Facilitating patient-oncologist communication in advanced treatment-resistant cancer: development and feasibility testing of a question prompt list

Abstract

Background

Patients’ expectations regarding medical information in advanced stages of cancer are still poorly understood. Tailoring information to advanced cancer patients is a subtle task. We developed a question prompt list (QPL) that serves as a patient-oncologist communication aid in France.

Methods

A four-step sequential mixed method involving patients with luminal B/triple-negative metastatic breast cancer or metastatic uveal melanoma (N = 110) and patients’ partners, oncologists, and researchers (N = 18) was used. In-depth interviews and questionnaires focused on the information needed at the disclosure of metastasis or resistance to treatment (step 1), the formulation of questions and procedures for use in oncology visits (steps 2 and 3), and the acceptability of the final tool (stage 4).

Results

The initial version of the QPL consists of 17 questions covering 5 themes (disease, current treatment, other options, living with cancer, prognosis). In step 2, 13 questions were added, 2 were merged, and 5 were deleted; a short form (4 questions) and recommendations for clinical use were proposed. In step 3, 2 questions were merged, and 6 were deleted. Four oncologists (27% of the target population) took part in step 4, and the QPL was discussed with 20 patients, revealing a positive appraisal.

Conclusion

We provide a rigorously developed, relevant, concise, and acceptable question prompt list for clinical application in the advanced cancer care setting in France. Further research needs to assess whether this tool actually facilitates oncologist–patient communication and improves satisfaction with care and health outcomes.

Trial registration

The study is listed on ClinicalTrials.gov (NCT04118062) and registered under identification n° IRRID “International Registered Report Identifier”: DERR1-10.2196/26414.

Key message regarding feasibility

1. What uncertainties exist regarding feasibility?

• The use of communication aid in the form of a question prompt list in advanced cancer-stage oncology consultations in France is uncertain.

• A culturally tailored, relevant, concise QPL rigorously co-designed with patients, oncologists, and patient partners should be acceptable and feasible for use by patients and clinicians.

2. What are the key feasibility findings?

• The final version of the QPL proved to be acceptable and satisfactory. It was used in two-thirds of the consultations where it was made available and was perceived as useful and easy to use by both oncologists and patient participants.

3. What are the implications of the feasibility findings for the design of the main study?

• The developed QPL comprises a comprehensive and concise list of questions to facilitate communication between patients and their oncologists in the specific context when the disease becomes metastatic or resistant to treatment.

• The use of this QPL procedure (e.g., availability on an intranet, provided by an oncologist, at initial consultations, in its long or short form to facilitate a progressive discussion about the disease prognosis, computerized warning system) is suggested.

Background

Informing patients about the metastatic status, progression, or resistance of their cancer is a sensitive and challenging task for oncologists. This crucial discussion involves imparting information about the incurability of the disease, available treatment options, potential side effects, and the effectiveness of these treatments in terms of symptom relief or survival. Most patients are eager to receive this information [1]. At this illness stage, patient-oncologist communication encompasses many sources of uncertainty that need to be addressed, especially since it could serve to uphold the patient’s sense of hope, which is crucial for their overall well-being and coping mechanisms [2]. Moreover, studies have shown that discussing prognosis does not appear to negatively impact the patient-oncologist relationship or the emotional state of the patient [3]. However, during these conversations, oncologists often grapple with the fear of exacerbating a patient’s anxiety or taking away their sense of hope. Additionally, many oncologists feel ill-prepared to answer questions about prognosis [4]. They may believe that patients will express their need for information if desired, while patients often expect oncologists to provide important information without the need for explicit requests [5]. Interestingly, some patients prefer to avoid discussions about dying and palliative care altogether and only desire this information if they directly ask for it [6]. It can be challenging for oncologists to gauge the specific type and extent of an individual patient’s information needs [7]. Furthermore, it should be noted that addressing treatment resistance or metastatic disease entails a high volume of complex information on various treatment options and prognostic consequences; this requires particular communication skills to tailor the communication approach to patients’ needs. Consequently, many patients do not receive the crucial information they require to actively participate in treatment decisions and prepare for their future, or they receive it at the wrong time [1, 8].

