, it provides an opening, to include in conversations, like well, we now see the deterioration, we discussed this policy back then: do you still agree with it, and then that is how we enter the final part of the final phase.” (Physician about Mrs. H.) “Luckily I asked before, before he became ill, more ill, and he said then, I said ‘what if you have to stay in bed for days?’ life was still worth living then. ‘What if you have to stay in bed all day and you are deaf and blind?’ no, that would be the tipping point. If there was nothing to enjoy, well, then he was done . It did help me a little, the thought that maybe he would be at peace with it now.” (Family about Mr. Z.) “This couple always said to each other that they did not want to be kept alive at all costs. So we agreed not to pursue pointless extension of life and to apply DNR.” (Physician about Mr. T.) “Many GPs don’t really understand, they just send the chart and that does not require a verbal transfer . And I also sometimes phone the GP if someone breaks down in the first 2 weeks and I haven’t had the chance to get to know the person.” (Physician about Mrs. V.) “The role of the GP in formulating decisions for the end of life; whether the GP should always bring it up. I wonder whether he should. The only thing I do think is that when a person is diagnosed with dementia, that you must give him the opportunity before he becomes completely incompetent.” (Physician about Mrs. H.)

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

Factors

(continued on next page)

983

984

Table 2 (continued ) Description

Identified in Interview(s) With

Quote

d. Family’s willingness

Initiating discussions about end of life and the pace at which decisions are made may be too slow according to the professional caregivers. Some families are open for discussions about the end of life and it is not difficult to make decisions. However, other families have to get used to the NH. They absolutely did not want discussions about end of life immediately. Difficulty talking about the end of life does not mean that the discussions about end of life should be postponed; discussions can be started and the direction of care can be formulated step by step. Not everything needs to be discussed in a single session.

Family Physician Nurse

e. Family involvement

Physicians and nurses said that involved family facilitates the scheduling of discussions about end-of-life care. However, it could also make it harder to agree on a stronger focus on comfort, because it may then be very emotional to realize that death is approaching.

Physician Nurse

f. Continuity of communication

Trust and open communication between family and caregivers emerged in the interviews with elderly care physicians and nurses as positive factors to give physicians an opening to start discussions about end-of-life care. Physicians said that having the same communication partner is important for the family. This could be the physician, but also the nurse could play a role. The interviews showed that if the communication partner of the family informs family about the possibilities and impossibilities of end-of-life care, it is easier for family to decide on end-of-life care. In addition, by informing family through the same person, he/she is familiar with what the family already knows and how much information the family wants.

Physician Nurse

g. Consensus with or within family

In interviews with all types of respondents, consensus was a theme. The interviews revealed that consensus within families and about the wishes of the patient facilitated ACP. Interviewees indicated there was not always consensus about the direction of care or which treatment decision should be taken.

Family Physician Nurse

h. General NH policy

The interviews revealed that the general policy facilitates discussions. Physicians and nurses of several NHs said it is standard practice to discuss the care policy within 6 weeks after admission and family also indicated that there were standard discussion moments. In addition, family and physicians of several NHs said that an information booklet with the general care policy was provided at admission, and that an admission policy form had to be completed on the day of or shortly after admission. The vision of the NH influences the type of care goals and decisions.

