The decision process in real time shared decision-making in consultations about palliative systemic treatment for advanced cancer
Aranka Akkermans 10262954
University of Amsterdam Faculty of Humanities Master’s Thesis
RMA Communication and Information Studies: Rhetoric, Argumentation Theory and Philosophy
Date of completion: 26 June 2017
Supervisor: Dr. A.F. (Francisca) Snoeck Henkemans Second reader: Dr. J.H.M. (Jean) Wagemans
1
Preface
First of all, I would like to thank my supervisor, Francisca Snoeck Henkemans, for providing excellent supervision along the way. I am very thankful for the input and motivation Francisca provided me with.
I would also like to thank Nanon Labrie, my co-supervisor, and Inge Henselmans for their feedback and input from a more medical psychological perspective. Nanon has
contributed a great deal to the project, and Inge was of great help by standing in for Nanon. I am very thankful to them both.
I greatly appreciate that I was allowed to use Inge Henselmans’ and Hanneke van Laarhoven’s data enabling me to write my thesis. I also want to thank Inge and Hanneke for their feedback.
A special thanks goes to Martijn Demollin for taking the role of the second coder and to Nanon for providing me an insight into shared decision-making education.
2
Table of contents
Preface ... 1 1 Introduction ... 4 2 Current characterization of SDM and its shortcomings ... 6
2.1 Current characterization of SDM in consultations about palliative systemic treatment for advanced cancer ... 7 2.2 Shortcomings of the current characterization of SDM in consultations about palliative systemic treatment for advanced cancer ... 10 3 A pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer ... 12
3.1 SDM in consultations about palliative systemic treatment for advanced cancer as an argumentative activity type ... 12 3.2 Institutional constraints for SDM in consultations about palliative systemic treatment for advanced cancer ... 13 3.3 A procedure for SDM in consultations about palliative systemic treatment for advanced cancer ... 15 3.4 Benefits of a pragma-dialectical characterization of SDM in consultations about
palliative systemic treatment for advanced cancer ... 17 4 Weighing and balancing in SDM in consultations about palliative systemic treatment for advanced cancer ... 18
4.1 The process of weighing and balancing in SDM in consultations about palliative
systemic treatment for advanced cancer ... 18 4.2 The basic prototypical argumentative patterns in SDM in consultations about palliative systemic treatment for advanced cancer ... 19
4.2.1 The basic prototypical argumentative pattern involving symptomatic argumentation ... 20 4.2.2 The basic prototypical argumentative pattern involving pragmatic argumentation 23 4.3 The basic prototypical argumentative pattern based on weighing and balancing in SDM in consultations about palliative systemic treatment for advanced cancer ... 24 5 Objectives of the content analysis ... 26 6 Methodology ... 29
3 6.1 Content analysis ... 29 6.2 Sample ... 29 6.3 Procedures ... 30 6.4 Measures ... 31 6.5 Validity ... 31 6.6 Pilot study ... 32 6.7 Study reliability ... 33
7 Results and discussion ... 33
7.1 Argumentation in SDM in consultations about palliative systemic treatment for advanced cancer ... 33
7.1.1 Symptomatic argumentation and pragmatic argumentation ... 34
7.1.2 Authority argumentation ... 37
7.1.3 A comparison between first consultations and follow-up consultations ... 38
7.2 The decision rule in SDM in consultations about palliative systemic treatment for advanced cancer ... 39
8 Conclusion ... 40
References ... 43
Appendix ... 47
Appendix I: Articles 448-450 of Wet op de Geneeskundige Behandelingsovereenkomst (WGBO) in Department 5 of Title 7 of Book 7 of the Dutch Civil Code ... 47
Appendix II: OPTION scale ... 48
4
1 Introduction
Over the last decades, interest in the communicative interaction between doctors and patients has increased. In the fields of medicine, communication and argumentation, the focus has mostly been on shared making (henceforth: SDM) as an ideal model for decision-making in medical consultation. SDM involves at least two parties – the doctor and the patient – participating actively in the treatment decision-making process and sharing information in order to come to a mutually agreed upon treatment decision (Charles, Gafni & Whelan, 1997). Evidence indicates that SDM improves patient outcomes by promoting patient’s involvement in their own health, which results in decision-making based on medical knowledge as well as on the patient’s preferences (Van Laarhoven et al., 2014: 434).
Still, from a study by the Dutch Federation of Patients in 2016 among 4600 people it has become apparent that 40% of the patients do not feel involved (enough) in the final decision about their treatment. To get an idea of what might be the causes for this patient judgement, it is interesting to zoom in on the decision process in SDM.
In this thesis, the decision process in SDM in consultations about palliative systemic treatment for advanced cancer is studied. The main research objective of this thesis is to lay bare how the decision process in SDM takes place in actual consultations and whether this matches the expectations based on the conventions and constraints of SDM as an
argumentative activity type. In order to succeed in this endeavour, a basic prototypical argumentative pattern in SDM in consultations about palliative systemic treatment for advanced cancer will be identified and the stereotypicality of the components of this basic prototypical argumentative pattern will be tested using content analysis.1
By focussing on SDM in consultations about palliative systemic treatment for
advanced cancer, this thesis aims to contribute to argumentation theoretical research regarding medical consultation. Until now, in the research undertaken by argumentation theorists, the focus has been on medical consultation in general or on medical consultation in a general practice encounter. These studies provide us with descriptions of the role of argumentation in doctor-patient communication based on SDM in a very broad sense (Labrie & Schulz, 2015), and of how doctors and patients can manoeuvre strategically in SDM (Labrie, 2012; Snoeck
1 According to Van Eemeren (2016: 12-16) a distinction can be made between prototypical argumentative
patterns and stereotypical argumentative patterns. A prototypical argumentative pattern is a pattern that is characteristic of a particular communicative activity type because it is instrumental in realizing the institutional point of that activity type, and can be laid bare by means of qualitative research. A stereotypical argumentative pattern is a pattern that occurs frequently in a particular communicative activity type, and can only be identified by means of quantitative research.
5
Henkemans & Mohammed, 2012; Pilgram, 2015). The context of SDM in consultations about palliative systemic treatment for advanced cancer differs from the contexts studied so far mainly in the severity of the disease for which the best treatment is to be found and in the uncertain benefits and potentially high burdens of treatment options.
While current research in the field of argumentation theory – and of pragma-dialectics more specifically – focusses on the identification of prototypical argumentative patterns manifesting themselves in communicative activity types in various domains, in this thesis the stereotypicality of components of a basic prototypical argumentative pattern is quantitatively tested as well. The aim is thereby to enrich the ongoing research regarding prototypical argumentative patterns.
While this thesis is of theoretical importance to argumentation theoretical research, this research could also prove to be relevant to the medical domain and society at large as it is a first step in determining potential causes for patient dissatisfaction with regard to the
decision process in medical consultation. In this way, this thesis could provide a starting point for improving medical consultation.