To overcome patient-clinician communication difficulties, tools such as question prompt lists (QPLs) are provided so that the patient can select and express his/her own questions during the oncology consultation [9, 10]. This approach facilitates the identification of the patient’s individual concerns to adapt the content of the information provided. These communication aids have proven to be inexpensive and easy to use in the limited clinical time of oncologists without increasing the length of consultations [10, 11]. Most of these tools were developed for the early stages of cancer or at the end of life [12]. For example, our team demonstrated the beneficial effects of a list of questions designed to facilitate end-of-life discussions during palliative care consultations [11]. Patients in the experimental group expressed questions more frequently without experiencing additional anxiety. This list, developed for the end-of-life phase of the disease trajectory, was found to be inappropriate for the period surrounding the diagnosis of metastatic cancer or resistance to cancer treatment, which provides many opportunities for discussion between patients and oncologists about disease status, prognosis, and therapeutic alternatives [13, 14].

Methods

The study took place at a specialist oncology center (Institut Curie, Paris, France). Details of the protocol have been published [12]. Ethics approval was received in July 2019 (research project no.: 19.06.21.64751; MS4 RIPH2 HPS N° ID RCB: 2019-A01713-54).

Written informed consent was obtained from all participants at each step of the project.

This study aimed to develop a QPL for the early advanced cancer care period, i.e., the diagnosis of metastatic disease or disease progression. Moreover, as little is known about patients’ expectations regarding medical information in this particular medical context in France [15], the development of this tool was based on iterative exploratory interviews. It involved potential users (patients and oncologists) to ensure that the resulting tool would be acceptable and usable in routine care.

Study design

A mixed qualitative and quantitative method was used [16]. The tool was built in four steps on the basis of data collected successively [17]. Based on a constructivist grounded theory approach, the aim of the first step consisted of interviews to explore information needs from the patient’s perspective and to generate a comprehensive list of questions covering these concepts. Then, we formulated questions likely to be addressed by patients during oncology consultations when cancer metastasis or resistance to treatment is identified. The second step involved oncologists, researchers, and patient partners in focus groups to review the content of this list and the questions’ specific wording and to propose modalities for using the tool in clinical practice. Step 3 involved the same participants as step 2, using an online survey, and aimed to select the essential questions for a short, concise, and feasible tool in practice and to review their wording, the QPL introduction, the format, and structure. Step 4 assessed the perceived QPL acceptability in clinical practice through patients’ and oncologists’ self-assessments.

Step 1–self-administered open question and interviews

Inclusion criteria and patient selection

Eligible patients were 18 years of age or older and had either metastatic triple-negative or luminal B breast cancer, or metastatic uveal melanoma for which the metastatic diagnosis or resistance to treatment was announced less than 3 months prior to the study. These cancer types were chosen because, although they are difficult to treat, specific oncological treatment options can still be offered for several months or even years; however, their prognosis remains poor [18,19,20].

Procedure

A sample of 80 consecutive patients (N = 40 per tumor site) was selected from medical oncology consultation lists from January 2021 to January 2022. They were asked to participate in the study by their oncologist and were invited to complete a self-administered questionnaire. Additionally, some of them, selected by chance (N = 20 per tumor site to ensure data saturation [21]), took part in an interview conducted by a research psychologist (AR) following the oncology consultation, either in person at the hospital, by telephone, or videoconference.

Self-administered open question

One question from the European Organization for Research and Cancer (EORTC) QLQ-INFO25 [22] was addressed: “Would you like to receive more medical information and, if so, on which subjects?”.

Semistructured interview

A semi-structured interview based on a guide developed from the literature [23,24,25,26,27] explored information needs in the period during which metastatic or treatment-resistant cancer was diagnosed.

