Introduction

In spite of evidence that advance care planning (ACP) conversations can increase both the quality of care and overall satisfaction of patients with advanced cancer [1,2,3], the uptake of ACP in clinical practice has remained low [4,5,6,7]. Both healthcare professionals (HCP) and patients have indicated that readiness constitutes a significant barrier to initiating these ACP conversations [8,9,10,11,12,13].

A primary goal of ACP is to help ensure that patients’ medical care and treatment aligns with their personal values, goals, and preferences, especially should they become unable to articulate these preferences themselves [14]. For patients with advanced cancer, an ACP conversation can serve as a valuable opportunity to discuss and refine their wishes and preferences before the onset of progressive and functional decline. An ACP conversation generally consists of four phases: preparation, initiation, exploration, and action. During the core phase, exploration, patients are encouraged to share their thoughts about various end-of-life topics [15]; often, they are also invited to discuss topics related to various aspects of their lives including their physical health, psychological, and social wellbeing [16].

Due to the sensitive nature of these topics, both healthcare providers and cancer patients consider readiness for ACP an important factor when deciding whether or not to engage in an ACP conversation. Some studies have defined readiness as being prepared for action [17] or being willing to engage in a discussion about one’s values and wishes with one’s family and HCP [16, 18, 19]. Studies of ACP based on these definitions often consider readiness as a prerequisite for a conversation. These studies focus on a patient’s state-of-mind prior to the start of the ACP conversation [14, 16] and consider readiness to be a predictor of a patient’s willingness to engage in an ACP conversation or ACP-related activites [18,19,20,21], an indicator for HCPs as to when they should initiate an ACP conversation [22], or an essential precondition for a patient to experience an added value of ACP [14, 16, 23]. Yet readiness has also been defined as a process outcome of successful ACP [24], and patients have reported that the ACP conversation itself can have a positive impact on their readiness [25].

Until now, the literature has shed little light on the manifestations of patients’ readiness during an ACP conversation. This means that we have very little practical knowledge about how patients respond to individual topics brought up during an ACP conversation or how their state of readiness might shift or change during the course of the conversation [26]. Therefore, this study aims to gain more insight into how signs of (not) being ready become manifest and the role that readiness plays in advanced cancer patients’ discussions of ACP topics throughout a conversation.

Methods

Research design

We conducted a secondary analysis of ACTION ACP conversation recordings. An inductive qualitative content analysis of ACP conversations was done in order to better understand how patients responded to the topics being addressed, and ultimately to arrive at a better understanding of the manifestations of readiness for ACP in these conversations [27, 28]. The study data were thematically analyzed. This study is embedded in the ACTION trial (ISRCTN63110516), Phase III multicenter cluster-randomized clinical trial designed to evaluate the ACTION Respecting Choices (RC) ACP intervention in six European countries (Belgium, Denmark, Italy, the Netherlands, Slovenia, and the UK) [29]. Patients with advanced lung or colorectal cancer were recruited to participate in the ACTION trial between May 2015 and December 2018 [29].

Sampling and data collection

For this sub-study, we purposively sampled ACTION RC ACP conversations of patients who completed the ACTION RC ACP conversations in both the qualitative and quantitative parts of the ACTION study at one of the participating Dutch intervention sites. A total of 150 patients were invited to participate in the intervention arm of the ACTION study. Sixty-one patients (21 colorectal patients and 40 lung patients) participated in the study (Fig. 1). ACP conversations were eligible for this study when (1) the facilitator involved had already conducted at least three ACTION RC ACP conversations and (2) the ACTION RC ACP conversation was fully completed. The conversations were led by a trained facilitator either at the patient’s home or in the hospital where they were being treated. Facilitators used a structured conversation guide consisting of open-ended questions on ACP topics as well as scripted explanations of key concepts [29]. Table 1 includes a list of the topics discussed and an example of a corresponding question per topic (see Supplementary Material 1 for more information about the intervention). We considered an ACP conversation to be completed when all scripted discussion guide topics had been broached. This could require one or two ACP conversations depending on the patients’ wishes and the presence of a so-called personal representative—a person appointed by the patient to express their preferences should they be unable to do so themselves. An ACTION RC ACP conversation lasted an average of 1 h and 25 min. To increase our understanding of readiness in all its iterations, maximum variation was sought while sampling with respect to patients’ underlying illness, treatment, and facilitators, as is common in qualitative studies [30].