In short, this thesis can be seen in terms of five research questions, the answers to which jointly contribute to the main research objective.
As to date the definition of SDM, and of its consequences for actual consultation in particular, remain unclear in important respects, the first research question is the following:
(1) How is SDM currently characterized?
To answer this question, section 2 clarifies how SDM is currently characterized and discusses the shortcomings of this characterization.
In order to provide a framework for systematic research into SDM in practice, the current characterization of SDM from a dialectical is reformulated from a pragma-dialectical perspective in section 3. The main research question addressed in this section can be formulated as follows:
(2) What does a pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer look like?
The pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer as an argumentative activity type and of its institutional point
6
and institutional constraints for critical discussion given in section 3 forms the basis for answering the following research question:
(3) What does the basic prototypical argumentative pattern in SDM in consultations about palliative systemic treatment for advanced cancer look like?
To answer this question, section 4 presents a description of the basic prototypical argumentative pattern in SDM in consultations about palliative systemic treatment for advanced cancer: a pattern based on weighing and balancing using symptomatic and
pragmatic argumentation. This basic prototypical pattern is central to the decision process in SDM in consultations about palliative systemic treatment for advanced cancer.
The theoretical analysis in sections 2-4 forms the basis for an empirical study. The main research questions addressed in this empirical study can be formulated as follows:
(4) What does the argumentative pattern in SDM in actual consultations about palliative systemic treatment for advanced cancer look like?
(5) In how far does the argumentative pattern in SDM in actual consultations about palliative systemic treatment for advanced cancer match the expectations based on the conventions and constraints of SDM as an argumentative activity type?
To answer these questions, the components of the basic prototypical argumentative pattern based on weighing and balancing using symptomatic and pragmatic argumentation will be tested qualitatively and quantitatively using content analysis. A codebook and coding sheet have been developed and used to analyse the transcripts of forty-nine SDM consultations about palliative systemic treatment for advanced cancer.
The analysis focuses on the decision process. More specifically, it focuses on the types of argumentation used and on the decision rule. After a discussion of the objectives and methodology of the content analysis in sections 5 and 6, section 7 discusses the results of the content analysis. Finally, a comparison will be made between the decision process in SDM in actual consultations and the expectations based on the conventions and constraints of SDM as an argumentative activity type.
2 Current characterization of SDM and its shortcomings
Over the last decades, interest in the communicative interaction between doctors and patients has increased. In the fields of medicine, communication and argumentation the focus has
7
mostly been on SDM as an ideal model for decision-making in medical consultation. Still, the definition of SDM, and of its consequences for actual consultation in particular, remains unclear in important respects. Section 3 of this thesis aims to provide a framework for systematic research into SDM in practice by characterizing SDM from a pragma-dialectical perspective. First it should be clarified how SDM is currently characterized and what the shortcomings of this characterization are. This will be done in section 2.1.
2.1 Current characterization of SDM in consultations about palliative systemic treatment for advanced cancer
Theorists have distinguished four prototypical treatment decision-making models based on the degree of control of the patient and the doctor (Charles, Gafni, Whelan, 1999: 653-658; Emanuel & Emanuel, 1992; Roter, 2000) (figure 1). In the paternalistic model, the doctor has high control and the patient has low control. The doctor sets the agenda for the consultation and the patients’ preferences are not actively taken into account. Consumerism forms the absolute counterpart of the paternalistic model. In consumerism, the doctor has low control and the patient has high control. The doctor’s role is limited to that of information provider. An (informed) patient sets the agenda for the consultation and takes sole responsibility for the decision. The patients’ preferences are taken into account in this treatment decision-making model. Indecision, a standstill, follows in a situation in which both doctor and patient control are low. Finally, in the SDM model, both doctor and patient control are high.
Figure 1: Doctor and patient control and prototypes of treatment decision-making. Adjusted version of figure 2 in Van Laarhoven, Henselmans & De Haes (2014: 434)
8
SDM is generally regarded as offering a potential middle choice between paternalism and consumerism (Charles et al., 1997; Labrie, 2012: 174). According to the most commonly used conceptualization, SDM involves at least two parties2 – the doctor and the patient –
participating actively in the treatment decision-making process and sharing information in order to come to a mutually agreed upon treatment decision (Charles et al., 1997). A main precondition of SDM is that at least two meaningful medically equally acceptable treatment options have to be available.3
Whereas this model considers the doctor as a medical expert possessing specialist knowledge, the patient is viewed as having knowledge about his/her own health and of his/her preferences. Even though they approach the medical consultation from very distinct angles, doctor’s and patient’s viewpoints are of equal importance.
Every doctor in the Netherlands is legally obliged to inform his patient on the nature, and the pros and cons of each qualifying treatment option4. Doctors also have to obtain the
patient’s explicit consent before starting treatment. The SDM model accounts for this legal obligation of informed consent as well as for the ethical concept of patient autonomy.
Furthermore, evidence indicates that SDM improves patient outcomes by promoting patient’s involvement in their own health, which results in decision-making based on medical
knowledge as well as on the patient’s preferences (Van Laarhoven et al., 2014: 434). Apart from these beneficial aspects of SDM, there are also several barriers to implementing this ideal in practice.
The main barriers to realizing SDM in practice can be divided over the following categories: healthcare system organisational factors, decision making interaction factors, knowledge, and power (Joseph-Williams, Elwyn & Edwards, 2014: 306) (figure 2).
In SDM in general, a main barrier is that there usually is an asymmetric relationship between the doctor and the patient (Ariss, 2009). This asymmetry arises from a factual difference in knowledge and from the fact that both doctors and patients view the doctor as having more epistemic authority. According to Bickenbach (2012: 10-12), this asymmetry may lead to doctor coercion, which could undermine respect for the patient’s autonomy, and lead to patient passiveness.
2 Often also family members or friends can play a variety of roles within (or outside) the medical encounter.
Also, several physicians often participate in the process with a single patient.
3 This precondition is taught to BA-students at Amsterdam Medical Center.
4 Articles 448-450 of Wet op de Geneeskundige Behandelingsovereenkomst (WGBO) in Department 5 of Title 7
9 Figure 2: Barriers on individual capacity to participate in SDM (Joseph-Williams et al., 2014: 306)
10
For SDM in consultations about palliative systemic treatment for advanced cancer in particular, the decision characteristics ‘disease severity’ and ‘treatment invasiveness’ may form barriers. For patients with advanced cancer survival time is very short and survival rates are zero, and treatments of these types of cancer involve uncertain benefits and potentially high burdens (Henselmans, van Laarhoven, van der Vloodt, de Haes & Smets, 2016: 2). As positive attitudes of patients and doctors toward SDM decline with disease severity (Frosch & Kaplan, 1999: 288; Pollard, Bansback & Bryan, 2015), one may expect SDM in consultations about palliative systemic treatment for advanced cancer to be limited to the minimum.
