Background

Patients with advanced chronic kidney disease (CKD) need to choose a kidney replacement therapy that best suits their needs, medical options and living situation [1]. In 2018, at least 17,670 Dutch patients received kidney replacement therapy [2]. Although various forms of dialysis have comparable clinical outcomes, the treatment options are preference-sensitive and impact patients’ life differently [3,4,5]. This leaves both patients and healthcare professionals in nephrology (HCPs) with the task of making a treatment decision together. To support this process, Dutch patients with CKD receive treatment modality education.

To make a decision, HCPs are legally required to fully inform patients about all available options, risks and consequences, apprise themselves of the patients’ situation and personal circumstances, and invite them to ask questions [6]. The Dutch Medical Treatment Act also obliges HCPs to make a treatment decision together with the patient [7, 8]. In such a process of shared decision-making (SDM), HCPs and patients choose the best evidence-based treatment option together after discussing all available options, their pros and cons, personal preferences, and the circumstances of the patient [7,8,9]. In other words, SDM results from weighing clinical guidelines against patient preferences.

Given the complexity of CKD and the preference-sensitive choice to be made, SDM in nephrology is challenging [10]. Our recent study shows that, notwithstanding best intentions, nephrologists tend to steer the decision-making process in a certain direction, thereby leaving less room for SDM [11]. Additionally, even after adjusting for the patient’s age, sex and disease stage, Dutch dialysis centres vary widely in the types of treatments provided, especially in the percentage of home dialyses out of the total number of dialysis treatments [12]. This suggests that SDM in Dutch nephrology is not yet widely implemented.

To our knowledge, no studies have so far investigated SDM during consultations in nephrology. Our first aim was therefore to assess SDM during consultations in which a decision for kidney replacement therapy was discussed. Our second aim was to have patients and HCPs reflect on their own communication and SDM behaviour for additional in-depth understanding of the practice of SDM in nephrology.

Methods

Study design

This study is embedded within a larger Dutch nephrology study called DIALOOG. Twenty-eight HCPs (i.e. nephrologists) from 21 Dutch hospitals were invited to participate (invited by AA and BvJ). In three academic and four general hospitals, data was collected between July 2019 and November 2020 by two researchers (RR and LS). Fourteen nephrologists and one specialist nephrology nurse who discussed final treatment options with patients with advanced CKD participated.

An explanatory sequential mixed method design was used with a quantitative observational part and a qualitative interview part [13]. The quantitative part started by collecting video-recorded outpatient consultations with the aim of 1) observing the extent to which SDM was applied in clinical practice, and 2) providing the basis for the stimulated recall interviews with patients and HCPs. Video recordings of consultations are a valid method for examining communication, a video recorder does not significantly influence behaviour [14,15,16]. SDM was investigated by the extent to which HCPs involve patients in SDM, using the ‘Observing Patient Involvement in Decision-Making’ (OPTION)5 instrument [17,18,19]. This instrument is used for analysing the video-recorded consultations by coding five items (see Table 2) for every individual video recording. This coding assesses the extent to which HCPs involve patients in SDM.

The theoretical framework underpinning the qualitative part of this study is based on a phenomenological approach, focusing on describing the meaning and significance of experiences. This design lets participants express their views on communication and SDM in nephrology in their own words. In the qualitative interview part, patients and HCPs reflected on their own SDM behaviour during individual stimulated recall interviews together with a researcher. During these interviews, video-recorded consultations were used to recall the outpatient consultation and to discuss patients’ and HCPs’ elicited perspectives, thoughts and reactions [20,21,22,23]. The participants were told that nothing they said would be passed on to others.

Procedure

Patients with CKD (eGFR < 20 ml/min) were invited to participate in the study by the participating HCPs, based on inclusion criteria and convenience sampling. They had to be aged =18 and awaiting a final consultation with their HCP to discuss a form of kidney replacement treatment. Patients were excluded from participating if they 1) were not able to speak Dutch sufficiently well, 2) had a severe intellectual disability, or 3) had a psychiatric problem or dementia. These consultations mark the end of a multi-consultation, educational process and aim to summarize all treatment aspects discussed previously.

