The place of the relative at the time of the announcement of cancer progression: BABEL – a mixed-methods study

Abstract

Objectives

This study aims to explore the place of the relative in these triadic consultations and how this influences communication.

Methods

A mixed-methods research strategy was used. Triadic consultations for the announcement of cancer progression were recorded and following the 3 participants completed questionnaires comprising mirror-items. Recordings and answers were further investigated in a few semi-structured interviews. Comparison of quantitative responses (questionnaires) used Wilcoxon’s test for matched series. Qualitative analyses (consultations, interviews) used grounded theory. Patients were over 18, followed for cancer in palliative phase, excluding brain tumors and malignant hemopathies, and presented renewed disease progression. Relatives were over 18 and authorized by the patient to participate.

Results

47 consultations (audio-recordings, answers to questionnaires) and 12 interviews conducted separately with 4 triads were collected. Half the relatives, while remaining in the background, nevertheless contributed to the discussion. For patients, the presence of a relative was considered beneficial and for oncologists it facilitated the announcement. However, symptoms perceived as intimate or private appeared difficult to express for some patients, and for relatives, prognosis was a difficult subject to broach. Although their relationship with time and their expectations may differ, patients and relatives found consultations positive. Oncologists appeared to underestimate the patient’s level of understanding (P<0.001) and perceptions of the seriousness of the disease (P=0.009) but not those of relatives. They did not evaluate the relative’s state of health and check what the dyad had retained.

Significance of results

Training via simulation sessions should be adapted to communication involving relatives.

Introduction

Breaking bad news, “any information which adversely and seriously affects an individual’s view of his or her future” is a key moment in the relationship between oncologists and their patients (Buckman 1992). For oncologists, announcing bad news can be a particularly distressing moment, and numerous negative effects are linked to patient and family reactions (Bousquet et al. 2015; Desauw et al. 2009). Quality communication contributes to improved satisfaction, treatment adherence, health results, and patient understanding (Ong et al. 2000; Zachariae et al. 2003), making patients readier to understand disease progression and accept further palliative chemotherapy in parallel with early palliative care sometimes many months before end of life or the end of palliative chemotherapy (National Cancer Institute (NCI) 2014). The relative will remember how the patient’s end-of-life period was managed (Boyle 2019). From oncologists’ viewpoint, family presence is part of “the oncologist’s balancing act,” especially when communicating on transition to palliative care (Bousquet et al. 2015). Skillful communication with patients and families is crucial in providing quality palliative chemotherapy and palliative care, and poor communication is deleterious (Seccareccia et al. 2015). Few studies have explored the place of relatives in this communication (Datta et al. 2017; Korfage et al. 2013; Laidsaar-Powell et al. 2016; Lee et al. 2018; Zamanzadeh et al. 2013).

Alongside, the role of the family varies across cultures (Datta et al. 2017; Laidsaar-Powell et al. 2016; Lee et al. 2018; Zamanzadeh et al. 2013). French medical culture encourages physicians to talk to patients and then to give the family coherent information if the patient agrees (Bousquet et al. 2015). Present recommendations for communicating bad news to patients with cancer (Baile et al. 2000; Gilligan et al. 2018, 2017) very rarely integrate relatives (Datta et al. 2017; Parikh et al. 2017; Tattersall 2018), and tools, like BAD or SPIKES strategies, available to facilitate communication during physician–patient consultations in cancer care (Licqurish et al. 2019) do not always include relatives.

Methods

Aim

To study the place of the relative in triadic consultations bringing bad news and how this influences communication, given the complexity of interactions between physicians, patients, and relatives, a mixed qualitative/quantitative multidisciplinary approach was implemented (Fàbregues et al. 2020) (oncology, public health, human, and social sciences), combining recordings of consultations, self-administered questionnaires, and semi-structured interviews (Lim et al. 2017).

Population

Oncologists from 16 administrative areas in western France, representing approximately 9 million inhabitants, recruited patients accompanied by a relative in a consultation, the main aim of which was to announce cancer progression leading to changes in care strategies. Eligible patients were over 18, followed for cancer in palliative phase excluding brain tumors and malignant hemopathies, and presented renewed disease progression. Any accompanying relative over 18 and authorized by the patient could participate in the study.