However, research has also shown that patients want to be involved in decisions about invasive, potentially risky interventions (Frosch & Kaplan, 1999: 288), and that advanced cancer patients’ preference for participation in treatment decisions increases towards the later phase of illness when life prolongation becomes more limited and quality of life becomes more important (Brom, Pasman, Widdershoven, Van der Vorst, Reijneveld, Postma & Onwuteaka-Philipsen, 2014).
2.2 Shortcomings of the current characterization of SDM in consultations about palliative systemic treatment for advanced cancer
The characterization of SDM in consultations about palliative systemic treatment for
advanced cancer as provided in section 2.1 covers general ideas about SDM in the literature from the late 90’s up to now. There are, however, inconsistencies between the various
characterizations of SDM, leading to problems for theorists and practitioners. Furthermore, it remains unclear how (facilitators and) barriers to SDM can be (institutionally) compensated within the current characterization. These shortcomings will be discussed in this section.
Inconsistency between different characterizations already becomes apparent in the basic level of assumptions and necessary conditions. There does not seem to be real consensus on, for instance, whether or not to take into account the roles of others outside the doctor-patient dyad.
Another point of disagreement is the communicative nature of SDM. Whereas some researchers claim SDM to merely allow for pure information sharing (e.g. Charles et al., 1997), others discuss the value of argumentation as an instrument in SDM processes, but do not view it as a necessary aspect of it (e.g. Rubinelli, 2013), and yet another group of authors goes even one step further by describing SDM as inherently argumentative (e.g. Goodnight, 2006; Labrie, 2012; Snoeck Henkemans & Mohammed, 2012).
11
The literature is also not clear on how the process of SDM should be conducted. Two main groups of accounts can be distinguished. The first only discusses the rules and conditions necessary for achieving the main goal of SDM. The simplicity of this type of accounts could both be an advantage – allowing for adjustment to the particular situation –and a disadvantage – a lack of guidance may impede successful execution of SDM in practice. The second group of accounts aims to translate the conceptual description of SDM into detailed models. The models proposed by these types of accounts provide doctors and patients with precise guidelines for the successful implementation of SDM. However, as each account advocates another model, doctors and patients will first have to decide on which model to use.
Furthermore, independent from which model is chosen, problems may arise when practice appears not to be completely in accordance with the situation assumed in the model.
In those cases, doctors and patients will have to decide whether or not to depart from the model. If they decide to do so, they are still left with the question of how to do this. If they decide to strictly adhere to the model, that may result in an unsatisfying decision process as the model does not fulfil the needs of the situation.
As a consequence of the inconsistencies between the various characterizations of SDM and the insufficient support of those, there is no clear framework available in which
(facilitators and) barriers to SDM in consultations about palliative systemic treatment for advanced cancer can be interpreted. Because of that, it remains unclear how to
(institutionally) compensate for barriers.
In conclusion, from my discussion it follows that the current characterization of SDM in consultations about palliative systemic treatment for advanced cancer has its shortcomings. These shortcomings can be summarized as follows:
1. There are inconsistencies between the various characterizations of SDM at the level of:
a. assumptions and necessary conditions for SDM; b. the communicative nature of SDM;
c. the practical implementation of SDM, leading to problems for theorists and practitioners.
2. Within the current characterization, it remains unclear how (facilitators and) barriers to SDM can be (institutionally) compensated.
12
Because of these shortcomings, the current characterization of SDM in consultations about palliative systemic treatment for advanced cancer does not enable us to systematically
analyse, understand, facilitate or improve medical consultation practice. In section 3, it will be discussed to what extent a pragma-dialectical approach to SDM can offer solutions for these shortcomings.
3 A pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer
To systematically analyse, understand, facilitate or improve SDM in consultations about palliative systemic treatment for advanced cancer, it first has to be established which characteristics of this specific type of medical consultation should be taken into account. In order to do this, the medical part of the pragma-dialectical theory aims to specify the characteristics of medical communication (e.g. medical consultation) as an argumentative activity type. In this research, it is established which institutional constraints may influence critical medical discussions and how they can do so.
In section 3.1, it will first be indicated how SDM in consultations about palliative systemic treatment for advanced cancer can be characterized as an argumentative activity type and as part of a critical discussion. Section 3.2 will discuss how these rules and conventions create the institutional constraints for critical discussion in SDM in consultations about palliative systemic treatment for advanced cancer. In section 3.3, a procedure for SDM in consultations about palliative systemic treatment for advanced cancer will be described. Benefits of a characterization of SDM in consultations about palliative systemic treatment for advanced cancer from a pragma-dialectical perspective will be discussed in section 3.4.
3.1 SDM in consultations about palliative systemic treatment for advanced cancer as an argumentative activity type
Medical consultation can be seen as an institutionalized communicative practice between a doctor and a patient (Pilgram, 2015: 19-24). Over the last decades, SDM has been advocated as the best model for medical consultation over treatment. The institutional point of SDM is to reach a mutually agreed upon treatment decision based on the best (medical) evidence about risks and benefits of all available options5 (including ‘doing nothing’) and on the preferences
of patients.
5 In practice, usually only those available treatment options that are relevant on the basis of the patient’s medical
13
The legal obligation of informed consent6 implies that even when the patient does not explicitly express doubt or opposition regarding the medical advice provided, the doctor should minimally assume critical doubt on the part of the patient (Labrie, 2012: 178).
According to the WGBO, the patient should be informed about the nature and the purpose of allavailable options, so that he/she can make a reasonable treatment decision. As a result, even if doctors have a strong treatment preference, they are obliged to always argue their case (Labrie, 2012: 178). In argumentation theoretical terms: the doctor should always anticipate doubt by the patient concerning what the best treatment would be. Therefore, SDM inherently involves a difference of opinion. Because of this, SDM in consultations about palliative systemic treatment for advanced cancer can be viewed as a critical discussion aimed at resolving a difference of opinion between the doctor and the patient.
To guarantee that the resolution process results in a mutually agreed upon decision based on the best (medical) evidence about risks and benefits of all available options and on the preferences of the patient, SDM in consultations about palliative systemic treatment for advanced cancer is subjected to (procedural) rules and conventions. In section 3.2, it will be indicated how these rules and conventions create the institutional constraints for critical discussion in SDM in consultations about palliative systemic treatment for advanced cancer.