Analysis

The answers to the self-administered questions were collected, and the interviews were recorded and transcribed verbatim. A thematic analysis of the textual data was carried out on the basis of a grid developed in two phases [28]: (1) from the interview guide and free comments in response to the self-administered question and (2) from an independent analysis of 10 interviews by two researchers (AR et JT). Intercoder reliability was tested and revealed an 87% agreement rate and a Cohen’s kappa coefficient of 0.85. The coding grid, initially developed by considering the two tumor sites separately, turned out to be similar; therefore, it was applied to all interviews (N = 40). Iterative reading of interviews made it possible to generate recurring themes and idiomatic formulations. Iterative discussions within the research team (AR, JT, SD, AB) helped establish a consensus on the meaning of the codes. A count of the code frequency was carried out, and these quantitative and qualitative data were integrated with the information garnered in the following steps.

Step 2: focus groups

Participants

Participants in two successive focus groups were selected among the institution’s oncologists, oncology researchers, and patients’ partners with an equal distribution among these three categories of stakeholders. They were contacted by an email that included a study information note. Fifteen subjects, 7 and 8 per group and 5 per type (oncologist, researcher, or patient partner), participated [29].

Procedure

Group interviews were conducted by a research psychologist (AR) accompanied by one observer per group (AB and SD). The interviews were recorded and transcribed verbatim. This step involved videoconference.

The initial list of questions from step 1 was sent to the participants beforehand. During group interviews, participants were invited to discuss the relevance of the questions, possible omission of themes, clarity and appropriateness of wording (questions, instructions), format and structure of the tool, and to suggest modalities for maximizing the use of the tool in clinical practice and minimizing implementation obstacles.

Analysis

Group interviews were recorded and transcribed verbatim. Their analysis consisted of listing and categorizing all the comments. Redundant, too specific, potentially disturbing, or unclear questions were deleted, merged, or reworded based on criteria discussed within the research team (AR, JT, SD, AB). When a consensus was not reached between participants on deleting a question from the list, it was decided to keep that question.

Step 3: survey

Participants

We approached 15 participants who participated in focus groups and 10 additional participants, including 5 oncologists, 2 researchers, and 3 patient partners.

Procedure

An online survey was conducted using REDCap software [30]. As the conciseness and clarity of the tool (and thus the number of possible issues to discuss during the consultation) was considered a prerequisite for its use in everyday practice [31], the aim was to retain the essential overarching questions. Therefore, in addition to free comments, opinions were sought regarding the degree of importance of each general question on a 5-point Likert scale ranging from 1 = not at all to 5 = completely. Any sub-questions could be added as examples.

Analysis

A quantitative analysis (frequency, mean, and standard deviation) of the importance of overarching questions was performed to obtain a list that covered all the themes in a comprehensive and acceptable manner while remaining concise to facilitate its use in clinical practice [32]. A 75% rate of “important” and “very important” responses and a score of 4 or more per question statement for the overall sample were determined to retain this question [33]. Qualitative comments on rewording or redundancy were taken into account. In the case of disagreement between quantitative and qualitative information, it was decided to keep the question.

Step 4: QPL acceptability in routine practice

Inclusion criteria

Step 4 included patients who met the same inclusion criteria as those in step 1.

Procedure

All oncologists treating metastatic triple-negative and luminal B breast cancer and metastatic uveal melanoma at our institution (n = 15) were contacted. The QPL was provided by the oncologist to consecutive patients meeting the eligibility criteria (n = 30) during an oncology consultation to announce the metastatic diagnosis or resistance to treatment. Patients were invited to read the list at home, identify their own questions, and address them, if they wished, at a subsequent consultation. The oncologist was invited to remind the patient about the possibility of using the QPL during that consultation.

