“I need to know what makes somebody tick …”: Challenges and Strategies of Implementing Shared Decision‐Making in Individualized Oncology

Although shared decision making is advocated in most fields of medicine, implementation in oncology is challenging. Considering the importance of patient involvement in the era of individualized oncology, this article analyzes barriers to shared decision making.

Abstract

Background.

Shared decision‐making (SDM) has been advocated as an ethical framework for decision‐making in cancer care. According to SDM, patients make decisions in light of their values and based on the available evidence. However, SDM is difficult to implement in cancer care. A lack of applicability in practice is often reported. This empirical‐ethical study explores factors potentially relevant to current difficulties in translating the concept of SDM into clinical practice.

Methods.

This study was conducted with nonparticipant observation of the decision‐making process in patients with gastrointestinal cancers for whom the benefit of adjuvant chemotherapy was uncertain according to clinical guidelines. Triangulation of qualitative data analysis was conducted by means of semistructured interviews subsequent to the observation. Observation notes and interview transcripts were analyzed according to the principles of grounded theory.

Results.

Deviating from the concept of SDM, oncologists initiated a process of eliciting values and medical information prior to conveying information. The purpose of this approach was to select and individualize information relevant to the treatment decision. In doing so, the oncologists observed used two strategies: “biographical communication” and a “metacommunicative approach.” Both strategies could be shown to be effective or to fail depending on patients’ characteristics such as their view of the physicians’ role and the relevance of value‐related information for medical decision‐making.

Conclusion.

In contrast to the conceptual account of SDM, oncologists are in need of patient‐related information prior to conveying information. Both strategies observed to elicit such information are in principle justifiable but need to be adapted in accordance with patient preferences and decision‐making styles.

Implications for Practice.

This study showed that knowledge of patients’ values and preferences is very important to properly adapt the giving of medical information and to further the process of shared decision‐making. Shared decision‐making (SDM) trainings should consider different strategies of talking about values. The right strategy depends largely on the patient's preferences in communication. To be aware of the role of values in SDM and to be able to switch communicative strategies might prove to be of particular value. A more systematic evaluation of the patient's decision‐making preferences as part of routine procedures in hospitals might help to reduce value‐related barriers in communication.

Introduction

“Individualized oncology” promises to deploy patients’ biological characteristics for an optimal treatment approach [1], [2], [3], [4]. Nevertheless, because of its inherently probability‐based nature, individualized decisions of patients and physicians are often decisions under uncertainty [2]. In other words, even with improved knowledge about biomarkers and evidence‐based guidelines, there are many treatment decisions in which there is uncertainty about the “medically best option.” In such situations, information about patients’ lifestyle and goals has been emphasized as particularly important for making good treatment decisions [5], [6].

The concept of “shared decision‐making” (SDM) has been advocated as an ethical framework, in particular for decisions under uncertainty about the medically best treatment. SDM is defined by three conceptual dimensions. Firstly, it describes a process of deliberation [7], [8] between at least two participants [9], [10] aiming at mutual agreement regarding treatment decisions [7]. Secondly, it has to be understood as a moral requirement stemming from the acceptance of patients’ basic rights such as autonomy and self‐determination [11], [12] and the acceptance of different realms of expertise. According to their rights and knowledge of aims, values, and preferences, patients may choose their preferred degree of involvement as well as their preferred amount of medical information. Physicians, in light of their expertise (i.e., knowledge and practical experience) are able to assess the medically available options. The aforementioned two characteristics, thirdly, assign certain roles and responsibilities to physicians and patients to make SDM happen [9]: Physicians have the task of bringing in their knowledge and experience to provide patients with diagnosis and treatment options. Patients have to articulate aims, values, and preferences to provide physicians with information on their preferred degree of involvement and to evaluate alternatives [13].

Shared decision‐making has been recommended as an approach to increase the chance that patients’ values and preferences are reflected in treatment decision‐making [14]. However, despite its widely acknowledged importance, current research indicates considerable differences between preferred and perceived levels of implementation. Physicians, when asked about their attitude toward SDM, strongly support the involvement of patients [5], [14], [15], [16]. This holds especially true when compelling evidence is missing and physicians do not feel the strong need to recommend a certain treatment option [5], [17], [18]. However, the reported support for SDM does not correspond with its level of practical implementation [14], [19], [20], [21], [22]. Physician‐reported barriers are time constraints, a lack of agreement with certain key elements of SDM, and a lack of practical applicability because of patient characteristics or the nature of the medical situation [21], [23], [24]. There is also evidence for a supportive attitude toward SDM on the side of the patients [25], [26], [27]. However, patients tend to underestimate their contribution in the process and do not often recognize the importance of their personal preferences [24], [25]. Moreover, some refrain from participation because of feelings of incapacity regarding the complexity of medical information or presumptions about “normal” patient behavior [25].