3.2 Institutional constraints for SDM in consultations about palliative systemic treatment for advanced cancer
The procedure for a critical discussion in SDM in consultations about palliative systemic treatment for advanced cancer is conventionalized in order to realize the institutional point, a mutually agreed upon treatment decision. The institutional point and the conventions in SDM in consultations about palliative systemic treatment for advanced cancer constitute the
institutional constraints that define the doctor’s and the patient’s argumentative possibilities. The way in which the procedure of critical discussion is institutionalized concerns different types of conventions that are related to the focus point of the different discussion stages: the activity type’s initial situation, its starting points, the argumentative means available in the activity type and its possible outcomes (van Eemeren, 2010: 152-158).
The initial situation in SDM in consultations about palliative systemic treatment for advanced cancer is a(n assumed) difference of opinion about the best treatment option for the
6 Article 448-450 of Wet op de Geneeskundige Behandelingsovereenkomst (WGBO) in Department 5 of Title 7
14
patient. This difference of opinion theoretically could be either non-mixed (the patient doubts whether the doctor’s standpoint is acceptable7), or mixed (the patient takes an opposing stance). However, specialists from the field indicate that mixed differences of opinion are rare in this type of consultations. This is in line with Frosch and Kaplan’s (1999) observation that patients show a more compliant attitude as disease severity increases.
Starting points of SDM in consultations about palliative systemic treatment for advanced cancer can be divided into two groups: material starting points (e.g. medical knowledge as external proof) and procedural starting points. The procedural starting points can be viewed as procedural rules and conventions, such as the doctor’s obligation to obtain the patient’s explicit agreement and to do the utmost to enable the patient to take an active part in the decision process (van Laarhoven et al., 2014: 435). The procedural starting points concerning the doctor’s institutional burden of proof (Goodnight, 2006; Snoeck Henkemans & Mohammed, 2012) and the legal requirement of informed consent imply that the doctor is obliged to provide argumentation for and/or against each available treatment option. The treatment decision will then be reached through a weighing and balancing of the arguments.
The process of weighing and balancing determines the argumentative means available to doctors and patients in SDM in consultations about palliative systemic treatment for advanced cancer. The doctor and the patient will try to figure out whether a treatment option is beneficial or not with regard to the patient’s preferences on the basis of medical facts and evidence (Snoeck Henkemans & Mohammed, 2012: 21).
Once the argumentative discussion in SDM has come to an end, the discussion outcome should be an agreement between the doctor and the patient about the best treatment option. If the doctor and the patient are unable or unsuccessful in reaching a shared decision, they could agree to continue the consultation at another moment, or the patient could request a second opinion.
In sum, this section described the institutional constraints of SDM in consultations about palliative systemic treatment for advanced cancer as an argumentative activity type. The institutional point of SDM in consultations about palliative systemic treatment for advanced cancer is to reach a mutually agreed upon decision about the best treatment option for the patient. To reach such a decision, the doctor and the patient will go through a process of
7 In the context of SDM, this means that the patient doubts whether treatment option discussed by the doctor is
15
weighing and balancing based on the best (medical) evidence about risks and benefits of all available treatment options and on the preferences of the patient.
3.3 A procedure for SDM in consultations about palliative systemic treatment for advanced cancer
Based on the characterization of SDM in consultations about palliative systemic treatment for advanced cancer as a subtype of the communicative activity type of medical consultation over treatment (section 3.1-2), a procedure for this particular type of decision-making can be described.
Three key phases of the SDM process can be distinguished: (1) a preparatory phase, (2) a cooperative argumentative conversational exchange phase, and (3) a decision phase (figure 3).
Figure 3: Procedure for SDM in consultations about palliative systemic treatment for advanced cancer
The first phase of the SDM process consists of (a) preparing consultation(s) between an oncology nurse consultant and the patient.8 Firstly, this phase is aimed at enabling the patient
8 This phase is not required and (a) preparing consultation(s) between an oncology nurse consultant and the
16
to actively participate in the decision process by providing him with medical and procedural information and by making him aware of the importance of his opinion. Secondly, in this phase it is established what the patient’s pre-consultation preferences are. In other words, in the preparatory phase the starting points for the decision process are determined. As such, the first phase is comparable to the opening stage from the pragma-dialectical ideal model of a critical discussion.
The second phase of the SDM process, a cooperative argumentative conversational exchange phase, consists of four main components. The first component resembles Van Eemeren and Grootendorst’s (2004: 36-37) internal second-order condition: the mental conditions necessary for a reasonable discussion attitude. Even though this component often remains implicit, both the doctor and the patient should be willing to and be (mentally) able to reasonably participate in the decision process for SDM to make sense. The second component is the mutual exchange of information, in which the starting points are explicated – this could be done by summarizing and, if necessary or desirable, complementing the outcomes of the first phase – and the type of difference of opinion is established – the standpoints and the positions of the parties are established. The doctor’s support for the available treatment options forms the third component. In this component, the doctor provides argumentation in support of the relevant treatment options based on the interpretation of medical facts and medical evidence and the patient’s explicated preferences. In addition, the doctor could also use framing strategies and presentational devices in order to influence or steer decision-making processes. The fourth component can be referred to as patient self-advocacy and may consist of asking questions for clarification, asking critical question with regard to the
doctor’s supported suggestion(s), sharing opinions, and providing argumentation in support of the relevant treatment options. The third and fourth component mirror the argumentation stage from the pragma-dialectical ideal model of a critical discussion.
The last phase of the SDM process is the decision phase. This phase resembles the concluding stage from the pragma-dialectical ideal model of a critical discussion. In the decision phase it becomes apparent whether the difference of opinion is resolved – in that case, the discussion outcome is agreement between the doctor and patient about the best treatment option – or not – the doctor and patient are unable or unsuccessful in reaching a shared decision. In the case that no decision can be made, the doctor and patient could agree to continue the consultation at another moment, or the patient could request a second opinion.
17
treatment for advanced cancer is described on the basis of the pragma-dialectical
characterization of SDM in consultations about palliative systemic treatment for advanced cancer in sections 3.1 and 3.2. Section 3.4 discusses the benefits of a pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer.
3.4 Benefits of a pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer
A pragma-dialectical characterization of SDM in consultations about palliative systemic treatment for advanced cancer has three main advantages.
First, the characterization provided in section 3 forms a solution to the inconsistency problem (section 2). The characterization is an ideal model, a design of what the process of SDM in consultations about palliative systemic treatment for advanced cancer would be like if it were optimally conducted. Ideal models form neutral, straightforward characterizations of concepts. As such, the characterization of SDM in consultations about palliative systemic treatment for advanced cancer as an argumentative activity type is more simplified, consistent and comprehensive than previous characterizations.9
A second advantage of the pragma-dialectical characterization is that it provides more insight into the assumptions and necessary conditions for SDM in consultations about
palliative systemic treatment for advanced cancer. As it specifies obligations and rights, and procedural information for the decision process, the characterization provides doctors and patients with a clear framework for practical implementation.