Fig. 1
figure 1

Inclusion, sampling, and data analysis

Full size image
Table 1 Topics ACTION RC ACP conversation
Full size table

Data about the patients’ characteristics were collected from their medical files. The location of the ACP conversation and the number of conversations per patient were extracted from the facilitators’ reports of the ACTION RC ACP conversations. The included conversations were transcribed verbatim by a professional transcription service and checked for reliability (MZ) prior to the data analysis.

Data analysis

The data selection and analysis occurred inductively and iteratively [31] by means of constant comparison [32]. The software program NVivo11 supported the data analysis. The members of the analysis team are experienced qualitative researchers; two of them (MK, MZ) have a background in nursing. From the conversations of 49 participants that met our eligibility criteria, we first sampled four conversations of patients with different diseases and facilitators in order to familiarize ourselves with the material. After reading the transcribed conversations as a whole, three members of the analysis team (MK, MM, MZ) independently reread the four transcripts in order to identify conversation fragments, or units of meaning [28], we thought signaled signs of being ready and of not being ready for ACP. At this phase of the analysis, we tried to focus on manifest rather than latent content, keeping close to the data and working with a low level of abstraction [27, 28]. We then started grouping these fragments into descriptive categories, resulting in our first code tree. Next, we sampled five more conversations varying in patient gender, levels of experience of the facilitators, and stage of illness. MM and MZ independently coded these transcripts. These independent coding results were compared and discussed during weekly meetings until consensus about the codes, provisional categories, and interpretations was reached. MZ then sampled and coded four more transcripts.

After the first stage of coding, we began a process of abstraction and interpretation [27]. We noticed that facing the past, present, or future played a role in patients’ readiness. As a result of these insights, we created a more focused code tree and fine-grained analysis of patients’ state of readiness for all 15 conversations in response to past, present, and future situations. Again, MK, MM, and MZ discussed differences in their interpretation during weekly meetings until consensus was reached. Using our analyses of variations in readiness in the preceding two stages as a guide, we then categorized the topics discussed during the ACTION RC ACP conversations into three levels of difficulty. Finally, MK, MM, and MZ tested for meaning saturation [33] by sampling and coding two new conversations with the most experienced facilitators and using the final list of signs of being ready and signs of not being ready.

Ethical considerations

Ethical approval for the ACTION trial, including the qualitative work package, was obtained from the Research Ethics Committee (REC) of Erasmus MC, University Medical Center Rotterdam (14-560/C). Written informed consent was obtained from all participating patients. Verbal informed consent was obtained and recorded from the relatives present at the ACP conversation. To ensure confidentiality, all transcripts were coded and any identifying information was removed.

Results

Fifteen of the 49 eligible completed ACP conversations were included for analysis (32%). Table 2 includes an overview of the patient and facilitator characteristics.

Table 2 Background characteristics
Full size table

Unraveling patients’ readiness

From the conversations, we identified both signs of not being ready and signs of being ready.

Signs of not being ready

Signs of not being ready were found in all conversations (Table 3). Patients signaled their unreadiness by expressing a reluctance to consider aspects of ACP, by minimizing the seriousness of their illness and condition, and by steering away from talking about the personal consequences of their illness progression. These signs indicated that a patient’s inability or unwillingness to talk about certain ACP topics was not just a reflection of the patient’s state of mind at that moment in the conversation. Many of the signs of not being ready also revealed that a patient was delaying or avoiding having to think about his/her own deterioration of health or death.

Table 3 Signs that a patient is not ready for aspects of ACP
Full size table

Signs of being ready

Conversely, the willingness and ability to discuss an ACP topic or to consider the personal relevance and impact of an ACP topic constituted important indications that a patient was ready for that ACP topic. Table 4 lists the signs of being ready identified. Although patients indicated their readiness to discuss an ACP topic in a variety of ways, each sign of readiness essentially revealed that a patient could face and talk about an aspect of ACP and could link his/her thoughts to future scenarios related to the end of life.

Table 4 Signs that a patient is ready for aspects of ACP
Full size table

Shifts in readiness within the conversation

Our analysis of the transcripts showed that patients could display both signs of not being ready and of being ready for ACP within one conversation and even within one topic. In fact, patients’ state of readiness could shift per sentence. We also noticed that if a patient had difficulty with one topic, this did not imply he/she would also have difficulty discussing the subsequent topic in the script. To illustrate, Table 5 provides a summary of the ACP conversations of two patients. For example, patient 9, who was unwilling or unable to talk about topics such as his diagnosis and potential future complications, and who openly struggled emotionally at multiple points throughout the conversation, could nevertheless clearly and resolutely articulate his preferences regarding resuscitation and his final place of care.