In summary, the aforementioned findings indicate that although SDM is often advocated, its implementation in oncology (as well as in other fields of medicine) is still challenging [24], [28]. Given the importance of patient involvement in individualized oncology, this qualitative empirical‐ethical study analyzes barriers to implementing SDM. The study focuses on the paradigm situation in which, according to current guidance, there is uncertainty regarding the benefit of tumor‐specific treatment. In the first step, we observed the challenges in implementing SDM and strategies used by oncologists to overcome these barriers. In the second step, we analyzed the data with a focus on ethically justifiable strategies for supporting the implementation of SDM in oncology.

Materials and Methods

Data Collection

A qualitative research approach that triangulated different data sources and methods of collection was chosen to reconstruct the challenges of SDM [29]. Data were collected from open nonparticipant observation of decision‐making in patients with gastrointestinal tumors for whom adjuvant chemotherapy was optional according to current guidelines. Observational research assumes behavior to be purposeful and to reflect deeper values and preferences [30]. The term “nonparticipant” denotes a style of observation that aims to minimize the researchers’ intervention in the observed situation. Researchers therefore took part in meetings between patients and physicians in which the further treatment course was discussed. However, they did not interfere with the communication between both parties. We hypothesized that this patient group might be typical regarding uncertainty in times of individualized oncology. Many gastrointestinal tumors are well known for their heterogeneity on the molecular level and their complex interplay of different factors [31]. Therefore, the treatment of gastrointestinal tumors by means of personalized medicine is a worthwhile strategy but nevertheless connected to uncertainty regarding treatment and prognosis. In addition, the population inherits a certain variability in relevant sociodemographic and medical factors. Our study population included cases with different diagnosis as well as curative and palliative cases from different sociocultural backgrounds. Those differences made it possible to observe a maximal bandwidth of different cases and decision settings. Because of the qualitative research method and sample size it is, however, not possible to compare the groups involved among each other for differences (e.g., palliative vs. curative decisions). Observations were conducted in an oncology department of a German university hospital. Considering the interactive and bilateral nature of decision‐making processes, semistructured interviews with patients were conducted subsequent to the observation. In addition, if hospital routines allowed, we also interviewed oncologists directly after the consultation.

Regarding nonparticipant observation, the conducting researcher (J.H.) followed an observation scheme that gave special attention to communicative breaches during the process of decision‐making, hypothesizing that disruptions and incongruity of the process might be a good indicator of value‐related challenges [32]. The observer took notes of the consultation and, when possible, of the oncologist's impression subsequent to the consultation. The notes were transcribed into a predefined observation protocol on the day of the observation. Quotation marks indicate literally transcribed passages.

Semistructured interviews with patients were recorded and transcribed by the researchers. The interviews followed a guideline that was structured in three main parts. The first part asked for biographical information and was used to enter the interview and to deepen the understanding of health‐related values and preferences within their biographical context. The second part addressed issues concerning the consultations observed directly. It was used mainly to understand the patient's experience of decision‐making and gave the interviewer the opportunity to ask for explanations or validation of details. The last part was concerned with the patient's assessment of the decision‐making process and his or her overall satisfaction with the physician and the decision made.

The first observations were carried out by two authors (J.H., S.W.). All the following observations and interviews were carried out by the first author. Cases to observe were sampled theoretically according to the principles of grounded theory methodology to maximize the opportunity of discovering varieties in the phenomena observed and to densify the categories to develop [33]. Participants for interviews were sampled purposively according to a criterion of maximal variation in the phenomenon. Physicians and patients were informed about the study and gave their consent. The study was approved by the ethics committee of the Ruhr University (15‐5310) Bochum.