Thirdly, because of its strong theoretical foundation combining insights from argumentation theory with practical insights, the provided characterization of SDM in
consultations about palliative systemic treatment for advanced cancer as a subtype of medical consultation over treatment enables us to systematically analyse, understand, facilitate or improve medical consultation practice. Within the here provided neutral framework, it can be explained more clearly why various situational aspects can function as barriers to the decision process and how they can be (institutionally) compensated. Moreover, as the characterization takes into account the contextual characteristics, it provides insight into the opportunities and constraints that SDM in consultations about palliative systemic treatment for advanced cancer
9 The first advantage not only applies to the pragma-dialectical characterization provided in section 3, but also to
other ideal models for SDM (e.g. Elwyn, Frosch, Thomson, Joseph-Williams, Lloyd, Kinnersley, Cording, Tomson, Dodd, Rollnick, Edwards & Barry, 2015; Stiggelbout, 2015).
18
offers for the contributions of the doctor and the patient.
Also, the ideal model is a tool for theorists to make predictions about (argumentative) patterns, possible strategies and other aspects of the decision process. Furthermore, it can be used as a starting point for determining the soundness criteria for the evaluation of SDM in consultations about palliative systemic treatment for advanced cancer.
In section 4, the ideal model will be used a tool to make predictions about the basic prototypical argumentative pattern in SDM in consultations about palliative systemic treatment for advanced cancer.
4 Weighing and balancing in SDM in consultations about palliative systemic treatment for advanced cancer
The main activity within the context of SDM in consultations about palliative systemic treatment for advanced cancer is the weighing and balancing of benefits and risks of the available treatment options by the doctor and the patient. This process of weighing and balancing manifests itself when dealing with the question of what is the best treatment option from various perspectives. These perspectives are related to (medical) evidence about risks and benefits and the preferences of patients.
In what follows, the basic prototypical argumentative pattern that results from the weighing and balancing of different treatment options based on different types of
considerations (i.e. medical facts and evidence, and the patient’s preferences), will be discussed.
4.1 The process of weighing and balancing in SDM in consultations about palliative systemic treatment for advanced cancer
Argumentation based on weighing and balancing is a form of complex argumentation in which judgments about the different available treatment options are weighed against each other on the basis of specific criteria.10 According to Huth (1994: 889), the process of
weighing and balancing consists of four steps: (1) a question or problem is posed;
(2) evidence for and/or against a potential answer or solution is presented;
10 The description of the basic prototypical argumentative patterns in SDM in consultations about palliative
systemic treatment for advanced cancer based on weighing and balancing provided in this section is based on Feteris (2017: 43-48). Feteris (2017: 43-48) discusses a basic prototypical argumentative pattern of weighing and balancing in the legal domain. Here, her ideas are adjusted in order to allow for an implementation within the medical domain (SDM in consultations about palliative systemic treatment for advanced cancer).
19
(3) the relative strengths of the evidence are weighed;
(4) the answer or solution most strongly supported is chosen.
This process can be translated into argumentation theoretical terms as follows: (1) the difference of opinion is established;
(2) argumentation for and/or against (a) particular standpoint(s) is put forward; (3) the arguments are weighed in accordance with a decision rule;
(4) the difference of opinion is resolved.
Applied to the case of SDM in consultations about palliative systemic treatment for advanced cancer, the decision making process will look as follows:
(1) there is a (non-)mixed difference of opinion between the doctor and the patient about the best treatment option for the patient diagnosed with advanced cancer; (2) argumentation concerning criteria (based on patient preferences) is put forward;
a. criteria (based on patient preferences) are put forward;
b. argumentation in support of the fact that the criteria are fulfilled is put forward; (3) the criteria are weighed in accordance with a decision rule;
(4) the treatment option that meets the most important criterion and is the treatment option that is most consistent with other criteria is chosen.
Through the process of weighing and balancing, the doctor and the patient can reach a shared decision – a final standpoint. In accordance with the institutional point, in which a judgment is pivotal, the main standpoint in SDM in consultations about palliative systemic treatment for advanced cancer will always be of an evaluative nature, concerning a recommended treatment option: ‘Applying treatment option X1 is desirable for this patient diagnosed with advanced
cancer’.
4.2 The basic prototypical argumentative patterns in SDM in consultations about palliative systemic treatment for advanced cancer
According to van Eemeren and Grootendorst (1992: 160), evaluative standpoints are commonly supported with symptomatic argumentation (i.e. argumentation citing a sign, symptom or distinguishing mark of what is claimed in the standpoint (van Eemeren,
Grootendorst & Snoeck Henkemans, 2002: 96-97)). Therefore, we may expect symptomatic argumentation to be prototypical in supporting the standpoint in SDM in consultations about
20
palliative systemic treatment for advanced cancer. This assumption is also empirically grounded as the results from Labrie and Schulz’s (2015) content analysis of general practice consultations point out that 86% of all arguments in their study were based on a symptomatic argument scheme.
There is, however, another argument scheme that also seems to be prototypical in supporting the standpoint in SDM in consultations about palliative systemic treatment for advanced cancer: pragmatic argumentation. In health brochures and over-the-counter medicine advertisements, pragmatic argumentation plays a central role. As these contexts as well as the context of SDM in consultations about palliative systemic treatment for advanced cancer revolve around a treatment decision, one may expect pragmatic argumentation to be prototypical in SDM in consultations about palliative systemic treatment for advanced cancer as well. That pragmatic argumentation can play a role in medical consultations is confirmed by Labrie and Schulz’s (2015) finding that 11.8% of all arguments in their study belonged to a pragmatic argument scheme11. Its percental appearance in this study, however, is considerably
lower than that of symptomatic argumentation.
In what follows, the basic prototypical argumentative patterns involving symptomatic argumentation (section 4.2.1) and pragmatic argumentation (section 4.2.2) in SDM in
consultations about palliative systemic treatment for advanced cancer will be discussed. 4.2.1 The basic prototypical argumentative pattern involving symptomatic argumentation The nature of the standpoint has consequences for its support. Following Snoeck Henkemans (2017), it can be predicted that the evaluative standpoint in SDM in consultations about palliative systemic treatment for advanced cancer will be supported by symptomatic argumentation. Therefore, this section will focus on an argumentative pattern using
symptomatic argumentation. It will be discussed to what extent such patterns are prototypical of SDM in consultations about palliative systemic treatment for advanced cancer in light of the institutional constraints of the argumentative activity type.
The general argument scheme for symptomatic argumentation is (van Eemeren et al., 2002: 97):
11 Labrie and Schulz (2015) refer to causal argumentation in general. However, from their description of the
scheme – the positive/negative consequences of accepting the advice – it becomes clear that they do not refer to the general causal argument scheme, but only to the subtype of pragmatic argumentation (a brief discussion of pragmatic argumentation is provided in section 2.2.2 of their article).