Table 5 Samples of shifts in readiness during the ACTION RC ACP conversation
Full size table

Understanding patients’ readiness

Deeper analysis revealed that the level of readiness was most visible in patients’ willingness and ability to face their future illness trajectory while taking the past and present into account. Patients’ level of readiness also hinged on the ability to imagine and face the personal consequences of their illness trajectory, both in the physical and psycho-social spheres.

Easy and difficult ACP topics

All patients in this study were able to participate in the ACTION RC ACP conversation, but we found a great variation in their willingness and ability to talk about the distinct topics. The topics discussed during an ACTION RC ACP conversation can broadly be categorized into three levels of difficulty (see Table 6), which correspond to the signs of being ready and of not being ready we identified per topic. In line with the insights presented above, the predominantly easy topics were ones that patients could discuss without linking it to their personal situation—such as the concept of ACP—or topics that facilitated a positive view about their life and illness. The most difficult questions were the ones that explicitly challenged patients to link their responses to their own lives, thoughts, and feelings and to imagine themselves in specific future situations.

Table 6 Easy and difficult topics in the ACTION RC ACP conversations
Full size table

The role of reflection and prospection in a patient’s readiness for ACP

The ACTION RC ACP script encouraged patients to reflect upon the past, the present, and the future at certain moments during the conversation. Most patients were ready to say something about the present and were ready to reflect upon the past. But when asked to link past experiences to their present situation or to think about the future—be it possible future complications or preferences regarding end-of-life care—we noticed more diversity in patients’ states of readiness. Many patients delayed or avoided talking about the future or of what could be learned from earlier experiences (see Table 3). And as Table 4 illustrates, those patients who were ready and able to think about the future could consider their own changes in health from the past, the present, and the future position. They demonstrated an informed view of their prognosis and could anticipate future scenarios; some patients could also shift between the past, present, and future spontaneously and independently of the script.

Rational versus experiential perspective-taking

We noticed a further differentiation in the manner in which patients articulated their stance toward an ACP topic: via rational and experiential perspective-taking. Most patients took a rational approach and spoke without discernible emotional distress about past- and present-focused topics such as experiences with illness in their social or familial circle or when describing what “a good life” entailed. Most patients could also rationally describe practical matters related to the future, such as funeral arrangements, financial arrangements, or the eventual reallocation of household tasks. While these future matters pertained to them directly, patients almost always discussed these matters in an abstract or generalizing manner. To illustrate, one patient answered the question regarding the completion of advance directives as follows: “my non-resuscitation wishes and the euthanasia form [living will], are signed and are all here [in a folder] and [also] with my doctor” (Patient 12). This patient shared only the technical side of his end-of-life wishes without giving any impression of what they meant to him personally.

Some patients who were ready to discuss an aspect of ACP could also imagine themselves in various situations or consider the significance of a specific topic for them personally. This experiential perspective was most apparent in patients’ answers to the future-oriented ACP prompts. For example, one patient repeatedly stated during a conversation that he did not want to end up in a vegetative state. When asked by the facilitator to expand upon this statement, the patient responded: “As long as my brain still works I think I can deal with a lot of physical burdens. But for me it’s all about brain function. If I don’t recognize people anymore. I think that would be terrible for the people around me, but also for me. That is what I consider a vegetative state” (Patient 6). This patient’s utterance indicates that he could imagine himself in a future situation of physical deterioration and suffering. Of the patients who were more ready to face and discuss their future, a sub-selection appeared to be able to imagine the course their illness would likely take and how they personally would react to the impeding changes in health.

Synthesis of readiness for ACP

Synthesizing our findings, we arrived at the following description of readiness for ACP. It is necessary to note that most patients in our study were partially ready for ACP, meaning that they could talk about some, but not all ACP topics.

Readiness for ACP is the willingness and ability to engage in a discussion about the progression of one’s illness, one’s current physical and/or mental state, and possible future scenarios related to the end of life; one is also optimally ready for ACP when one can both rationally articulate one’s stance toward end-of-life topics, can articulate one’s corresponding emotions, and can imagine oneself in future situations.

Discussion

Main findings

This study of recordings of scripted and facilitator-led ACP conversations in patients with advanced cancer revealed that patients could display both signs of being ready as well as signs of not being ready when discussing ACP topics. We noticed that signs of not being ready and signs of being ready frequently occurred when patients discussed future-oriented topics related to a deterioration of health and the end of life. Patients who were most ready to talk about an ACP topic were able to envision their own future deterioration of health and to describe what this meant to them personally.