Recruitment

We used the following procedure, which is the same as described in our previous paper reporting results of another analysis of the same dataset [11]. The HCPs of the participating hospitals were the contacts for the researchers. After local approval on feasibility from the participating hospital and selecting and inviting patients by the HCPs, the HCPs and/or researcher phoned eligible patients about 1 week before the scheduled visit to the hospital. During this phone call, the patients were informed in more detail about the aim and procedure of the study. The patients who agreed to participate were asked to sign an informed consent (IC) form before seeing their HCP. The HCPs signed an IC form too. For patients who have difficulty reading and understanding health-related information, plain language versions of the information letter and the IC form were made [11]. Before seeing the HCP, patients filled out a questionnaire with questions about their sociodemographic characteristics and their preference for participating in treatment decision-making. The outpatient consultation was then recorded using an unmanned camera. During the recording, only the HCP was visible. Afterwards, the video recordings were coded in a secured, locked room at Nivel [11].

Quantitative video-observations

OPTION5 is a reliable and valid instrument for investigating SDM [17,18,19]. Using OPTION on every video recording allows the extent to which HCPs involve patients in SDM to be assessed. Five SDM items (Table 2) are coded on a 5-point Likert scale, (0 = ‘zero effort observed’ to 4 = ‘exemplary effort’). The total OPTION5 score is generated by converting the scores to a 0–100 scale and then calculating the average. The higher the score, the higher the level of SDM. All 29 video-recorded consultations were coded by the main observer (RR), and four consultations were also coded by a second observer (JN). These four consultations were randomly selected. A rule of thumb is to have at least 10% of records independently coded by two observers [24, 25]. The coding results of the main coder (RR) are presented in Table 2. Inter-rater reliability between observers was calculated using Cohen’s kappa (.412), indicating a moderate inter-rater agreement [26, 27].

Qualitative stimulated recall interviews

After the video-recorded consultation, individual patients and HCPs looked back at carefully selected fragments of their own consultations with a researcher (RR, male, or LS, female). Recall interviews with patients were conducted at home or another place convenient for the patient (e.g. at the hospital in a private room), with the researcher, the patient, and in some cases, the significant other of the patient present. Recall interviews with HCPs were conducted at their workplace, with only the researcher and HCPs present. Only the researcher and the participants were present. Both researchers were occupied as communication researchers (MSc) at the time of the study, and as such, both investigated (and had general interests in) communication in multiple domains and settings in healthcare (e.g. in palliative care, primary care, and nephrology for DIALOOG). The researchers were experienced in conducting qualitative in-depth semi-structured interviews and stimulated recall interviews with patients and HCPs to reflect on their own SDM and communication behaviour (see e.g. 32). Both researchers were trained by co-authors (SvD and JN) to conduct interviews with HCP and patients, and both had no profound prior relationship with the participants. Furthermore, the participants were aware that the researchers were not medically trained or involved in patient care, and participants generally knew the goals and reasons behind the interviews (i.e. investigating clinical practice and perspectives on communication in nephrology).

Characteristics of the interviewed patients (age, sex and presence of significant other) and HCPs’ interviews (number of consultations discussed) were noted and so was the duration of the interviews. All interviews were audio-recorded, transcribed verbatim and anonymized; no field notes were made. Transcripts were not returned to participants for comments or corrections. During the interviews, patients and HCPs reflected independently on their own communication and SDM, and on the communication and SDM of the other person in the video-recorded consultations [28]. To do so, a topic list was developed (Appendices 1 and 2) based on literature [29], experience from previous research (RR, LS, JN and SvD) and feedback from two nephrologists (AA and BvJ). The list was not pilot-tested. For each interview, three fragments were selected to show to the patient as well as to the HCP. These fragments were approximately 2 min in length, based on the occurrence of SDM (using the OPTION5 protocol). The average duration of the interviews with patients, per consultation, was 54 minutes (range between 26 and 104 minutes). The average duration of interviews with HCPs was 37 minutes (range between 23 and 67 minutes).