Family Physician Nurse

“I mean I wasn’t blind to the situation. And I also constantly thought about what would be the best option in view of the situation and her own ideas. And then a few conversations are enough.” (Family about Mrs. E.) “This wasn’t new to his wife and children. They had a broader perspective, so to speak. We agreed very quickly that this was not a person who should have futile extension of life” (Physician about Mr. T.) “We started focusing more on the palliative aspect very quickly. It could not go fast enough for us. Still: everybody has to move in the same direction of course, and we had a long distance to cover, so we had no choice but to go step by step until we came to the actual decision to discontinue the life-prolonging dialysis treatment.” (Physician about Mr. K.) “The family were actually always willing to come and they were there regularly so it was very easy to . with them good timing was easy. I mean there are situations where at a certain point we think ‘now is the time you would want to tell the family that .’ Some relatives are more difficult to reach or they are on holiday or live far away.” (Physician about Mrs. M.) “A very solid marriage, even now . people who so obviously love each other, despite the level of deterioration, painfully slow deterioration. On the one hand this makes it right to discuss things and easy to discuss things. And on the other hand you also see how hard this is on the family. So there are difficult aspects to it, to such tremendous involvement, and on the other hand also really good aspects.” (Physician about Mrs. M.) “I already knew the family from their home situation . My presence probably removed some barriers. I was there a lot, the same person all the time - they appreciated that.” (Physician about Mrs. L.) “I am much closer to the people than the doctor is and so I alert the physician: I observe a declining line of life, I notice a change in the patient, maybe we should talk to the family, otherwise it will be such an unexpected blow later . and you can also assess the family’s position in this, how much support they will need.” (Nurse about Mrs. M.) “In many cases, it is quite doable to always stay in touch with the family, but all those changes of physician are really unpleasant for the family. A constant factor is experienced as very positive. Especially in the case of this type of policy: this is often quite a process you go through. To them it is important to see the same face all the time. And not to have to explain everything again and again. But sometimes you are faced with situations you can’t always influence, of course.” (Physician about Mrs. L.) “Had this been my mother, I would have done it sooner, but it was her mother. I think the brother also wanted to shift the focus to comfort at an earlier stage, and not to extend life anymore . Because on the physically good days, so to speak, I also saw a very sad, vulnerable woman who enjoyed very few good things at that stage. (Physician about Mrs. V.) “Of course it helps that all children were in agreement. And in agreement with the father as well. There would always be 2 or 3 of them at these discussions. They never came alone and they informed each other.” (Nurse about Mr. S.) “We do bring up the policy during the first care planning discussion after arrival, within 6 weeks., It is nearly always mentioned during the first discussion because it is a good thing that the family starts thinking about it . Some people are totally clear about it and you can talk about it immediately. And it makes the others start thinking about it. (Physician about Mrs. O.) “We have a booklet with the general medical policy . we depart from the principle that your time is in God’s hand: that means that you should not shorten a life, but also not extend a dying process. In this context you start assessing: are any treatments needed and possible here that can simply continue to support life? Or is this by now an irreversible dying process, of which this infection is a part and so we are not going to treat it.” (Physician about Ms. M.)

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

Factors

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

Always discussions about CPR and hospitalization CPR and hospitalization of patients were almost always discussed shortly after admission, according to almost all interviewees when asked about the timing and content of discussions and decisions. In general, the result of the discussions was a do-not-resuscitate (DNR) and a do-not-hospitalize (DNH) order. Decisions about CPR and hospitalization are necessary, according to all interviewees, as there is no time for discussions at the critical moment. In addition, many physicians and nurses referred to a general DNR policy in their NH. The general institutional policy for patients with dementia is not to resuscitate and to hospitalize only with selected diagnoses: “In principle, this is a ‘do-not-resuscitate’ policy for the patients suffering from dementia. Hospitalization basically only takes place in case of fractures when it has added value.” (Nurse about Mrs. M.) and “A donot-resuscitate policy has been agreed upon. This is discussed for every patient.” (Elderly care physician about Mr. B.). Multiple factors related to the timing and content of ACP discussions and decisions about other treatments The timing and content of discussions and decisions on other specific treatments, such as antibiotics, rehydration and tube feeding, or specific conditions, such as pneumonia, varied. The interviews revealed the following related factors (Table 2, aeh): general strategies that guided physicians in initiating ACP discussions, patient’s condition, wishes expressed by patients or family, family’s willingness, family involvement, continuity of communication, consensus with or within family, and general NH policy. The factors were identified in interviews with all 3 interviewee types, except for family involvement, which we identified only from the interviews with the physicians and nurses. Consensus was mentioned in the context of the timing and the content of the treatment decisions. Continuity of communication was mentioned in relation to the initiation of careplanning discussions. The other factors were mentioned in the context of the initiation of the discussions as well as the timing of the treatment decisions and the content of the treatment decisions. In the following, we describe the physicians’ strategy (Table 2, a) in detail, as nearly all interviewees indicated that the physician took the initiative to start discussions about the end of life. Physicians’ strategies were therefore influential and the strongest factor for the initiation of discussion about other treatment decisions. We found that 1 of 2 different general strategies guided the physicians in their planning of end-of-life care: (1) wait for a reason to initiate discussions, such as a change in health condition or families bringing it up themselves, and (2) take initiative to discuss possible treatments (actively, including describing scenarios). The first strategy was to discuss and arrive at decisions about CPR and hospitalization of the patient, and subsequently wait for a change in the patient’s condition or situation to trigger further planning of end-of-life care. The second strategy was more active. In addition to decisions about CPR and hospitalization, physicians initiated discussions about other treatment decisions. However, within this second strategy, the level of detail in which other treatment decisions were discussed and made by physicians differed. For example, some used scenarios to arrive at a number of advance decisions, whereas others used scenarios to make family agree that life prolongation is not preferred, but would specify treatments only when a specific health problem develops. The reasons underlying the 2 strategies are presented in Box 1. Family indicated that, in retrospect, the appropriate moment to discuss treatment decisions was mostly the moment that was consistent with the strategy followed by the physician. With the first strategy, the patient’s condition (Table 2, b) triggers discussions. However, a decline in condition is also a reason to evaluate decisions for physicians who actively discuss treatment decisions in advance (second strategy). In addition to patient wishes