We defined readiness not only as a willingness to engage in a discussion about the progression of one’s illness, one’s current physical and/or mental state, and possible future scenarios related to the end of life, but also as an ability. For example, we identified the following levels in patients’ ability to respond to questions about their future: most patients could rationally articulate their stance toward end-of-life topics, some patients could also articulate their corresponding emotions, and some could even imagine themselves in future situations and reflect on what this meant to them. Patients varied in this ability, and this in turn became manifest in signs of not being ready.

Our study revealed that patients do not have to be ready for all elements of ACP to participate in an ACP conversation. When asked about various ACP topics, patients can respond to questions they feel ready to discuss. Exposure to topics that might trigger signs of not being ready can at least make a patient aware of an end-of-life topic, a first step in the circle of awareness, recognition, acknowledgment, and acceptance. This hypothesis is supported by bereavement theory, showing that “‘adaptive’ coping with loss is a dynamic regulatory process of oscillation between loss and restoration stressors, whereby the grieving individual at times confronts, at other times avoids, the different tasks of grieving.” [34] Although this model was developed for bereavement with loss, we suggest it also fits the situation of coping with anticipated loss due to a progressive illness that necessitates facing the end of life. On the one hand, ACP conversations touch upon loss-oriented stressors; they ask participants to consider how they will address or process various aspects of the loss experience itself. Examples include questions pertaining to life-prolonging treatment and questions related to the anticipated loss of independency due to expected physical deterioration. On the other hand, ACP conversations also address restoration-related stressors. This includes questions about how participants maintain a good life or what they do to add value to lives, which also requires efforts and energy [35]. ACP conversations therefore entail more than identifying and sharing values, goals, and preferences. From the perspective of coping with loss, ACP can also have a therapeutic value as it might add to the patient’s preparedness. This is in line with previous studies that the ACP process itself can have a positive influence upon the patient’s readiness [25]. Just as bereaved persons have to adjust to their new reality in the absence of a beloved person, seriously ill patients also have to adjust to a continuously changing or an anticipated “new reality.” An aspect of this adjustment is preparedness. Studies show that preparedness is supported by discussions that include prognoses and future care [36].

In our study, we also found that patients could answer difficult questions even if they were emotionally difficult. Taking these points into consideration, readiness should not be seen as an unequivocal prerequisite for starting an ACP conversation, but rather as a state of mind that fluctuates throughout an ACP conversation. The fact that ACP conversations trigger grief simultaneously serves as an argument for dosing such grief, and for consciously dosing ACP conversations as well [34]. Patients do need moments of respite from dealing with grief stressors as an integral part of adaptive coping. As such, ACP conversations should be clearly announced and planned to allow the patient and family to prepare for such a conversation.

Strengths and limitations

A strength of this study was that investigator triangulation was applied by including three researchers with different professional backgrounds and expertise in the data analysis team. This lead to in-depth discussions about how to interpret and categorize the signs of being ready and of not being ready. Another strength is that we studied facilitated conversations that were structured by a conversation guide. The facilitators were trained to bring up and to explore all the listed ACP topics. As a consequence, the topics discussed were the same in all conversations, which enabled us to study readiness in relation to a broad range of ACP topics.

However, the fact that we only studied facilitated and structured conversations can also be considered a study limitation. We could not compare these conversations with open-interview ACP conversations conducted by a patient’s physician or nurse, for instance. Another limitation of this study is the patient sampling; the cases we analyzed were predominantly married and were advanced cancer patients receiving palliative care; we had far fewer cases of patients receiving curative cancer treatment. Finally, it should be noted that patients who were willing to participate in the ACTION trial might have self-selected as being receptive to and probably more ready to discuss ACP in general.

What this study adds

Our study has concrete implications for practice. First, while it is important that patients express general readiness by agreeing to participate in an ACP conversation, HCPs should not use the patient’s readiness as the only indicator for whether or not to initiate or postpone an ACP conversation. Readiness can fluctuate and change during the course of the ACP conversation itself. Instead, HCPs should initiate an ACP conversation with the awareness of the patient’s individual needs, signs of being ready and of not being ready, and potential triggers of signs of not being ready. For example, if a patient seems to show a lot of signs of not being ready to discuss certain aspects of ACP, the HCP can adjust the order of the topics or switch between easy and difficult topics. In addition, knowing that patients can alternate in readiness depending on the topic that is being discussed can help HCPs guide the patients through the conversation accordingly. A topic that deserves further exploration is the impact of facilitators’ attitude and communication skills on patients’ readiness to discuss ACP topics.