Data Analysis

Transcribed protocols and interviews were analyzed according to the principles of grounded theory, following an iterative process of data analysis and data gathering. The use of grounded theories has proven to be of advantage when new fields or social processes are explored and basic theories are created. Their purpose is to gain in‐depth understanding of phenomena in contexts that have attracted little attention so far and were therefore deemed suitable for our field of interest [34]. Preliminary categories were constructed based on the first interviews and observations that were independently coded by two of the researchers (J.H., S.W.). These served as a basis for the analysis of the prevailing material. In the end, the categories developed were summarized through similarity or interrelatedness into more general categories. The main themes of the ongoing analysis and emerging codes were discussed within the research group and were then presented to further experts from the fields of oncology, medical ethics, philosophy, and social sciences. Data analysis was ended subsequent to theoretical saturation, the moment at which alterations of existing codes and change of interpretation by new observations or interviews were judged by the research team to be unlikely.

Results

Patients were asked for permission to observe their consultation on 20 occasions. A total of 18 observations with five different oncologists were finally conducted. Two patients (P‐003; P‐006) were excluded from the study because of their refusal to give consent. The consultations each took an average of 20 minutes (ranging from 5 to 55 minutes). Physicians gave additional impressions regarding the decision‐making process subsequent to the consultation on four occasions. In addition, we conducted six in‐depth interviews with patients that were conducted in a period of 3 days to 3 weeks after the consultation. The interviews lasted an average of 55 minutes (ranging from 25 to 70 minutes). Table 1 gives an overview of the characteristics of the observations and interviews.

Table 1. Settings and patients.

General Structure of the Decision‐Making Process

Analysis of the observational data revealed a typical structure of a three‐phase decision‐making process. The starting point was an exchange about the patient's perception of his or her health condition and the development of the disease. This narrative of the patient at this stage was often connected to events in his or her biography. In the second phase, the decision‐making process went through an interim conclusion by which physicians tried to sum up the information given in a short and precise manner. The third phase started with the oncologists elaborating relevant medical information of the patient's case, followed by the pros and cons of the treatment options given—that is, whether or not to administer tumor‐specific treatment. Although the patient's narration mainly shaped the first phase, oncologists took the lead in this last phase by presenting and explaining treatment options, their risks and harms as well as possible benefits. The decision‐making process ended in mutual agreement regarding the next steps. A model of this structure can be seen in Figure 1.

Figure 1.

Structure of decision‐making observed; structural outline of the structured decision‐making process process.

Differing Interests of Participants and Mutual Inhibition of Information Transfer

Observation, especially at the beginning of the consultation, revealed the different interests of the participants shaping the communication. Patients’ questions and demands for information displayed in the conversation were mainly directed to medical information, that is, information on their condition from a medical point of view, possible treatments, their alternatives, and their consequences. When one of the patients, for example, was told that a decision regarding the option of an adjuvant chemotherapy needed to be made and was then asked how he felt, he replied

that he never actually thought about such a situation in concrete terms, but rather in a way that there is such a possibility that needs to be considered. “But okay then. Then I need some information now … length, which drug, how many cycles and the like …” Observation Protocol, P012

Patients seemed to be considerably less interested or did not have the ability to share information on their values and preferences, defined here as assignment of interest depending either on one's moral beliefs (values) or one's personal aims and goals in a certain context (preferences). They felt that physicians did not belong to a sphere that is concerned with or amenable to discussing personal preferences or sharing value judgments. Interviewees, thus, expressed the opinion that value‐related considerations were issues suitable rather for the inner circle of family and friends. Asked what she would discuss with her oncologist in decision situations and if values and preferences might be a part of it, one patient denied that she would discuss anything and explained:

“Because, I mean … she [the physician] is just there concerning to my body. My … Me … What is regarding the soul, or the like [snorting], I don't know …” Interview with P002

Physicians, on the other hand, were highly interested in value‐related considerations of the patients and their preferred decision‐making style. One reason offered for this interest was the relevance of such information to individualizing the information conveyed to the patients. One oncologist, for example, giving her impressions of a consultation, stated:

“I need to know what makes somebody tick!” The oncologist describes this as the most important part of the decision‐making. After that, she says, it is about medical facts, advantages, and disadvantages. “Finally, it is important that people are okay with their decision. And that, after all, seems to depend on their character.” Observation Protocol – P002

Our observations can be explained as the result of a different distribution of knowledge of medical information and values at the beginning of the conversation. Patients have access to the necessary value‐related content for the decision‐making process while they are mainly interested in medical information. Oncologists, on the other hand, have access to those facts but need to learn more about the patient's values and preferences to properly adapt the giving of information. In cases of disagreement concerning the preferred decision‐making style or underlying values, or in cases of unawareness or inability to communicate values and preferences, the distribution and different amount of knowledge leads to a mutual inhibition of information transfer at the beginning of the process that needs to be overcome.