985

Box 1. Reasons Underlying Physicians’ Strategies That Guided Physicians in Their Initiative to Discuss Other Treatments and Conditions in Addition to CPR and Hospitalization Reasons for strategy 1: Wait for a reason to initiate discussions (such as a change in health condition) - Feeling there is enough time to discuss other issues at the critical moment; - Feeling that discussing scenarios of possible future events is pointless, these events will not necessarily happen to this patient; - Belief that it is hard for families to think in terms of scenarios, which is therefore not helpful; - Experience that not all families are open to ACP discussions and the physicians avoided it because of uncertainty about the reaction of families, or even reluctance to go into this matter. Reasons for strategy 2: Take initiative to discuss possible treatments - Benefit of having detailed instructions as to what to do in acute situations, in particular relevant for the “physician on duty” in acute situations; - To allow time for discussions after admission and before a patient gets sick, whereas time in acute situations may be limited; - Wish to inform family on time; - Allow family to prepare for and think about what they want at the end of life of their family member so they will not be surprised in an acute situation.

(Table 2, c), we identified a variety of factors related to communication of practitioners with and communication within families (family wishes, willingness, involvement, continuity of communication and consensus with or within families) (Table 2, ceg). These factors are also interrelated and also affect timing. For example, a patient’s clear wish not to pursue pointless extension of life, or the opposite, a wish to live as long as possible, facilitates treatment decisions, whereas without consensus about the content of treatment decisions within family or between family and physician, the making of a decision can be postponed. The last factor, general NH policy (Table 2, h), is an important determinant of the content of decisions about CPR and hospitalization, but also facilitates discussions about other treatments. Several NHs as a default discussed the care policy within 6 weeks after admission, and some interviewees also indicated that an information booklet with the general policy was provided on admission.

The Experiences With ACP of Family, Physician, and Nurse of 1 Patient Boxes 2, 3, and 4 present 3 information-rich example patient cases to illustrate the different and overlapping perceptions of ACP of different actors regarding 1 patient. In case 1, the physician waited for a trigger to discuss treatment decisions other than CPR and hospitalization, whereas in patient cases 2 and 3, the physician also discussed other treatment decisions shortly after admission. Case 2 shows that specific decisions were made in advance about lifethreatening and nonelife-threatening conditions. In addition, decisions were made about artificial feeding. In case 3, physician and family agreed from the beginning that the care goal should not be to prolong life, but the physician prefers to elaborate specific treatments in the case of concrete calamities. The 3 cases point to the different strategies of planning end-of-life care, but in all 3 cases, a change in

986

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

Box 2. Case 1: Mrs. D., Aged 82 at Death, Length of Stay 1.7 Years, Perspective of Family and Physician*

Box 3. Case 2: Mrs. H., Aged 95 At Death, Length of Stay 2 Years, Perspective of Family, Physician, and Nurse*

Mrs. D. was admitted to the NH in the beginning of 2008 by authorization of the court. The spouse indicated that the admission of his wife to the NH was hard for him, as he had to admit that he could not cope with it at home.

Mrs. H. was transferred from a residential home to an NH in the spring of 2008. A year later, the physician we interviewed started work at the NH. There was no face-to-face transfer of patient information between the physicians, and the patient file did not include information about care goals and treatment decisions. The physician indicated that this was the case for many patients, and he changed this for his patients. In his opinion, it is very important that the family has time to think about a decision and decide whether to treat or not treat. His experience is that if you wait until the last moment, family often chooses the life-prolonging treatment. Afterward, family, in his experience, often regretted their decision. He usually received positive reactions from family on advance discussions about care goals and treatments. The nurse underlines the importance of making decisions in advance and reporting these decisions in the patient file. In case of changes in the patient’s condition during the night, she and the physician on duty know which decisions had been made and how family feels about these decisions.