Trial regulation

This study was approved by the Comité Consultatif pour le Traitement de l’Information en matière de Recherche dans le domaine de la Santé (CCTIRS) on November 26, 2014 (file n°14 744) and received regulatory authorization from Commission Nationale de l’Informatique et des Libertés (CNIL) on November 9, 2015 (file DR-2015-362). Ethics approval was granted by the Espace de Réflexion Ethique de la Région Centre (ERERC) on March 4, 2015 (project n° 2015-008).

Design – data collection

Recording of the consultation

Consultations were recorded after the patient–relative dyad had received verbal and written information on the study aims from the oncologist and had signed a consent document.

Self-administered questionnaires

Three self-administered questionnaires (patient, relative, and oncologist) were designed via 10 simulation sessions with volunteering senior physicians and professional actors (playing patients and relatives) then validated by linguists and psychologists (Hureaux et al. 2016).

At the end of consultations, oncologists gave patients and relatives questionnaires to complete independently, collecting information on their sociodemographic characteristics (gender, age, educational level, and professional situation) and comprising mirror-items for each version in the form of closed and open questions on their comprehension of the consultation, their expectations, their perceptions of the illness and of the therapeutic decisions. The oncologists also completed questionnaires with the same mirrored themes, enabling comparison of perceptions, and provided their sociodemographic characteristics (gender, age, specialty, and qualification) and clinical data on the tumors (supplementary material).

Interviews

Some oncologist–patient–relative triads were interviewed 2 to 3 weeks after the consultation by an independent psychologist using semi-structured interviews to further investigate the recordings of the consultations and answers to the questionnaires. These interviews allowed the patient, relative, or oncologist to address points not covered in the questionnaires. Because the answers to questions on overall feelings and understanding of the consultation were particularly informative, we did not conduct any further triad interviews. For patient and relative, to facilitate open discussion, these interviews were conducted separately, face-to-face, outside the usual consultation premises. The interviews with the oncologists took place separately in their workplace.

Data analysis

Descriptive statistics were used for respondents’ demographic and medical characteristics. Means and standard deviations for responses to items on a 10-point Likert scale were calculated. Comparative analysis of patient–relative–oncologist responses was performed using the Wilcoxon test for matched pairs, with 2-sided 5% significance level (SAS 9.4).

Consultations and interviews were fully transcribed, preserving participant discourse, and anonymized. Our analysis used grounded theory (Glaser and Strauss 2010) and thematic analysis (Ramos 2015). Using an inductive approach, we developed our theoretical sampling by analyzing our material as the research deployed. The potential of mixed methods was used not only in the data collection phase but also in the data analysis phase.

After qualitative exploratory work to construct the questionnaires (simulation sessions), not described in this paper, we continued fieldwork with the collection of the recordings of consultations (qualitative data) and questionnaires (quantitative data), processed with an equal priority shared between the 2 forms of data, and we supplemented these results with qualitative analysis of some of the interviews. Theory triangulation (Patton 1999) and multidisciplinary interpretation of the data were performed by oncologists (TL, OC, JE, AA, JH, PC, EG, and HB), sociologists (EL), and public-health professionals (II). The researchers read the transcripts independently, re-read them together and discussed them until they reached a joint solution. Then they scheduled the interviews. Strategies for combining data such as integration by connecting, merging, or nesting enabled integrative analysis of the results (Creswell and Plano Clark 2018).

Results

Themes emerging from the analysis were the course of the consultation, the place of the relative, and representations and perceptions of the consultation by the triad.

Characteristics of the triads

During 2016, among the 45 oncologists who agreed to take part, 20 included 62 patients (Fig. 1). The analysis concerned 47 consultations with complete data (3 questionnaires and an audio-recording of the consultation) conducted by 16 physicians aged 33 to 56 (Table 1), and 12 interviews conducted with 4 of the 47 triads (different oncologists and work settings).

Figure 1. Flowchart.

Table 1. Characteristics of investigators (n = 16)

Patients were aged 46 to 87, and 68% were women. The tumor sites varied, time since diagnosis ranged from a few months to 17 years. Relatives were aged 25 to 82, and most were spouses (Table 2, supplementary material).