Strategies to Overcome the Informational Imbalance

We identified two different strategies that the participating oncologists used to gain access to values and preferences relevant to the decision at stake in order to overcome the mutual inhibition of the dialogue partners and to further the necessary informational transfer. We call them the biographical approach and metacommunicative approach.

The biographical approach attaches more weight to the first structural part of the consultation exploring the patient's preferences and value decisions through his or her narrative of the disease's history. In such cases, oncologists actively strived to initiate a narration of the patient's experience and gave special attention to those parts of the stories that were connected to former contact with medical professionals or former decision‐making situations.

“So then, tell me,” she begins anew, “What brought you here? Are you from the city of [X]?” The patient replies she lives in [Y]. “Ah, okay,” the oncologist answers, mildly surprised. “And some ten years back or so you already had to deal with issues of cancer, right?” The patient confirms and starts telling her story: How she underwent surgery and spent over four months in hospital because of a wound infection. The oncologist gasps and displays sympathy. “For God's sake! How did you realize you were in a bad condition?” Now, the patient starts with a long explanation, carefully connecting the stages of her course of illness. Observation Protocol – P011

The patients’ endeavor to provide a consistent and comprehensible picture of the events led them to explicate certain underlying preferences and values to which they gave importance and, thus, informed the oncologists about crucial value‐related aspects.

Using the metacommunicative approach, on the other hand, oncologists addressed the situation of the decision‐making itself and tried to evoke a certain understanding regarding the need to talk about values and preferences as well as medical facts. Typical issues brought up were the physicians’ own interest in the situation. One oncologist, for example, tried to explain her repeated attempts to gather information on the patient's preferences by explaining her intentions:

After a short pause, the oncologist again takes the lead: “On the other hand, I do not want to scare you, but it may well be that in one or two years something [the tumor] comes back again. This is why we are talking.” She then explains that it is not her intention to talk the patient into something. “You can do it this or that way […].” Observation Protocol – P004

In addition, physicians talked about the problems of uncertainty or missing data to give the patient an impression of the importance of their value consideration. One oncologist, for example, directed attention to uncertainty issues by telling her patient that

“Therapy can reduce the likelihood of future illness,” the oncologist explains. She emphasizes “can.” “[…] In medicine, one often looks at great numbers over a period of five years,” she continues. “One builds classes before and after to determine who is still alive. Anyhow, these numbers between the groups show differences between three and five percent in your case. That means it is not very much. It is neither white nor black. If you refuse to have chemotherapy, that does not mean that you will not get cancer again. However, if you do have therapy that does not mean you're free from cancer, either. One simply does not know exactly.” Observation Protocol – P014

Discussion

This empirical‐ethical study provides a combined conceptual and empirical analysis of barriers to SDM and possible strategies to overcome them in situations in which patients and oncologist have to make a decision about tumor‐specific treatment of which the benefit, according to current guidelines, is uncertain. The first main finding that our data suggest is that a patient's values and preferences are not only important for the actual decision‐making—as this is usually conceptualized in SDM—but rather that such information is already an important prerequisite for oncologists to individualize the information given to the patient. The second main finding is the two strategies – the biographical and the metacommunicative approaches – that were used to elicit the patient's values and preferences.

Regarding the first finding, it is notable that conceptual analysis of SDM usually focuses on the question of how information about values and preferences can be integrated into decision‐making next to the information of medical facts. Such a conceptual account implies that all necessary medical information has been exchanged independently from information about individual preferences regarding the treatment. However, our empirical analysis suggests that such strict distinction between fact and value is not feasible in clinical practice. Values and preferences already seem relevant to the process of conveying medical information. One reason for this may be that oncologists need to have this information to be able to adapt their provision of information in accordance with the patient's interests. Given the increasing amount and complexity of medical information in oncology, it seems plausible that oncologists do not perceive it to be appropriate to convey all the medical information but rather select pieces of information according to their relevance for the individual situation.