The physician started to work in this NH in mid-2008. The physician decided together with the family that the patient will not be hospitalized, but if necessary, they will discuss what else to do. The family indicated that the patient did not express wishes about the end of life. In June 2009, the previous decisions were evaluated with the patient’s spouse. In retrospect, the physician indicated that this evaluation was quite late. In this conversation, the spouse and physician agreed that further discussions about hospitalization were not needed anymore. Hospitalization was excluded and they agreed to a “proportionally active policy.” The patient had compulsive behavior and parapemphigus. The physician indicated that these conditions were not lifethreatening, so there was no need for a palliative treatment goal. The physician found it hard to communicate with the first contact person, the spouse, because the spouse could not oversee the situation. In September 2009, a further conversation was held to discuss the treatment of a life-threatening condition; ileus. This time the daughter and son of the patient were also present. The children supported their father. In this conversation, the policy was converted to a palliative policy. The physician said that in his experience there is still time to discuss the situation that arises at that moment with the family and a palliative treatment goal can be implemented at relatively short notice. The physician’s opinion is that family also suffers by discussing scenarios in advance. To clearly explain these scenarios, concrete examples are needed. In his experience, family often indicated that they do not want to decide yet; they prefer to wait until the situation actually develops. The spouse felt he had had little contact with the physician and he would have liked to have more. When the health condition of the patient declined, the situation was hard for him to accept. *In bold are the care decisions and the factors that determine the timing of care-planning discussions and the content of treatment decisions.

condition triggered having discussions and making or evaluating treatment decisions. Further, in all 3 cases, most contacts were with the nurse. However, the families varied in their experience of contact with the physician. The family member in case 1 would have liked more contact with the physician, whereas the family in case 2 felt they could contact the physician at any time. This feeling of being able to contact the physician is important for family. In case 3, the family also said they appreciated the openness of communication, even if it did not concern good news. Discussion This qualitative interview study aimed to describe experiences of family, physicians, and nurses on the process of ACP in Dutch NHs and to explore factors related to ACP. The study showed that it was standard practice to make advance decisions about CPR and hospitalization soon after admission. However, the timing and content of

For this patient, the first multidisciplinary consultation took place with the physician, responsible nurse, and the family in October 2009. Care goals and possible treatment decisions were discussed. The physician and nurse said this family knew what they wanted and the following decisions were made: In case of a life-threatening condition, the patient would be treated in the NH. Hospitalization will not be considered. If active treatment would not result in improvement, symptom relief would be the next step. In nonelifethreatening conditions, hospitalization would be considered in consultation with the family if improvement of comfort was to be expected. Fluids and food are offered, but refusal was accepted; no artificial fluids and nutrition. The family said they thought it was good to have ACP discussions. Their mother already indicated that she does not want to be resuscitated. Family said they respected and understood her wish, as she was almost 96 years old. After the first multidisciplinary consultation, nurses had the most frequent contacts with family. The physician said this was on purpose, because they are closely involved. If the nurses indicate that family has specific questions, the physician takes the initiative to contact the family. The decisions made in the multidisciplinary consultation were not changed later during admission, as there was no reason for this. Family confirmed they had little contact with the physician, but they had contact with the nurses and felt they could always contact the physician when needed. They said the decisions made in the multidisciplinary consultation were evaluated once and reaffirmed. During her admission, the condition of the patient was stable until May 2010. In May 2010, the patient became nauseated and had abdominal discomfort. A number of laboratory tests showed that the patient had a bowel infection. The fluid and food records, completed by the nurses, showed a poor intake of fluid. The nurse indicated that based on earlier decisions, the physician decided not to use artificial feeding. Over the next days, her discomfort increased, oral medication was discontinued, and morphine was started. Physician, nurse, and family indicated that, within a week after the patient’s condition started to decline, the patient quietly died in her sleep, in the presence of her son. They all have a positive feeling about the delivered care. *In bold are the care decisions and the factors that determine the timing of care-planning discussions and content of treatment decisions.