Questionnaire

On a five-point Likert scale ranging from 1 “strongly disagree” to 5 “strongly agree”, patients and oncologists appraised the QPL referring to the consultation where the QPL was discussed on a five-point Likert scale on the following items: “this document is easy to use” (1); for the patient, “this document helped me ask my oncologist questions” (2a); “this document helped me ask questions about subjects that are important to me and that had not been discussed with my oncologist” (3a); “I plan to use this document for my next consultations with my oncologist” (4a); and “my oncologist encouraged me to ask the questions I had selected in this document” (5a); and for the oncologist, “this document has helped me in my communication with my patient” (2b), “I plan to use this document with other patients” (3b), “I have encouraged my patient to ask me the questions he selected in this document” (4b) and “using this document hasincreased the time of the consultation” (5b).

Analysis

Score means and standard deviations were calculated for patients’ and oncologists’ item responses. The item 5 score for oncologists was inverted so that all items with a high score indicated favorable judgment.

Results

Step 1

Sample description

Seventy-eight of 80 patients took part in stage 1, 39 by tumor site. The mean age was 62.1 years (standard deviation = 11.6) and 52 patients were women (Table 1). Two patients were lost to follow-up. Of the 40 patients who participated in the interviews, the average age was 61.4 years (standard deviation = 12.2), and 29 were women.

Table 1 Sociodemographic and clinical patient characteristics in Step 1 and Step 4^a

List of issues

For the self-administered question, 50 (64%) patients expressed the wish to receive more information about the following areas: current treatment (n = 22), prognosis (n = 15), current medical situation (n = 6), “living with the disease” (n = 6) and other treatment options (n = 5). Data saturation was obtained after 18 and 17 in-depth interviews for patients with metastatic triple-negative and luminal B breast cancer and with metastatic uveal melanoma, respectively, raising information needs on these same themes: current treatment (n = 12 and 12), prognosis (n = 9 and 8), current medical situation (n = 10 and 6), other possible treatments (n = 4 and 6), and living with the disease (n = 6 and 2).

Analysis of the interviews by tumor site did not reveal differences in terms of the revealed general themes; as a result, we chose to develop a single QPL.

With respect to patients’ idiomatic formulations, 41 questions, including 7 general questions and 34 specific sub-questions, were developed.

Step 2—Focus group

Sample

Among the participants in the focus groups, the mean (standard deviation) age was 54.7 (11.9) years, and 12 were women. The interviews lasted 70 and 74 min per group.

Initial QPL revision

Among the modifications made to the initial QPL (see Table 2), 13 questions were added, 5 were deleted, and two questions were merged, resulting in 48 questions. A two-tier QPL structure comprising general and specific questions (in italics) was proposed to lighten the tool’s visual aspect while providing a maximum number of suggested questions. Summary instructions and a diagram explaining how to use the tool (i.e., provided at initial consultations, read at home, and referred to at a subsequent consultation depending on the timing of the patient’s information needs) are provided. Figure 1 shows a snapshot of the first and last page of the final QPL. During focus groups, a short version of the tool was suggested in the form of 4 questions considered essential (i.e., the objective and duration of treatment, side effects, and curability of the disease). This would allow oncologists to gradually address more delicate questions about prognosis during subsequent consultations.

Table 2 The QPL main questions or sub-questions added, deleted, or modified after step 2 (focus groups) and step 3 (online survey)

Fig. 1

QPL format, first and last page*

Step 3—Survey

Sample

Eighteen of the 25 subjects approached took part in step 3, including 8 oncologists, 3 researchers, and 7 patient partners, with an average age of 51.3 years (standard deviation = 12), and 14 women.

QPL revision from step 2

Comments from participants (N = 118, with an average of 4 per question) led to reformulations and to the deletion of 6 of the 48 questions resulting from step 2 (see Table 2). The degree of importance of the 15 general questions and 26 sub-questions resulting from step 3 is high, with an average ranging from 3.89 to 4.89 (Table 3).