Regarding the second main finding—the biographical and metacommunicative approaches to overcome the informational imbalance—we have observed that both can be successful in eliciting preferences and values, but both also yield certain pitfalls from an empirical‐ethics perspective. Biographical communication strategies attempt to reconstruct the patient's preferences regarding the degree of involvement in decisions from the patient's biographical narrative. However, patients were neither aware nor informed about the importance oncologists gave to their implicit value considerations. In this sense, the approach runs a risk of nontransparent interpretations on the side of the physician. One reason for a patient's missing awareness of the aims of oncologists might be that patients assign their doctors to a medical sphere that is primarily concerned with facts and information regarding the body and not to a social setting, where values and preferences need to be discussed. Quite similar observations have been made elsewhere [35]. Secondly, the reconstruction of implicit preference and value judgments requires one to distinguish clearly between one's own judgments and those provided by the patients. The risk of conflating perspectives and judgments, especially for the less experienced, and hence of attributing one's own preferences to the patient, seems to pose a serious threat for the quality of SDM. In addition, it is known from biographical interview research that in settings where patients strive to give a consistent and comprehensive picture of former decisions and behavior, they tend to picture themselves not as they actually behaved, but rather as they would have liked to have be seen by others [36]. This implies the risk of misunderstanding preferences and values that enter the decision‐making process.

The metacommunicative approach often displays a surface appearance of rationally weighing pros and cons and talking explicitly about one's own interests and perspectives. It might be more transparent, but, simultaneously, it conceals the risk of overwhelming the patient. Metacommunicative approaches require the patients to develop a clear and concise perspective regarding their own underlying preferences and value assumptions and the ability to express these considerations. However, not all patients might understand the need to talk about values or be able to express their thoughts. More importantly, especially in difficult situations in which decisions need to be made, the demand to verbalize value considerations that are connected with deep‐rooted concerns, for example, the fear of dying, might result in harsh emotional reactions instead of clear and satisfying decisions.

Our data suggest that the success of both communicative approaches depend on the patient's preferred decision‐making style and the skills and abilities of the oncologists. From an ethical perspective, it is important to note that both approaches have their advantages and can lead to normatively acceptable results. Their adequacy may, therefore, be assessed not against any external standard but rather in relation to the patient's needs and preferences and the oncologist's skills.

Strengths and Limitations

Qualitative research offers the opportunity to explore complex issues, such as underlying value assumptions, communicative strategies, and structures, within their native social horizon. Together with an explicit normative framework, it opens the opportunity to generate empirically based and generalizable hypotheses about hurdles and obstacles in the implementation of normatively required concepts such as SDM. However, because of the methodological design, the results gained in this study are not representative of all oncologists. Furthermore, as is the case with all studies of this kind, the effect of social desirability has to be taken into account. Other limits that need to be considered are that of the single‐center study as well as the focus on patients with gastrointestinal cancer and the differing treatment contexts (i.e., differing prognosis of patients with colon cancer compared with pancreatic cancer in this sample).

Conclusion

Even in the light of the limits above, our findings warrant further research into possible barriers to SDM and problem‐solving strategies. The findings of our study suggest a critical reevaluation of SDM concepts regarding the inherent interdependencies of facts and values. Furthermore, there are possible implications for practice. Firstly, SDM training should consider the different strategies of talking about values before conveying medical information. Because the right strategy depends largely on the patient's preferences in communication, to be aware of and to be able to switch communicative strategies might prove to be of particular value. Secondly, a more systematic evaluation of the patient's decision‐making preferences as part of the routine procedures in hospitals might help to reduce the informational imbalance at the beginning of the process. Eliciting the patient's preferences for autonomy or their preferred decision‐making style, for example, might be an easy way to address those issues. Such investigation into value‐related aspects might turn patients’ and physicians’ attention to the importance of values. This could blaze the trail for the insight that they are an integral part of every decision‐making process.

Author Contributions

Conception/design: Joschka Haltaufderheide, Jan Schildmann

Provision of study material or patients: Bernhard Bertlich, Anke Reinacher‐Schick

Collection and assembly of data: Joschka Haltaufderheide, Sebastian Wäscher

Manuscript writing: Joschka Haltaufderheide, Sebastian Wäscher, Bernhard Bertlich, Jochen Vollmann, Anke Reinacher‐Schick, Jan Schildmann

Final approval of manuscript: Joschka Haltaufderheide, Sebastian Wäscher, Bernhard Bertlich, Jochen Vollmann, Anke Reinacher‐Schick, Jan Schildmann

Disclosures

The authors indicated no financial relationships.