Table 2. Characteristics of patients and relatives (n = 47)

EVA = visual analog score , WHO = World Health Organization.

The course of the consultation

Consultations (Table 3) mainly took place in the oncologists’ consulting rooms (94%), and in most cases (87%) the objective was to offer the patient a new treatment line. Exchanges focused on the announcement of disease progression, therapeutic alternatives, treatment goals, supportive oncological care, or palliative treatment options, with few digressions resulting from a need to change the subject or to reassure a distressed patient.

Table 3. Characteristics of the consultation (n = 47)

^a Information on a new chemotherapy protocol to be instated; a few words on the recording; recall of the previous treatment.

Oncologists dominated the exchanges (ratio of the number of words to the total number = 63% (39–87%)). They began the consultation by questioning the patient on his/her state of health, symptoms (83%) and/or by explaining examination results (66%). To announce disease progression, they adopted 4 methods: preparation of the dyad, slower pace of exchanges, careful choice of words, and valorization of positive aspects.

The disease is still around the liver and the lymph node, (..). The good news is that there is no new damage in the lungs. All the other organs are spared. (Consultation 6)

To describe the situation, oncologists recalled using words like “disease progression,” “cancer,” “metastases,” “tumor,” and “nodules,” with few references to “discontinuing treatment,” “palliative chemotherapy,” “palliative care,” “life expectancy,” or “prognosis.” This is confirmed by the recordings. They mentioned prognosis in 6 consultations (13%), following questions from the patient or relative, generally toward the end of the consultation.

To refer to palliative care, oncologists mentioned measures associated with it: “home hospitalization,” “respite hospitalization,” “maintaining in the home as long as possible with assistance,”; and for pauses in treatment they attempted to explain that continuing chemotherapy was not beneficial: “they (the treatments) are not doing you much good,” and announced it in the manner of a reasoned decision: “we consider it is not wise to resume treatment.”

According to oncologists, both patients and relatives asked for information on disease status (79% and 60%, respectively) and on the treatment plan (68% for patients and relatives alike), i.e. the objective of the consultation, and only a minority broached the issues of a therapeutic alternative (36% and 34%, respectively) or prognosis (30% and 19%, respectively).

The relative

Their place in the consultation

Half (49%) the relatives favored the patient–physician relationship, remaining relatively in the background although contributing to the discussion, but a quarter (26%) took over the consultation. In contrast, another quarter (23%) intervened very little. Relatives generally did not intervene at the start of the consultation. They only spoke when areas where they considered they could contribute were broached, patient quality-of- life, treatments, and logistics. Husbands were attentive to the quantifiable elements of the disease and wives to the day-to-day management of the illness. Patients’ children mainly focused on their parent’s comfort.

While my wife is undressing, there’s something that puzzles me (..) in her results, and that is the Gamma GT. (Husband, Consultation 12)

So, I have adapted his diet, with more protein.. I try to find the right foods. (Wife, Consultation 43)

That (loss of appetite) was the reason for our visit to her (the mother). To try and give her new tastes, something different. (Son, Consultation 9)

To contribute to the consultation certain relatives called on resources derived from their professional activity (medical, relational, or linguistic skills), mostly approaching the illness from a relatively outside (non-somatic and nonprofessional) point of view or based on their own experience.

He (the patient) says to me “I’d like to see you (in my position)”. But I’ve already been there, when I was in a car accident with my parents, I had broken bones all over and I stayed one/three months in bed! (Wife, Consultation 46)

The patient–oncologist dyad was dominant, but according to oncologists, in 62% of the triadic consultations, both verbal and nonverbal interactions contributed to the way the consultation unfolded.

A presence that is beneficial

According to 93% of the patients, the presence of a relative helped in the consultation (85% positive comments in questionnaires). Relatives were able to detail, correct, or add elements related to the patient’s health, and help patients understand the results.