Table 3 Frequency of information questions in step 1 and estimated importance per general question in step 3

Step 4—QPL acceptability in clinical practice

Sample

Four out of 15 oncologists (27%) agreed to participate in step 4 and included 30 consecutive patients in total over the period July to November 2022. Oncologists did not provide the list of questions to 7 of the 30 patients included, either because it was forgotten (N = 6) or because he/she considered the patient to be psychologically unable due to denial of the severity of the cancer (N = 1). Discussion with oncologists about the QPL was carried out for 20 (67%) of the included 30 patients during a subsequent consultation, on average 3 months later (range 1 to 10 months). The QPL was not discussed either because the patient forgot to read it (N = 1) or because the oncologist forgot to refer to it during the consultation (N = 2).

QPL appraisal

The QPL appraisal was based on its use during 4 to 5 consultations, depending on the oncologist and patient. One patient was excluded because he had participated in step 1, and two patients did not return their answers.

The mean appraisal scores ranged from 3.61 (helps to ask questions not raised during the consultation) to 4.56 (easy to use) for the patients and from 2.00 (longer consultation time) to 4.29 (easy to use; helped to encourage the patient to ask questions) for the oncologists (Table 4).

Table 4 Mean (SD) response to the QPL acceptability assessment^a (N = 20 patients; N = 4 oncologists)

Discussion

A communication aid in the form of a question prompt list (QPL) was co-designed with patients affected by cancer with a severe prognosis, patients’ partners, oncologists, and oncology researchers. It comprises a comprehensive and concise list of questions to facilitate communication between patients and their oncologists when the disease becomes metastatic or resistant to treatment. This tool is original in specifically targeting patients’ information needs at a particular time when oncological treatments, which are potentially effective in terms of survival, can still be offered, even though the disease is currently incurable [14]. Adequate patient-oncologist communication is of utmost importance, especially at this stage of the disease trajectory, when addressing the uncertainties related to care, treatment options, and prognosis can have a profound psychological impact on patients navigating the complexities of advanced cancer care [2].

Throughout the study’s steps, particular attention was given to the tool’s clarity and comprehensiveness of content, its format and structure, and the delicate process in which it is expected to be used in clinical practice. The final version of the QPL proved to be acceptable and satisfactory. It was used in two-thirds of the consultations where it was made available and was perceived as useful and easy to use by both oncologists and patient participants, which is promising for systematic implementation in everyday practice. Moreover, only one patient out of 30 was identified by their oncologist as potentially psychologically unable to use the QPL. This low incidence emphasizes the tool’s potential to bridge communication gaps effectively and suggests a high degree of patient readiness and receptiveness, ultimately enhancing its overall acceptability and utility in clinical practice.

This tool validates patients’ various concerns and reminds them to discuss subjects they perceive as important, even though they may seem trivial (e.g., the continuation of non-oncological treatments), inadequate (e.g., a planned pregnancy), or embarrassing (e.g., the request for a second medical opinion; the use of complementary medicine). Particular care was taken in formulating questions about the prognosis to ensure that they were acceptable not only to the patient but also to the oncologist, who is often uncomfortable communicating on this issue.

The suggested procedure of this QPL (i.e., provided by the oncologist, at initial consultations, in its long or short form as the oncologist feels is appropriate, patients’ invitation to read it at home and to select questions they wish to discuss at a subsequent consultation) was designed to encourage patients to address questions of interest to them personally and in the time frame suiting their needs [34, 35]. To facilitate its use in real life, the QPL is now available on the institution’s intranet portal, and oncologists are regularly informed or reminded of its existence. In conjunction with our IT, we also plan a computerized warning system when an oncology consultation is planned to announce treatment resistance. The provision of the QPL is considered a medical act that is tracked in the patient’s medical record for quality assurance. Short training sessions could also be offered to help oncologists address sensitive communication issues such as prognostication in early advanced cancer care [36].