21 Y is true of X,
because Z is true of X,
and Z is symptomatic of Y.
This general scheme can be specified by pinpointing which subtype of symptomatic argumentation is used.
In SDM in consultations about palliative systemic treatment for advanced cancer, the (value) judgment in the standpoint (i.e. ‘Applying treatment option X1 is desirable for this
patient diagnosed with advanced cancer) is based on criteria relating to the patient’s preferences. The subtype of symptomatic argumentation in line with this institutional
constraint is the one in which a (value) judgment is defended by argumentation claiming that certain criteria for considering the judgment acceptable have been fulfilled. This
argumentation could be seen as a decision rule which specifies the weighing factors of the criteria (i.e. it specifies which criterion/criteria weigh(s) heavier than other criteria).12
Because of the uncertain benefits and potentially high burdens involved in palliative care, decisions about palliative systematic treatment cannot be based solely on evidence and clinical expertise but should incorporate patients’ preferences (Henselmans et al., 2016: 2). As in SDM in consultations about palliative systemic treatment for advanced cancer only
medically equally acceptable treatments can be discussed, the weighing factors (e.g. decisive factors) should be based on patient preferences. In the case of SDM in consultations about palliative systemic treatment for advanced cancer, the decision rule should take the form of a combination of the lexicographic decision rule (van der Geest, 2015: 76) and the satisfying rule (van der Geest, 2015: 76). In accordance with the lexicographic rule, the patient and the doctor decide for the treatment option which is most in accordance with the most important criterion. However, in accordance with the satisfying rule, the other criterion/criteria have to be fulfilled in a satisfying way as well. An example is a patient who wishes to achieve a maximal tumour development delaying effect (maximum life expectancy), but only without constantly being sick (sufficient quality of life). The treatment option chosen by the doctor and the patient is the treatment option offering the highest life expectancy with sufficient quality of life. This lexicographic-satisfying rule is essential in order to ensure maximal patient satisfaction with regard to the treatment decision, since other decision rules do not take into account other criteria besides the most important criterion and/or do not take into account
12 Feteris (2017: 45) discusses the decision rule in the legal context. Here, her definition is adjusted for
22 the hierarchy in the patient’s set of criteria.
The general argument scheme for symptomatic argumentation based on criteria (i.e. the decision rule) can thus be specified for SDM in consultations about palliative systemic treatment for advanced cancer as follows:
Applying treatment option X1 is desirable for this patient diagnosed
with advanced cancer,
because applying treatment option X1 meets the most important criterion and is
most consistent with other criteria,
and that applying treatment option X1 meets the most important criterion
and is most consistent with other criteria is symptomatic of the fact that applying treatment option X1 is desirable for this patient diagnosed with
advanced cancer.
Apart from this general use of the symptomatic argument scheme, a more specific use of symptomatic argumentation in a lower layer of the argumentation structure is expected to be prototypical for SDM in consultations about palliative systemic treatment for advanced cancer. The use of symptomatic argumentation is namely very frequent in medical
consultations (Labrie & Schulz, 2015). An example of symptomatic argumentation of this sort is:
Applying the treatment with Oxazepam is a responsible possibility, because there is no reason why that applying the treatment would be
problematic
and the fact that there is no reason why that applying treatment would be problematic is symptomatic of applying the treatment with Oxazepam being a responsible possibility.
In this example, the fact that there is no reason why that would be problematic, is used as a basis for concluding that applying the treatment with Oxazepam is a responsible possibility.
Symptomatic argumentation does not cover all institutional requirements for the argumentative means in SDM in consultations about palliative systemic treatment for advanced cancer. According to the legal doctrine of informed consent, doctors are legally obliged to inform the patient of all treatment options available, including the corresponding advantages (i.e. positive effects) and disadvantages (i.e. negative effects). This means that
23
benefits or risks must always be presented in the doctor’s argumentation. The basic prototypical argumentative pattern accounting for discussing these risks and benefits, pragmatic argumentation, will be discussed in section 4.2.2.
4.2.2 The basic prototypical argumentative pattern involving pragmatic argumentation As we may expect that the main reason for patients and doctors to decide for or against a particular treatment option is that its consequences are beneficial or unbeneficial in view of the patient’s preferences, it may be taken that a prototypical argument in this activity type is a pragmatic argument.
Pragmatic argumentation is a subtype of causal argumentation in which it is argued that a certain measure is (un)desirable because it will lead to a result that is (un)desirable (Schellens & Verhoeven, 1988: 86). The general argument scheme for pragmatic
argumentation is (van Eemeren, Houtlosser & Snoeck Henkemans, 2005: 211; Schellens & Verhoeven, 1988: 86):
Action X is (un)desirable
because action X will lead to result Y and result Y is (un)desirable
Pragmatic argumentation as described here can be employed in support of an evaluative standpoint and would therefore be suitable in SDM in consultations about palliative systemic treatment for advanced cancer.13 Therefore, and due to its essentiality in successfully fulfilling the institutional point in accordance with the legal obligation of informed consent, pragmatic argumentation could be regarded as a prototypical argument scheme for the standpoint in SDM in consultations about palliative systemic treatment for advanced cancer. The general argument scheme for pragmatic argumentation can thus be specified for SDM in consultations about palliative systemic treatment for advanced cancer as follows:
13 Some researchers claim that pragmatic argumentation can only be used in support of a prescriptive standpoint
(e.g. van Eemeren, 2016: 18). According to that approach, pragmatic argumentation would not be allowed in SDM in consultations about palliative systemic treatment for advanced cancer, in which an evaluative standpoint is supported. However, from the context it can be inferred that the evaluative standpoint implies a prescriptive standpoint as in SDM, a decision is made about which treatment a patient will undergo. Therefore, concluding that ‘applying treatment option X1 is the best available treatment option for this patient diagnosed with advanced
24
Applying treatment option X1 is (un)desirable
because applying treatment option X1 will lead to result Y
and result Y is (un)desirable
An example of the positive variant of pragmatic argumentation is: Applying the Folfirinox treatment is desirable,
because applying the Folfirinox treatment will most probably lead to the longest
extension of life
and the longest extension of life is desirable An example of the negative variant of pragmatic argumentation is:
Applying hyperthermia is undesirable,
because heating the tumour with all those surrounding blood vessels can lead to danger and risks
and danger and risks are undesirable
The results referred to in pragmatic argumentation can relate to long term lasting effects and to short term lasting effects. This time aspect in relation with the patient’s preferences will be taken into account in the weighing process. A patient may, for instance, either prefer a
relatively short life of a relatively high quality or a lower quality longer life.
In section 4.3, it will be indicated how the process of weighing and balancing forms the basis of the basic prototypical argumentative pattern involving symptomatic and
pragmatic argumentation in SDM in consultations about palliative systemic treatment for advanced cancer.