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

Box 4. Case 3: Mr. T., Aged 86 At Death, Length of Stay 2.4 Years, Perspectives of Family, Physician, and Nurse* Mr. T. was admitted to the NH in the autumn of 2007. The physician indicated that the patient accepted the dementia and his other disabilities. In the beginning, the patient walked small distances, but over time his condition declined. In the final 6 months, he understood less and less. He could not come to the living room by himself anymore and eating and drinking also became difficult. From the beginning, care goals and treatment decisions were discussed with his wife and children. The physician and nurse said they accepted the situation and soon after admission all agreed that futile life-prolongation was not desired. Physician and family said the patient agreed with his wife that he did not want to be kept alive at all costs. A DNR decision was made and they agreed to discuss what to do when necessary, but no extensive interventions, such as tube feeding, were allowed. The physician said she always discusses the care policy in general and follows up in more detail in case of concrete calamities. In retrospect, the physician was satisfied with the family communication and with the timing of discussions. The daughter said she and her mother had 4 or 5 conversations with the physician about the care policy: soon after admission, when they stopped the Alzheimer medication, after death, and 1 or 2 other conversations in-between. In addition, they had a lot of contact with the responsible nurse in the joint living room. During admission, the patient had several pneumonias and was given antibiotics to cure this infection. The condition of the patient gradually declined, and approximately a half year before death, the Alzheimer medication was stopped, but the patient still died somewhat unexpectedly in his sleep. Family was very satisfied with the communication and provided care. They appreciated the honesty and openness, even if it did not concern good news. *In bold are the care decisions and the factors that determine the timing of care-planning discussions and content of treatment decisions.

discussions on other treatment decisions and on general goals of care depended on a variety of factors, and the terminology used to describe care goals varied between facilities. A variety of factors emerged from the interviews as being related to ACP: physicians’ strategy, the patient’s condition, wishes expressed by the patient or family, family’s willingness, family involvement, continuity of communication, consensus with or within family, and general NH policy. With regard to the physicians’ strategy, the interviews of all 3 types of interviewees showed that physicians generally take the initiative to start ACP discussions, and we found that the physician used 1 of 2 underlying strategies: (1) wait for a reason to initiate discussions and (2) take initiative to discuss possible treatments.

987

professional caregivers may not initiate discussions based on the reasoning that some scenarios are unforeseen and unplanned.30,32 In contrast to a Dutch review of barriers and facilitators of palliative care discussions among GPs, we did not find resistance to discussing a poor prognosis.33 It may be more obvious that many patients in NHs have a poor prognosis, and, therefore, it may be a less sensitive issue to talk about. However, it could also be explained by the research design. All interviews were after death, so all respondents looked back on the patient’s stay in the NH and did not have to make difficult decisions anymore. The condition of the patient with dementia was found to be one of the factors influencing the timing and content of ACP discussions in our study as well as other Dutch and US studies.26,27,29,34,35 However, initiating discussions about treatment following a change in the patient’s condition is more responsive than proactive, whereas ACP and palliative care in general aim to be proactive.1 Qualitative research in patients with incurable cancer identified 2 comparable strategies of physicians in preparing patients for the last phase of their disease trajectory: one more proactive and the other more passive.36 In addition, research among inpatient populations with cancer and heart failure has shown that addressing future issues in the process of ACP is limited mostly because the future condition is uncertain.37,38 The Right Timing of Initiating ACP The “right” timing for physicians to take the initiative in planning care for patients with dementia in the last stage of life is unclear. Several studies have suggested that ACP should be started as early as possible so the patient is allowed time to consider or accept the diagnosis yet still has the capacity to consider any preferences with regard to the future.1,32,39 Several authors have stated that ACP just needs time, which underlines the need to start early.13,29,35 The diagnosis of dementia is often made when a GP is still coordinating the patient’s care. Our study suggests that GPs transfer little to no information about patient wishes and advance treatment decisions when a patient is admitted to an NH. GPs also may have difficulties defining the right moment to initiate ACP for patients with dementia.40 Access to specific health care services or admission to a health care facility for long-term care, such as an NH, may trigger the initiating or reviewing of ACP discussions.32 Also, regular and planned meetings among family, professional caregivers, and, if possible, the patient, may help to initiate discussions about end-of-life care and help patients and families to visualize possible disease trajectories and prepare for future care.27,29,35 Time should be made available for such discussions, as physicians’ lack of time is found to be a barrier to both initiating discussions and decision-making.25,33,41 In addition, many of the factors related to the timing of care-planning discussions identified in this study may be modifiable factors that may be influenced by professional caregivers anticipating and providing information. For example, continuity of communication helps prepare family to be open to ACP discussions. Professional caregivers should be aware of these factors to choose the “right” moment to take the initiative in planning end-of-life care.