He helped me ask questions it would not have occurred to me to ask, and his being there reassured me. (Patient, Questionnaire 32)

It’s more than essential for the two of us to be there, to hear everything, and to understand, because sometimes when you get home you find you haven’t heard the same thing. (Wife, Interview 4)

For oncologists, the presence of most relatives was also seen as beneficial, making the announcement easier, avoiding misunderstandings, or enabling better understanding of the illness. One oncologist confirmed that it was more “comfortable” to announce bad news to a patient accompanied by a relative.

There’s nothing worse than announcing (bad news) to a person on their own. You won’t use the same words, you won’t have the resource person alongside who will provide support after the consultation, so it does change the content of the interview a little, and the way it is conducted. (Interview 11)

During the consultations, relatives were allies for patient and oncologist – the oncologist relying on the relative in cases of noncompliance, refusal of treatment, or misunderstanding by the patient, and the patient relying on the relative to back up his/her therapeutic choice. Thus, care management could be influenced by interactions within the triad.

Patient: We talked about it. I don’t know if you remember, about tablets, but I don’t know if… maybe at least try it? (short silence). Try it. But I don’t know. It’s your decision, isn’t it?

Wife: Maybe it would not be as strong as the normal chemo, I don’t know (…) Maybe it could temper things a bit, chemo-tablets… (Consultation 1)

The relatives attending the consultation had acquired knowledge of the disease and favored information-sharing in the family, acting as an interface with other family caregivers.

Oncologist: Would you like me to draw another diagram?

Wife: Yes, so that I can get things clear for my son – for our children. It was he who wanted it mostly. (Consultation 28)

I don’t want to miss consultations because I like to have the doctor’s opinion. (Wife, Interview 16)

A source of complexity

While certain patients wished to be informed in direct manner (“she didn’t beat about the bush,” “since pretenses are not really the order of the day”), others were not ready (“I would have other questions to ask, but I’m not ready yet to hear the answers”). Thus, certain relatives exposed the patients to questions they did not wish to broach.

To deliver information our first job is to listen, you need to listen to the person, and then to understand what the person really wants to hear, in fact it’s the patient who guides us. (Oncologist, Interview 4)

The relationship with time differs between patients and relatives since expectations can differ. Patients envisaged the short term “How many chemo sessions?” The same applies to physicians “I’ll see you again in two weeks”; “Let’s move forward step by step, if you agree.” Relatives had questions on the more distant future: apprehensions about what will happen, life expectancy, how to prepare for the future.

In these triadic consultations, oncologists also expressed the desire to examine patients privately, preserving their intimacy, in a suitable setting. Symptoms perceived as intimate or private appeared difficult to express for some patients in the presence of a relative. For relatives, prognosis was a difficult subject to broach in presence of the patient.

I had a spell when I was having colored stools – not red, or brown, but… well it was… but now it’s really brown, and – well – quite a lot… Anyway… is it losing blood that could have contributed to me feeling weaker? (Patient after her son went out, Consultation 9)

Of the 47 consultations, only 2, on the initiative of the relative, broached the subject of his/her own physical and mental health.

In fact, when you accompany a patient, what I find difficult for the helper is that people always ask about the patient. If I had been able to see the psychologist, it would have been a time just for me. (Wife, Interview 11)

Representations and perceptions of the triad – outcome of the consultation

During the consultations, oncologists maintained a balance, broaching the subject of palliative care while preserving hope, all within a short time-lapse. Both patients and relatives retained words that characterized the illness (“aggravation,” “cancer,” and “progression”), but also words of hope (“not serious,” “confident,” and “hope”).

According to oncologists, they provided complete information in only 17 consultations (36%) and thought they would need to return to the subject in later consultations. Although they broached the subject of prognosis very little, they (rightly) considered that in 41 consultations (79%), both patient and relative had understood that the disease was incurable.

The disease has progressed, and the treatment has had to be altered and adapted, because the cancer has attacked the lumbar vertebrae – it’s distressing”; “We both know that the disease has developed a lot and that the drugs are there just to slow it temporarily. (Patient and relative, Questionnaire 41)

One patient even added: “Because at one stage I could see in your face that I wasn’t going to get over this. I was very upset, doctor. Very upset” (Patient, Consultation 34).