4.3 The basic prototypical argumentative pattern based on weighing and balancing in SDM in consultations about palliative systemic treatment for advanced cancer
The process of weighing and balancing can be translated into a pragma-dialectical
argumentation structure (figure 4) representing the basic prototypical argumentative patterns involving symptomatic and pragmatic argumentation in SDM in consultations about palliative systemic treatment for advanced cancer. Such an argumentation structure can be understood as a reconstruction of how the arguments relate to each other and to the standpoint (Van Eemeren et al., 2002: 63-78). As such, it follows a reverse pattern in comparison to the process description in section 4.1: it works down from the standpoint to the argumentation.
25
1 Applying treatment option X1 is desirable for this patient diagnosed with advanced cancer
1.1a Applying treatment option X1 meets the most important criterion and is most consistent with other criteria
1.1b Applying treatment option Xn does not meet the most important criterion and/or is not most consistent with other criteria
1.1a-b’ If applying treatment option X1 meets the most important criterion and is most
consistent with other criteria, and applying treatment option Xn does not meet the most important criterion and/or is not most consistent with other criteria, then applying treatment option X1 is desirable for this patient diagnosed with advanced cancer
1.1a.1a Applying treatment option X1 meets criterion C1
1.1a.1b C1 is the most important criterion
1.1a.1c Applying treatment option X1 meets criterion C2-Cn
1.1a.1d C2-Cn are the relevant other criteria
1.1a.1a-d’ If applying treatment option X1 meets criteria C1 and C2-Cn while C1 is the most important criterion and C2-Cn are the relevant other criteria, then applying treatment option X1 meets the most important criterion and is most consistent with other criteria
1.1b.1a Applying treatment option Xn does not meet criterion C1
1.1b.1b C1 is the most important criterion
1.1b.2a Applying treatment option Xn does not meet criterion C2-Cn
1.1b.2b C2-Cn are the relevant other criteria
1.1b.1ab -1.1b.2a-b’
If applying treatment option Xn does not meet criteria C1 and/or C2-Cn while C1 is the most important criterion and C2-Cn are the relevant other criteria, then applying treatment option X1 does not meet the most important criterion and/or is not most consistent with other criteria
Figure 4: Prototypical argumentation structure in SDM in consultations about palliative systemic treatment for advanced cancer
26
The argumentation structure representing the basic prototypical argumentative pattern in SDM in consultations about palliative systemic treatment for advanced cancer starts with the evaluative standpoint ‘Applying treatment option X1 is desirable for this patient diagnosed
with advanced cancer’14 which is supported with argumentation considering the applicability
of the lexicographic-satisfying decision rule to the relevant treatment options (1.1a-b). The argumentation concerning the lexicographic-satisfying decision rule is supported with argumentation implementing the different specific criteria and related requirements (1.1a.1a-d; 1.1b.1a-b and 1.1b.2a-b)15. These criteria are inherently completely patient based. Still, the role of the doctor is considerable with regard to these criteria as he is obliged to elicit the patient’s preferences by asking questions if necessary. The treatment-related
argumentation implementing criteria (1.1a.1a and 1.1a.1c; 1.1b.1a and 1.1b.2a) prototypically takes the form of (comparative) pragmatic argumentation referring to (un)desired results of applying a treatment option or of (comparative) symptomatic argumentation referring to (un)desired characteristics of (applying) a treatment option. The information used in these arguments is mostly derived from the doctor’s input as the doctor is the medical authority in medical consultations. Moreover, as treatment decisions in consultations about palliative systemic treatment for advanced cancer are very crucial decisions because of the uncertain benefits and potentially high burdens of the treatment options (Henselmans et al., 2016: 2), it is of importance that the criteria are correctly understood and combined to construct an informed well-considered treatment decision. The doctor can help in this respect by asking questions.
Based on the specification of criteria and related requirements, the process of weighing and balancing will result in a particular evaluative standpoint.
5 Objectives of the content analysis
Sections 2-4 provide a point of departure for the content analysis with the twofold aim of (1) testing whether the components of the basic prototypical argumentative pattern I have identified are stereotypical in the context of decisions about palliative systemic treatment for advanced cancer, and (2) evaluating the decision process in SDM in consultations about palliative systemic treatment for advanced cancer by comparing the ideal with actual consultations.
14 ‘Doing nothing’ is also regarded as a treatment option.
15 These criteria can all be grouped under the two main criteria ‘life expectancy related to the treatment’ and
27
The basic prototypical argumentative pattern in decisions about palliative systemic treatment for advanced cancer consists of four main components: the discussion of criteria, weighing and balancing, the use of symptomatic and pragmatic argumentation, and the decision rule. The discussion of criteria in SDM in consultations about palliative systemic treatment for advanced cancer (component 1) has been the topic of investigation in a study by Henselmans et al. (2016). The aim of their research was to examine the expression of values, appraisals and preferences (e.g. criteria) by patients, as well as the oncologists’ communicative behaviour facilitating these expressions. Henselmans et al. (2016) found that criteria were discussed in the consultations. As I will use a subsample of the sample used by Henselmans et al. (2016) in this study, I can regard the presence of the first component of the pattern
ensured. Only those consultations that went in accordance with the principles of SDM according to the OPTION scale16 (Elwyn et al., 2003) were included in the present study. By
using the OPTION scale, the presence of the second component of the basic prototypical argumentative pattern, weighing and balancing, has been ensured17.
The third and fourth component of the basic prototypical argumentative pattern in decisions about palliative systemic treatment for advanced cancer remain untested so far. These components will be the focus of this study. Two questions are addressed:
1. Which arguments do medical oncologists and patients use when decisions about palliative systemic treatment for advanced cancer are made?
a. Which arguments do medical oncologists and patients use when decisions about palliative systemic treatment for advanced cancer are made in first consultations?
b. Which arguments do medical oncologists and patients use when decisions about palliative systemic treatment for advanced cancer are made in follow-up consultations?
c. How do first consultations and follow-up consultations compare with regard to the quantitative use of argumentation?
2. Which decision rule do medical oncologists and patients use when decisions about palliative systemic treatment for advanced cancer are made?
16 Appendix II
17 Weighing and balancing is covered by point 2 and points 4-7 in the OPTION-scale: multiple treatment options
28
With regard to the first question, I will focus on three (sub)types of argumentation: symptomatic argumentation, pragmatic argumentation and authority argumentation. Symptomatic argumentation and pragmatic argumentation have been discussed in section 4.2.1 and 4.2.2 respectively. Authority argumentation is a subtype of the main type of symptomatic argumentation (Pilgram, 2015: 35). For authority argumentation, the main scheme of symptomatic argumentation can be specified by regarding the authority’s opinion as a sign of the acceptability of the standpoint. The authority referred to in the argumentation can be an external source’s authority (argumentation from authority) or a discussion party’s own authority (argumentation by authority). The general argument scheme for authority argumentation is (an adjustment of Pilgram, 2015: 35):
Y is true of X,
because authority A is of the opinion that X
and A’s opinion is symptomatic (or indicative) of Y An example18 of authority argumentation is:
You should undergo an additional blood test, because the lab advises that,
and the lab’s advice indicates that you should undergo an additional blood test
In this example, the fact that the lab advises the patient to undergo an additional blood test, is used as a basis for concluding that the patient should undergo an additional blood test.