Analysis

The interview transcripts were analysed using deductive thematic content analysis [30]. Deductive means that the themes presented in the results section are similar to the questions in the topic lists (see Appendices 1 and 2). Thematic content analysis indicates the process of reading, coding, comparing and discussing the answers provided by patients and HCPs and finalizing these answers to give presentable results. Results from patient interviews could be categorized into three main questions: 1) what is important in SDM for you, 2) what can you do better, and 3) what can your HCP do better? For the HCPs, these questions were 1) what is important in SDM for you, 2) what contextual factors could have influenced SDM, 3) what role do you give yourself in SDM, and 4) what could you do better? All transcripts were read carefully by one coder (RR) and parts in which elements of SDM were mentioned were selected. An initial coding was applied to the selected segments, discrepancies between researchers were resolved through discussion by two additional coders (JN and SvD, with RR), and modifications to the initial categories were made where necessary. All categories and patterns that emerged during analysis are illustrated by quotes (Tables 3 and 4). Participants did not provide feedback on the findings.

Results

Sample characteristics

Twenty-nine patients and 15 HCPs participated (Table 1). The patients had a mean age of 71 and 14 were women. Consultations lasted around 30 minutes. Thirty consultations were video-recorded, of which one was excluded because the patient appeared to have an intellectual disability. Fourteen stimulated recall interviews were conducted with patients and 16 with HCPs.

Table 1 Characteristics of participating patients and professionals
Full size table

3.2. Quantitative video observations

The mean SDM score (0–100 score) was 51 (SD = 14). The highest average score (see Table 2), was observed for items 1 and 4, the lowest for item 2. This indicates that, on average, the effort put into involving patients in SDM in practice was moderate (basic phrases or sentences used).

Table 2 OPTION5* scores from observations of patients and professionals (n = 29)
Full size table

Qualitative stimulated recall interviews

The stimulated recall interviews took place between September 2019 and November 2020, conducted by two researchers (RR and LS) within, on average, 7 weeks after the recorded consultation. The patient interviews lasted 54 minutes on average and the HCP interviews 37 minutes (Table 1). Tables 3 and 4 show questions, themes and illustrative quotes regarding SDM perspectives of patients and HCPs respectively. In this section, the most salient examples of diverse perspectives are presented.

Table 3 Questions, themes and patients’ quotes
Full size table
Table 4 Questions, themes and HCPs’ quotes
Full size table

Patients’ perspectives

What is important in SDM for patients?

Most patients value the HCP’s opinion about suitable medical options because they believe that the HCP has more knowledge (Table 3). Nevertheless, the degree of significance assigned to the opinion of the HCP differs between patients. Some patients believe that the opinion of the HCP is very important for making a decision and that opposing that opinion is useless. Other patients perceive the HCP’s opinion as non-mandatory advice. These patients underline that the final decision still has to be made by them, not by the HCP.

The type of information provision in SDM is also deemed important by patients; however, patients differ in this respect. As illustrated in Table 3, some patients appreciate visiting, meeting and talking to other patients who are treated with dialysis, which supports them to conceive this particular treatment option. Other patients, however, strongly oppose meeting other dialysis patients as this would negatively affect their perception.

What can patients themselves do better in SDM?

Taking control and asking more questions were mentioned as strategies by patients to acquire more knowledge about the different treatment options. Another strategy that might work is taking notes during the medical visit.

What can HCPs do better in SDM according to patients?

All patients want their HCP to be consistent and reliable in the information that is provided during the SDM process (Table 4). Furthermore, according to some patients, errors in scheduling hospital appointments arouse distrust, whereas HCPs’ attentiveness and availability is much appreciated and provides patients with a sense of security.

HCPs should provide more information and ask the patients more questions about their treatment preferences. In addition to getting a time indication for dialysis, and transplantation in general (Table 3), patients wanted more information about who is to decide if the patients remain in doubt, the prospects and prognosis of having limited kidney function, the causes of kidney failure (e.g. an unhealthy lifestyle), conservative treatment, the burden of dialysis on everyday life and on significant others.

Healthcare professionals’ perspectives

What is important in SDM for HCPs?

According to most HCPs, SDM is a continuing process (Table 4). Most HCPs state that when a decision is made, this does not mean that this choice is irreversible. Some HCPs furthermore appreciate the input from significant others. After all, most patients will make their decision together with those significant others before visiting the hospital. In the process of SDM, HCPs see themselves only as confirming and documenting the final decision. Furthermore, HCPs vary in the significance they attach to decisions that have to be made, e.g. from a medical perspective, the choice for transplantation or conservative treatment is considered more important than the choice between haemodialysis (HD) or peritoneal dialysis (PD). Yet they do underline that this might not be the case for patients, for whom practical and everyday consequences of these treatment options are equally important.