Perceptions of the consultation (Table 4) were approached through the triads’ responses to questionnaires. While patients and relatives had apprehensions of the seriousness of the disease that differed little (P = 0.56), oncologists underestimated patient perceptions of the seriousness of their disease (P = 0.0090), but not those of the relatives (P = 0.47). They also underestimated patients’ understanding of the words they used (P = 0.0005), but not the relatives’ understanding (P = 0.63). Alongside, the oncologist questioned the relative on what he/she thought of the patient’s situation in only 1 consultation, while more patients than relatives (52% vs 39%) reported no difficulty in understanding the oncologists’ words.

Table 4. Cross-representation of patient, relative, and oncologist (n = 47). Significant differences are indicated as bold value

Following the consultation (Table 5), relatives seemed to underestimate patients’ ability to deal with the oncologist’s explanations, considering patients were less serene than they claimed (P < 0.0001). However, all felt trustful after these explanations (P = 0.14).

Table 5. The outcome of the consultation (n = 47). Significant differences are indicated as bold value

^a Seven relatives because of the patient’s presence.

^b For the other 7, the oncologist does not know.

Around twice as many relatives as patients (26% vs 16%) would have liked more information, on disease progression, prognosis, and treatments, and 7 relatives (15%) said they had not been able to ask all the questions they wished during the consultation because the patient was present. Certain relatives (11%) considered that it was not desirable to give the patient full information.

Nevertheless, both patients and relatives appreciated being supported in their plans and welcomed treatment postponement if it did not interfere with disease progression.

Finally, patients and relatives considered that oncologists listened to them. However, it can be noted that at the end of the consultation, no oncologist checked the understanding by the patient or the relative of the consultation by asking them to reword what had been said.

Discussion

The degree of empathy felt by patients and their relatives, and the experience of the consultation were overall very positive. Although none of the oncologists checked what patient and relative had retained from the consultation, they appeared to underestimate the patient’s level of understanding and perception of the seriousness of the disease but not those of relatives. The possible hypotheses, for oncologists, are the impact of the emotional load in the consultations, and the distress generated by what they had to say (Granek et al. 2017, 2016; Paiva et al. 2018). Thus, these supposedly greater difficulties in understanding for the patient than for the relative could be linked to the oncologist’s secret hope that the patient has not fully understood the announcement. They provided information, seeking to protect the patient more than the relative, also protecting themselves from possible guilt at having said too much to the patient, or because the patient had understood too well. It is also possible that, because certain words are impossible to utter, the oncologist’s nonverbal communication was more important for the relative than for the patient, possibly at moments before or after the clinical examination.

During the consultation, oncologists tried to include relatives, who were willing involve: only half of them actually contributed to the consultation, but without taking over from the patient. Very few oncologists broached the physical and psychological health of relatives by asking them directly about their own situation (van Oosterhout et al. 2021). Yet while accompaniers provide support, they also need support (Laidsaar-Powell et al. 2016; Ream et al. 2013). The relative is not just a partner but also a “second patient in distress” (Bréchot 2007).

For the patient, symptoms perceived as private or intimate, and for the relative, prognosis, were more difficult subjects to address in triadic consultation. Relatives did not ask all the questions they had in mind, and would have liked more information. Their demand for information was greater than that of the patients, and some appeared to wish to restrict the patient’s access to full information. The study confirmed that, in this context of treatment discontinuation and orientation toward palliative care, while patients were in a logic of seeking honest information (Datta et al. 2017), they were sometimes unwillingly confronted with questions they were reluctant to raise (Furber et al. 2015; Laidsaar-Powell et al. 2016). Patients and relatives did not always move at the same pace and did not have the same information needs (Saltel et al. 2001). Oncologists should identify the needs and expectations of patients, but also those of relatives, to detect discrepancies within the dyad and avoid forcing the pace of information delivery. They should adapt to changing alliances, ensure that dialogue is maintained within the dyad, and monitor interactions.