Timing and Content of ACP in Previous Studies

Strengths and Limitations

Our findings with regard to the finding that professionals often do initiate ACP discussions in the qualitative interview study are consistent with the conclusions drawn in our recently published review about initiation of ACP in dementia.30 Several studies have shown that encouragement by professionals of patients and families to talk about the end of life is essential, and professionals should guide the patient and family in discussions and making decisions.25,28,29 Studies in the United States also found that

A strength of this study is that, as recommended by Stewart et al,13 we included more perspectives in the interviews. As a result, we uncovered more factors related to ACP and we could compare experiences within individual patients. We also identified a factor in interviews with professional caregivers that was not identified in family interviews: family involvement. However, we were unable to recruit families participating in the DEOLD study with highly variable perceptions of whether the planning of care

988

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989

was too early or too late, as family was usually satisfied with the timing according to the questionnaires completed during the patient’s stay in the NH. We conducted 8 face-to-face interviews. These inspired the exploration of new issues, which were subsequently examined in telephone interviews that then lasted longer than the earlier telephone interviews. However, the face-to-face interviews yielded no information that could not also be collected in telephone interviews. Face-to-face interviews are more common than telephone interviews, but there is little evidence that missing the visual cues provides bias.42 Telephone interviews also may provide advantages when discussing sensitive issues. The context of the DEOLD study in which families had been involved for some time may also have facilitated establishing rapport during the telephone interviews. Further, whether face-to-face or by telephone, access to the patient’s chart during the interviews with professional caregivers facilitated more detailed discussions. Further Research and Cross-National Issues Participatory observational research may provide more in-depth knowledge about the effects of physicians’ proactive or responsive strategies to ACP. Quantitative research of the identified factors and cross-national comparison may increase our understanding of how important the factors are in different health care settings and different countries. Our study was conducted in a country with a high presence of specialized physicians in NHs compared with other countries. 43 Physicians who are less present are possibly less likely to wait for a reason to discuss treatment decisions and are less inclined to follow the condition of the patient. On the other hand, less presence implies physicians have fewer opportunities to develop a relationship of trust with the family, and may therefore feel less certain of family preferences, which may result in physicians opting for more aggressive treatment.43 We recommend further qualitative and quantitative research in countries with other models of care. In the future, more patients with dementia will continue to live at home longer and ACP will become increasingly relevant for GPs. Further research on a combined model of ACP that accommodates transitions to other settings is recommended. Additional research is needed to examine what is required for ACP and to develop tools to help physicians achieve optimal ACP across different settings. Conclusion This study highlights that ACP in dementia appears to be influenced by multiple factors, but 2 physicians’ strategies shape the context and other factors, such as the patient’s condition and family factors, potentially modify the timing and content of ACP. Although the most acute decisions that may have an immediate effect on survival (hospitalization and resuscitations) are covered in advance, a proactive as well as a more responsive “wait-and-see” style is seen with regard to palliative care principles. ACP also requires eliciting preferences, discussion of care goals, and discussion of other possible treatments, also in the Dutch NH setting with its high physician presence. Research is needed to increase understanding of whether and how the physicians’ strategies affect care processes and outcomes. Acknowledgments We thank all the family caregivers, physicians, and nurses who contributed to this study. We thank the Netherlands Organisation for Scientific Research (NWO, The Hague; Veni 916.66.073; Vidi 917.11.339), ZonMw, The Netherlands Organisation for Health