Many HCPs said that patients need to be aware of all options and risks. Furthermore, according to the HCPs, it is important to explore and validate the motivation of patients for certain treatment options (Table 4). Many HCPs state that when patients have understandable, valid and medically realistic reasons, the decision is acceptable to them, even when that decision differs from their own preference. When patients’ reasons are not valid or based on incorrect arguments or not medically realistic, this expectation needs to be corrected by the HCP and the decision should be reconsidered.

HCPs indicated varying degrees of taking control during SDM. Many HCPs stated that taking control as medical experts and when medical technicalities are important (Table 4). However, within this role, some HCPs say that it is important to keep listening to patients’ preferences. For instance, when the cognitive abilities of patients are limited, HCPs indicated to take responsibility and control over the decisions of patients, and within this role, continuously try and elicit the preferences of patients. Some of the HCPs mentioned strategies that facilitate taking control during SDM: referring the patient to the transplantation-team and using the team’s information provision and advice to back the decision for deciding on a transplant; using the patient’s age as an argument for deciding on a transplant; tailoring information arguments with knowledge and information to persuade the patient towards a decision.

What are contextual factors that influence SDM according to HCPs?

According to some HCPs, not every patient wants to hear or talk about dialysis (Table 4). Patient characteristics mentioned by the HCPs that diminish patients’ active participation in decision-making are their psychological condition (e.g. distressed or depressed patients); their attitude (e.g. patients being rigid in their arguments and beliefs); their cognitive processes and abilities that are sometimes hard to follow.

Furthermore, HCPs experience differences between patients’ preferred treatment option and the medical possibilities (Table 4). Another aspect mentioned by HCPs as influencing SDM is when patients and HCPs differ in their expectations of the consultations. Also, according to most HCPs, differences in medical knowledge and expertise influence SDM.

What role do HCPs give themselves in SDM?

In general, HCPs see five roles for themselves in SDM (Table 4). When patients are cognitively too impaired to decide, HCPs’ role is to coach and support the patient. In this role, HCPs emphasize being explorative, empathic and comforting to the patient. However, when patients are in doubt, their role of a responsible practitioner with a more rigid, directive and persuasive communication style prevails. Some HCPs said that, from a medical perspective, the HCP should always take the lead, taking responsibility for accurate information provision and SDM in general.

What can HCPs themselves do better in SDM?

In addition to maintaining control and providing structure in the conversation, other ways to be more clear towards the patient and take control were summarizing more often, using fewer words, refraining from too much in-depth conversation, being more clear and honest about the possibility for transplantation and the end-stages of conservative treatment.

Discussion

This explanatory study assessed the extent to which SDM was accomplished during consultations in nephrology in which a final decision for kidney replacement therapy has to be made. Overall, SDM is applied by the HCPs to a moderate extent. Compared to SDM scores in Dutch oncological and palliative care settings [31, 32], the average SDM score in Dutch nephrology seems promising. However, improvement is still needed, given recent survey results indicating that 30% of patients with kidney failure say they did not receive information or support when choosing a kidney replacement therapy [11]. Our interviews showed that all patients rely on the information provision and opinions of HCPs during SDM, expect consistency and support, and want a proactive role in SDM. They would like to receive more information from their HCPs and to be questioned about their preferences. HCPs appear willing to incorporate patients’ preferences in SDM when there are no medical contraindicators, the patient is fully informed and decides using valid arguments. HCPs also like the patient to take a more prominent role in SDM which is in accordance with previous research [33].

Our quantitative observations showed that HCPs least often reassure patients or reaffirm that they will support them by giving information or deliberating the options (2nd item of the OPTION instrument). This element of SDM seems particularly important in nephrology, though, given the difficulties patients face in choosing a preference-sensitive treatment option [2,3,4]. In addition, patients with limited health literacy, i.e. a limited ability to access, understand, appraise and apply health information in making healthcare decisions [34], are prevalent in kidney care [35]. These patients are known to be relatively passive during SDM, less inclined to take control during the conversation and prone to follow the HCP’s advice [36, 37]. For them, additional support to let them become better informed or deliberate options seems needed.