The relative provides support for the patient but is also a resource for the physician in helping to maintain compliance with treatment and providing emotional support. Oncologists can feel relieved of the need to provide particularly close emotional support, since it is provided by the relative (Burkhalter and Bromberg 2003). For the oncologist, this enables identical information to be given to both members of the dyad, and provides complementary information on the patient, since the relative is familiar with the patient’s life. Thus, the presence of the relative is often beneficial (Laidsaar-Powell et al. 2016). However, certain dyadic alliances can make exchanges more difficult, such as when relatives take the patient’s side to demand inappropriate treatments or monopolize the conversation (Beisecker and Moore 1994; Laidsaar-Powell et al. 2016; Rosenberg et al. 2017; Saint-Marc 2012). Furthermore, some relatives, contaminated by their own perceptions of the situation, appeared to underestimate the patient’s ability to cope with the oncologist’s explanations. Oncologists should consider these different contexts. Shared decision-making tools could facilitate exchanges (Sloan et al. 2021; van Oosterhout et al. 2021).

This study shows that each consultation is unique, depending on each participant’s manner of being and doing, but this does not exclude a common typology. To announce disease progression, oncologists relied on preparedness for the situation (Bousquet et al. 2015), a slower pace of exchanges (McHenry et al. 2012), careful choice of words, or the valorization of positive aspects (Ménoret 2007; National Cancer Institute (NCI) 2014).

Patients and relatives entered the consultation with different objectives, patients to discuss the present situation, and relatives to identify the best treatment solution (Lee et al. 2018). Relatives generally let the patient–physician dialogue unfold around the symptoms and the illness, and only spoke when treatments were broached. With the changes in sociocultural representations, taboos concern metastatic tumors more than the cancer itself (Fainzang 2013). For the oncologists, the difficulty was more how to position themselves toward prognosis (Singh et al. 2017), or palliative care, although it is particularly relevant to improve patient quality-of-life and reduce medical costs (Epstein and Street 2007). They tended to prefer associated themes, and an earlier study showed that while the words “palliative,” “supportive,” and “hospice” are commonly used interchangeably, they have differing impact (Fishman et al. 2018). The exchanges between oncologists and dyads confirmed that the words of the physician often dominated the exchanges.

This study echoes a recent systematic review that concluded that further research was needed to understand communication with relatives of patients approaching the end-of-life in different settings (Anderson et al. 2019). It showed that the presence of an accompanying person (here a relative) in a consultation can be beneficial, but that it can also complicate the announcement of disease progression, showing the need to improve tools facilitating communication among protagonists. There is thus a need to determine how oncologists can support relatives (Schulz et al. 2018) and integrate them into the consultation to foster the therapeutic alliance. Training via simulation sessions has already shown promising results in improving communication between doctor and patient and could be adapted to communication involving relatives.

Limitations

The method used is original, and the number of triads included (47 and 20 physicians) is larger than in earlier studies (Korfage et al. 2013; Lee et al. 2018). However, 1 limitation is the difference between the potential study population and the oncologists who actually took part, and may also have selected patients, and recruitment was hampered by the small numbers of volunteers. Lack of time in consultations was the main reason given for non-participation. Furthermore, these findings are only transferable to countries with a similar cultural context, where doctors are encouraged to talk to patients and provide consistent information to their families. Finally, our use of grounded theory was only partial in that we did not reach the end of the theorization process. This study has cast light in particular on announcement techniques and the types of alliance within triads, 2 themes where the recurrences observed made it possible to reach data saturation.

Conclusion and perspectives

This study highlights possible improvements in triadic oncology consultations:

1. (i) Optimizing the consultation premises layout to ensure patient privacy,

2. (ii) Taking time to determine what patients and relatives know and/or understand about the situation, what they expect from the consultation, and what they are feeling,

3. (iii) Taking an interest in patients’ plans,

4. (iv) Taking an interest in the accompanier, assessing his/her state of health and preoccupations,

5. (v) Managing the consultation in more interactive mode, availability to respond to questions and asking participants to reword what has been said.

To address the issue of not enough time for these consultations, physicians should refer patients and families to psychosocial team members to provide robust support to patients, families, and loved ones transitioning to end-of-life care, and also to diminish clinicians’ psychological pressure. There is a need is to raise awareness among physicians of the challenges of this situation. By way of simulation sessions, they need to appropriate these recommendations to acquire the skills to integrate them into their practice.