Oncologists’ preferences for informing patients about a severe disease prognosis and resulting treatment alternatives vary according to culture [37, 38]. This communication task is difficult for any oncologist, requiring sensitivity to both individual and cultural factors (coping attitudes, communication skills, etc.) [39]. Therefore, in this study, we adapted the QPL content and administration modality as follows. The participants suggested that the oncologist propose the QPL at initial consultations instead of providing it via open access. This approach limits its accessibility; however, it was found to be important that the purpose of the QPL be first explained by the oncologist, who encourages the patient to ask questions at subsequent consultations [10]. With this approach, oncologists show personal commitment to QPL use, and they can also assess patients’ ability to address upsetting issues. When patients are given bad news, they are particularly vulnerable; therefore, information must be delivered gradually and sensitively [40, 41]. This modality of QPL use is a cultural adaptation to France; it was meant to avoid discouraging its use in clinical practice. The QPL short form allowed us to take into account the timeframe of patient needs when discussing disease prognosis.

Among the few studies showing QPL cultural adaptation, Walczak et al. [42] reported the need to reformulate US English with respect to Australian English to allow idiomatic expressions and to place questions on prognosis at the end of the document. For Italy, the formulation of questions originally developed in Australia [43] was revised to be less direct, preferring, e.g., the term “disease” rather than “cancer”. The Norwegian adaptation of the same QPL [44] emphasized the importance of the general practitioner as a source of information [45].

The present QPL covers questions about the current disease status, disease curability, treatment purpose, disease duration, side effects, options including participation in a clinical trial, impact on daily life, lifestyle recommendations, sources of support, and disease prognosis. This content includes some questions similar to those developed for end-of-life cancer care [8, 46,47,48], although these questions may be given less emphasis. This phase differs from the early cancer diagnostic phase [44, 49], which includes questions relating to diagnostic tests and the specific diagnosis of cancer, and from the end-of-life phase, which raises questions about multidisciplinary and palliative care and the involvement of relatives [13, 47]. With respect to disease prognosis, patient-oncologist communication about advance care planning is particularly important for future end-of-life care decisions [50]. The QPL addresses this topic despite divergent opinions among participants in this study who worried that it would be raised too early at this point in the patient’s initial disease progression.

At the end-of-life phase, our team showed that patients did not discuss the disease prognosis more often using a QPL [11]. According to the literature [51], the interviews in this study highlighted the importance of raising the prognostic issue, even if the oncologist is afraid to do so because of the uncertainty about the figures involved. Patients are often unaware of poor prognoses [52]. However, being informed and able to participate in medical decision-making confers psychological resources to the patient [53, 54].

A specific QPL per tumor site was initially considered, but the similar content identified during patients’ interviews led to the development of a single tool. The nature of the disease seems to have less of an impact on information needs than the specificity of the time of treatment, whereas in this case, needs were particularly acute around questions about the stage of the disease and how it develops over time, irrespective of the type of cancer.

Limitations of this study include single-center participation, selected cancer sites, and likely participants’ interest in cancer care communication. More participants among the patient partners than among the oncologists and researchers at each step of the study could have given more voice to the patient perspective; however, we were limited by the number of patient partners available. Only 25% of the targeted oncologist population agreed to participate in the study. However, this figure is in line with preliminary data on new intervention implementation [55], and the QPL was raised by the patient or the oncologist during at least two-thirds of the consultations following the initial consultation where the QPL had been handed (Supplementary file 1).

Conclusion

With the use of a rigorous co-design approach, we provide a relevant and simple tool to aid patient-oncologist communication in the advanced cancer care setting in France. Consensual recommendations for the application of this tool in routine clinical practice are also given. Further testing of this QPL is needed to determine its clinical benefits. The obstacles to its use, such as institutional opportunities (e.g., availability of communication training; facilitating change in clinical habits) and personal characteristics (e.g., motivation) [32], should also be addressed.