Supporting patients in their decision-making process should not be restricted to the interaction during scheduled outpatient visits. A patient who has received information about the various treatment options beforehand will participate more actively in the decision-making process with the HCP. To that end, various tools have been developed in the Netherlands as well as in other countries. An example of this is the decision aid on kidney failure (in Dutch only) developed in close collaboration with nephrologists. In addition, the Dutch patient organization for patients with kidney disease developed an educational website (Nierwijzer.nl) with specific information about the various treatment options [1]. For patients with lower levels of health literacy and known to have difficulty reading and understanding written language, educational tools are available with visual information and illustrations that help them understand better what the different options entail. These tools can also be discussed during outpatient visits. However, discussing treatment options and participating in decision-making requires more than being informed. As our study shows, specific communication skills are needed such as asking questions, expressing concerns and interrupting an HCP when the information provided is too complex or upsetting. Patients often experience communication barriers, as was also demonstrated in a previous study among patients with different types of chronic diseases [38]. Designing communication-enhancing interventions together with patients could help them to overcome these barriers. Such interventions can, however, only be effective when HCPs also take their communicative role seriously and realize that inquiring about patients’ treatment preferences and listening to their needs and worries is crucial for achieving the goal of person-centred care and preventing the implicit persuasion which now occurs quite often and hinders patient participation in decision-making [11].

An important strength of this study is that we assessed SDM in everyday practice in multiple centres and discussed patients’ and HCPs’ contributions to this process afterwards. Furthermore, real-life video-recorded consultations were collected, representing valid and reliable data. In addition, the stimulated recall interviews were perceived as educational and worth the invested time. A final strong point was that various elements of the communication were unravelled in this study, giving nephrologists an understanding of many educational aspects of SDM.

Nevertheless, there are also some methodological weaknesses. First of all, the inter-rater reliability between the two observers was only moderate and the average SDM score was somewhat higher according to the second observer. This could point to an underestimation of SDM. However, the main coder had more experience in nephrology and therefore the (lower) coding by the main coder is represented in the results section. Even so, both observers were trained and have ample experience in coding with the OPTION. Second, in addition to how it is applied by HCPs, SDM also depends on patient characteristics and the context of the consultation [39]. Considering this, the relatively small number of hospitals (n = 7) and consultations (n = 29) and the explorative research design could have reduced the study’s external validity. Additionally, because of the explorative design and relatively small number of participants available, data saturation in the qualitative part of this study is also difficult to discern. Still, within the clinical context of strict privacy regulations and time constraints, 14 nephrologists allowed us to make recordings of their consultations and participated in the stimulated recall interviews. Thirdly, one consultation was recorded and analysed per patient. However, as HCPs’ interviews showed, SDM is a continuing process that involves multiple consultations with multiple HCPs. Information not discussed in the observed consultation could therefore have been mentioned in earlier consultations, indicating a possible underestimation of SDM. Fourthly, 15 HCPs participated in this study. They were the ones who agreed to participate out of a group of 28 HCPs who were invited to join. We did not record the characteristics of the non-responders. It is perfectly possible that they are not representative of the overall group of nephrologists in the Netherlands: maybe more skilled communicators volunteered, maybe less-skilled ones did. Our SDM ratings could therefore be an overestimation, although we did find similar ratings in other medical specialties [31, 32]. Lastly, the OPTION5 instrument does not measure the overall quality of communication [17]. This could mean that although SDM scores are moderate, the overall quality of communication is still adequate. For a more conclusive and comprehensive assessment of communication and SDM, other communication and decision-making aspects should be measured as well. An example of this is (implicit) persuasion, in which HCPs, despite their best intentions, convey a treatment preference and steer the decision [11]. Persuasion is mentioned by HCPs in our interviews when describing their role in SDM.

Conclusions

In conclusion, in Dutch nephrology, SDM is applied by HCPs to a moderate extent. However, every patient should be involved in SDM. In the case of kidney failure, encouraging patient participation is even more important, given the complexity of the disease, the relatively high prevalence of limited health literacy in patients [40] and the life-changing choices they are required to make. Reductions in SDM could result in decisional regret [41], harm the quality of life [42, 43] and lead to (unwarranted) practice variation [44]. Offering patients tools to support decision-making, such as educational websites and decision aids, could prevent these unwanted outcomes.