I will focus on symptomatic argumentation and pragmatic argumentation because these argument schemes seem to be prototypical in supporting the standpoint in SDM in consultations about palliative systemic treatment for advanced cancer (see: section 4.2). This study will test the stereotypicality of these argument schemes. If 75% or more of the
arguments advanced in the consultations pertain to symptomatic and pragmatic
argumentation, the argument schemes will be considered stereotypical for this context. Recently, interest in authority argumentation within the medical domain has increased (e.g. Pilgram, 2015; Snoeck Henkemans & Wagemans, 2012; Wierda, 2015). This increase seems to be a result of the paternalistic medical tradition shifting towards a more participation oriented approach, which has led to authority in medical consultation being perceived in
29
another light. Even though one would not expect authority argumentation to play a significant role in shared decision-making as this context disvalues doctor authority, the number of studies about authority argumentation within the medical domain suggests that authority argumentation is stereotypical in these contexts. This assumption has, however, never been supported quantitatively. Authority argumentation is included in this study to test its assumed stereotypicality.
The sub questions of the first question involve a comparison between first
consultations and follow-up consultations. Commonly, in first consultations more standpoints are discussed because more relevant treatment options are available. Furthermore, in first consultations the patient usually is less familiar with treatment goals and side-effects than in follow-up consultations. One could imagine that these consultation characteristics could affect the amount of argumentation and the type of argumentation.
The second question addresses the decision rule – the fourth component of the basic prototypical argumentative pattern. If 75% or more of the decisions are made in accordance with the decision rule, the decision rule will be considered stereotypical for this context.
6 Methodology 6.1 Content analysis
Content analysis is a standard method for the systematic and quantitative analysis of messages in their communicative context (Labrie & Schulz, 2015: 35; Mohammed & Schulz, 2012: 257; Neuendorf, 2002). In that sense, content analysis relies on the scientific method and pays attention to a priori design, hypothesis testing, objectivity-intersubjectivity, reliability,
validity, generalizability and replicability (Neuendorf, 2002: 10).
6.2 Sample
The sample used in this study was taken from an existing data set of dr. Inge Henselmans and prof. dr. Hanneke van Laarhoven. They described their data set as follows: “All medical oncologists, staff and in training, at the Academic Medical Centre Amsterdam were eligible. Patients were eligible if they: (1) met with the oncologist to discuss either the start of (a new line) of chemotherapy or the (dis)continuation of current chemotherapy, (2) had a median life expectancy of <1 year without chemotherapy and (3) if chemotherapy offered a median survival benefit of ⩽6 months. These included (1) patients with metastasised or inoperable tumours of the pancreas, oesophagus, stomach, liver, gall bladder, bladder or sarcoma
30
scheduled for an initial or an evaluative follow-up visit including the discussion of computed tomography (CT)/positron emission topography (PET)-CT results; or (2) patients with any type of cancer under an additional line of palliative chemotherapy (>first line) scheduled for an evaluative follow-up visit. We aimed to include >40 patients, of whom at least 20 were scheduled for an initial visit, and to audio-record two consultations per patient.” (Henselmans et al., 2016: 2-3).
For this study, consultations were excluded if they did not meet the criteria for SDM in consultations about palliative systemic treatment for advanced cancer. The OPTION scale (Elwyn et al., 2003) has been used to measure whether the consultations went in accordance with the principles of SDM.
6.3 Procedures
The study consisted of two phases: (1) a development phase in which the coding instruments
were developed and (2) a coding phase.
The instruments developed in the first phase are a codebook19 and a coding sheet. In the codebook, all variables are specified and coding instructions are provided. The coding sheet is a tool in which one can keep record of all codes. The development phase consisted of three stages. In the first stage, a literature study was done in order to root the instruments in theory. The theory was then tested in a qualitative study of six subsamples from the sample. The second stage was meant to increase the validity and reliability of the study. In this stage, two coders – one of whom is the principal investigator – took part in a training session. The completeness and clarity of the codebook and coding sheet were elaborately discussed by the two coders and the coding instruments were adapted where necessary. The third stage was a means to familiarize the coders with the coding procedure and measures, and to improve the coding instruments. In a pilot study, both coders coded five transcripts individually and the results were compared. Variables were reformulated, or even removed from the codebook if necessary on the basis of careful examination and discussion of the results by the coders.
In the coding phase, the first coder coded forty-four transcripts using the codebook and coding sheet. The second coder coded a random subset of five transcripts to allow for a
reliability calculation for each of the variables under study upon completion of the data
19 Appendix III
31
collection.20 Data analysis was performed using the statistical package SPSS Statistics 24.
6.4 Measures
The codebook contains four types of variables: consultation level variables, treatment level variables, argumentative statement level variables, and decision level variables. The
consultation level variables are formal and technical variables concerning the transcripts and coding itself, coder initials, transcript name and date coded, and were included mainly for the sake of coder and consultation identification.
The treatment level variables concern the specification of the treatment options discussed in a consultation and a short description of the characteristics of the situation in each consultation.
The argumentative statement level variables concern the argumentative statements of the doctor and the patient throughout the consultation. Coders were instructed to code the different argument schemes by highlighting the parts of the transcripts in which an argument relevant to the treatment options discussed is put forward with the right colour. Also, the coders were instructed to indicate the number of arguments of a certain type in the coding sheet. The categories of arguments the coders could choose from are symptomatic
argumentation (with the exception of authority argumentation), authority argumentation, pragmatic argumentation, and ‘other’ argumentation.
The decision level variables concern the decision rule used in the consultation. The coders were instructed to indicate whether a decision was made, and, if so, which treatment option was chosen. Furthermore, the coders were instructed to indicate whether (they thought that21) the treatment option chosen was the treatment option that met the most important
criterion and whether it was the treatment option that was most consistent with other criteria.
6.5 Validity
In order to assure the validity and reliability, attention was paid to various types of validity. First, the development of the codebook was strongly rooted in theory. The pragma-dialectical theory and analyses in accordance with that theory formed the starting point for the development of the measures (see: sections 2-4). Aiming to do justice to the inherent
dialogical aspect of an argumentative discussion, and to the collaborative aspect of SDM in
20 To determine the inter-rater reliability, a random sub-sample of at least 10% should be taken (Neuendorf,
2002: 158-159).