Research and Development (Palliative Care in the Terminal Phase program, grant number 1151.0001), and the Department of General Practice and Elderly Care Medicine, and the Department of Public and Occupational Health of the EMGO Institute for Health and Care Research, VU University Medical Center, Amsterdam, for funding this research project. References 1. van der Steen JT, Radbruch L, Herthogh CM, et al. White paper defining optimal palliative care in older people with dementia: A Delphi study and recommendations from the European Association for Palliative Care. Palliat Med 2014;28:197e209. 2. van der Steen JT, Helton MR, Ribbe MW. Prognosis is important in decision making in Dutch nursing home patients with dementia and pneumonia. Int J Geriatr Psychiatry 2009;24:933e936. 3. Sepúlveda C, Marlin A, Yoshida T, et al. Palliative Care: The World Health Organization’s Global Perspective. J Pain Symptom Manage 2002;24:91e96. 4. Gove D, Sparr S, Dos Santos Bernardo AM, et al. Recommendations on end-oflife care for people with dementia. J Nutr Health Aging 2010;14:136e139. 5. Hertogh CMPM. Advance care planning and the relevance of a palliative care approach in dementia. Age Ageing 2006;35:553e555. 6. Brinkman-Stoppelenburg A, Rietjens JA, van der Heide A. The effects of advance care planning on end-of-life care: A systematic review. Palliat Med 2014;28: 1000e1025. 7. Dening Harrison K, Jones L, Sampson EL. Advance care planning for people with dementia: A review. Int Psychogeriatr 2011;23:1535e1551. 8. Robinson L, Dickinson C, Rousseau N, et al. A systematic review of the effectiveness of advance care planning interventions for people with cognitive impairment and dementia. Age Aging 2012;41:263e269. 9. Teno JM, Lindemans Nelson H, Lynn J. Advance care planning. Priorities for ethical and empirical research. Hastings Cent Rep 1994;24:S32eS36. 10. Gillick MR. Advance care planning. N Engl J Med 2004;350:7e8. 11. Sudore RL, Fried TR. Redefining the ‘planning’ in advance care planning: Preparing for end-of-life decision making. Ann Intern Med 2010;153: 256e261. 12. Henry C, Seymour J. Advance Care Planning: A Guide for Health and Social Care Staff. London: Department of Health; 2007. 13. Stewart F, Goddard C, Schiff R, et al. Advanced care planning in care homes for older people: A qualitative study of the views of care staff and families. Age Ageing 2011;40:330e335. 14. van der Steen JT, Heymans MW, Steyerberg EW, et al. The difficulty of predicting mortality in nursing home residents. Eur Geriatr Med 2011;2: 79e81. 15. Robinson L, Dickinson C, Bamford C, et al. A qualitative study: Professionals’ experiences of advance care planning in dementia and palliative care, ‘a good idea in theory but.’. Palliat Med 2013;27:401e408. 16. Torke AM, Sachs GA, Helft PR, et al. Timing of do-not-resuscitate orders for hospitalized older adults who require a surrogate decision-maker. J Am Geriatr Soc 2011;59:1326e1331. 17. Schulz R, Boerner K, Klinger J, et al. Preparedness for death and adjustment to bereavement among caregivers of recently placed nursing home residents. J Palliat Med 2015;18:127e133. 18. van der Steen JT, Onwuteaka-Philipsen BD, Knol DL, et al. Caregivers’ understanding of dementia predicts patients’ comfort at death: A prospective observational study. BMC Med 2013;11:105. 19. Vandervoort A, Houttekier D, Van den Block L, et al. Advance care planning and physician orders in nursing home residents with dementia: A nationwide retrospective study among professional caregivers and relatives. J Pain Symptom Manage 2014;47:245e256. 20. Van der Steen JT, Ribbe MW, Deliens L, et al. Retrospective and prospective data collection compared in the Dutch End Of Life in Dementia (DEOLD) study. Alzheimer Dis Assoc Disord 2014;28:88e94. 21. Van der Steen JT, Hertogh CMPM, de Graas T, et al. Translation and crosscultural adaptation of a family booklet on comfort care in dementia: Sensitive topics revised before implementation. J Med Ethics 2013;39: 104e109. 22. Lynn J, Adamson DM. Living well at the end of life. Adapting health care to serious chronic illness in old age. Santa Monica, CA: Rand Health; 2003. 23. Teno JM, Lynn J. Putting advance-care planning into action. J Clin Ethics 1996; 7:205e213. 24. Gillick MR. A broader role for advance medical planning. Ann Intern Med 1995; 123:621e624. 25. Gessert CE, Forbes S, Bern-Klug M. Planning end-of-life care for patients with dementia: Roles of families and health professionals. Omega (Westport) 2000; 42:273e291. 26. Hirschman KB, Kapo JM, Karlawish JH. Identifying the factors that facilitate or hinder advance planning by persons with dementia. Alzheimer Dis Assoc Disord 2008;22:293e298. 27. Forbes S, Bern-Klug M, Gessert C. End-of-life decision making for nursing home residents with dementia. J Nurs Scholarsh 2000;32:251e258.

M.C. van Soest-Poortvliet et al. / JAMDA 16 (2015) 979e989 28. Livingston G, Leavey G, Manela M, et al. Making decisions for people with dementia who lack capacity: Qualitative study of family carers in UK. BMJ 2010;341:c4184. 29. Hurley AC, Volicer L, Rempusheski VF, et al. Reaching consensus: The process of recommending treatment decisions for Alzheimer’s patients. ANS Adv Nurs Sci 1995;18:33e43. 30. van der Steen JT, van Soest-Poortvliet MC, Hallie-Heierman M, et al. Factors associated with initiation of advance care planning in dementia: A systematic review. J Alzheimers Dis 2014;40:743e757. 31. Nederlandse Vereniging van Verpleeghuisartsen (NVVA), Sting, Verpleegkundigen & Verzorgenden Nederland (V&VN): [Concepts and accuracy demands as regards decision making at the end of life]. Utrecht: NVVA, Sting, V&VN; 2006. 32. Mullick MA, Martin J, Sallnow L. An introduction to advance care planning in practice. BMJ 2013;347:f6064. 33. Slort W, Schweitzer BP, Blankenstein AH, et al. Perceived barriers and facilitators for general practitioner-patient communication in palliative care: A systematic review. Palliat Med 2011;25:613e629. 34. Pasman HRW, The BAM, Onwuteaka-Philipsen BD, et al. Participants in the decision making on artificial nutrition and hydration to demented nursing home patients: A qualitative study. J Aging Studies 2004;18:321e335. 35. The BAM, Pasman HRW, Onwuteaka-Philipsen BD, et al. Withholding the artificial administration of fluids and food from elderly patients with dementia: Ethnographic study. BMJ 2002;325:1326.

989

36. Pfeil TA, Laryionava K, Reiter-Theil S, et al. What keeps oncologists from addressing palliative care early on with incurable cancer patients? An active stance seems key. Oncologist 2015;20:56e61. 37. Low J, Pattenden J, Candy B, et al. Palliative care in advanced heart failure: An international review of the perspectives of recipients and health professionals on care provision. J Card Fail 2011;17:231e252. 38. Barnes KA, Barlow CA, Harrington J, et al. Advance care planning discussions in advanced cancer: Analysis of dialogues between patients and care planning mediators. Palliat Support Care 2011;9:73e79. 39. Poppe M, Burleigh S, Banerjee S. Qualitative evaluation of advanced care planning in early dementia (ACP-ED). PLoS One 2013;8:e60412. 40. De Vleminck A, Houttekier D, Pardon K, et al. Barriers and facilitators for general practitioners to engage in advance care planning: A systematic review. Scand J Prim Health Care 2013;31:215e226. 41. Cavalieri TA, Latif W, Ciesielski J, et al. How physicians approach advance care planning in patients with mild to moderate Alzheimer’s disease. J Am Osteopath Assoc 2002;102:541e544. 42. Novick G. Is there a bias against telephone interviews in qualitative research? Res Nurs Health 2008;31:391e398. 43. Helton MR, van der Steen JT, Daaleman TP, et al. A crosscultural study of physician treatment decisions for demented nursing home patients who develop pneumonia. Ann Fam Med 2006;4:221e227.

Appendix 1. Themes That Emerged From the Qualitative Interviews About Advance Care Planning in NH Patients With Dementia Themes

Condition/situation patient (in general or in relation to discussions/decisions/care goals) Consensus/(dis)agree Contact/communication/knowing (in general or in relation to the discussion/decision) Contact/communication/family (discontinuity in care [by physician]) Content policy/discussion/decision/care goal (in general or for a specific patient) Do more than asked of them Definition and use of terminology (Former) treatments General way of working in the NH (content of policy or opinion physician on the general policy or moment/frequency/embedding of policy discussions) Guidelines Importance to make decisions/establish care goals Initiative Moment(s) of policy discussions/decisions for this patient Perception of dementia (Quality of) physical environment Quality of care/satisfaction Willingness (Un)certainty Who decides/influence/law/wish

Identified in Interviews With

Findings Relevant to Purpose*

Family

Physician

Nurse

x x x x x

x x x x x

x x x

x x x

x x x x x x x x x

1 and 2 1 and 1 and 1 and 1 and 1 NA 1 and

x x x

x x x x x

x x x x x x x x x x

1 2 2 1 and 2 1 and 2 NA NA 2 2 1 and 2

x x x x x

x x x x

2 2 2 2 2

2

NA, not applicable. *Purpose 1 ¼ Describe the process of ACP in nursing home patients with dementia; Purpose 2 ¼ Determine factors related to ACP in nursing home patients with dementia.