Trust and shared decision‐making among individuals with multiple myeloma: A qualitative study

Abstract Background Multiple myeloma (MM) is an incurable cancer with complex treatment options. Trusting patient–clinician relationships are essential to promote effective shared decision‐making that aligns best clinical practices with patient values and preferences. This study sought to shed light on the development of trust between MM patients and clinicians. Methods Nineteen individual semi‐structured interviews were conducted with MM patients within 2 years of initial diagnosis or relapse for this qualitative study. Interviews were recorded and transcripts were coded thematically. Results We identified three main themes: (1) externally validated trust describes patients’ predisposition to trust or distrust clinicians based on factors outside of patient–clinician interactions; (2) internally validated trust describes how patients develop trust based on interactions with specific clinicians. Internally validated trust is driven primarily by clinician communication practices that demonstrate competence, responsiveness, listening, honesty, and empathy; and (3) trust in relation to shared decision‐making describes how patients relate the feeling of trust, or lack thereof, to the process of shared decision‐making. Conclusion Many factors contribute to the development of trust between MM patients and clinicians. While some are outside of clinicians’ control, others derive from clinician behaviors and interpersonal communication skills. These findings suggest the possibility that trust can be enhanced through communication training or shared decision‐making tools that emphasize relational communication. Given the important role trust plays in shared decision‐making, clinicians working with MM patients should prioritize establishing positive, trusting relationships.


| BACKGROUND
Multiple myeloma (MM) is an uncommon hematologic malignancy that primarily affects older adults, with around 35,000 new cases diagnosed annually in the United States. 1,2 Considered incurable, MM follows a relapsing and remitting course, where each period of remission is expected to become progressively shorter until treatment eventually fails. 3 Management of MM involves a series of complex treatment decisions across the disease trajectory, with the goal of both prolonging life and maximizing quality of life. 4 Many novel and effective therapies have been introduced in recent years, yielding multiple treatment options to address each relapse. However, responses to and side effects from treatments vary greatly, and few clear guidelines exist on which to base treatment decisions beyond first-line therapy. 3,5 Further complicating treatment decisions, many therapies can be burdensome to patients in terms of their physical effects, financial impact, and frequency of administration or monitoring requirements. 5,6 Consequently, both clinical acumen and clear insight into patients' goals, values, and preferences are needed to choose wisely.
Shared decision-making-a process in which patients and clinicians partner to make treatment choices that optimize clinical outcomes while honoring individual values and preferences-is essential in circumstances like these, where there is limited or conflicting evidence to support which treatment is best. 7,8 While there is growing consensus that shared decision-making is necessary to patientcentered care for individuals with serious and life-limiting illness, 9 the best strategies to achieve shared decisionmaking in practice remain elusive. 10,11 For effective shared decision-making to occur, clinicians must be able to understand and effectively communicate the risks and benefits associated with various treatment options, while patients must feel empowered to communicate their goals and concerns. 12 Trust between patients and clinicians is essential in this context.
Trust can be conceptualized both in terms of general trust in the medical profession or institutions as a whole, and as relating to specific interpersonal relationships between an individual clinician and patient. 13 Importantly, interpersonal patient trust in clinicians has been found to influence decision-making preferences and behaviors. In previous work among individuals with chronic conditions, lower level of trust in the clinician was independently associated with lower quality of shared decision-making communication. 14 Among hospital inpatients, trust in the clinician has been associated with patient-reported shared decision-making behavior. 15 Similarly, among patients admitted to intensive care, trust in clinicians was correlated with dimensions of patient-reported psychosocial care, including clinician shared decision-making behaviors. 16 In other work among individuals with breast cancer, however, trust in the oncologist was not associated with either patient-reported or observer rated shared decisionmaking. 17 These disparate findings suggest that the relationship between trust and shared decision-making is complex and challenging to quantify. Given that most studies of trust and shared decision-making have been cross-sectional in nature, the direction of the relationship between trust and shared decision-making remains unclear. Does trust lead to better shared decision-making, does shared decision-making behavior promote trust, or is there a bi-directional relationship? A recently developed model of relational trust in the clinician-patient relationship supports the idea that trust itself is dynamic. This model suggests that while an individual's assumptions and predisposition to trust are the main factors leading to trust in a new clinician-patient relationship, interpersonal interactions over time reinforce or erode trust over time. 18 In MM, specifically, greater trust in the healthcare system at large has been associated with more active communication patterns (e.g., asking clarifying questions, stating individual preferences) in healthcare encounters. 19 Development of trust between clinicians and individuals with MM may be especially complex, given that they often receive care from a multidisciplinary team of clinicians, including medical oncologists, bone marrow transplant specialists, primary care providers, palliative medicine specialists, and advanced practice providers. In the context of life-threatening illness, prior work suggests that trust in clinicians may develop in part as a coping response to vulnerability and distress. 20 Further supporting this idea, a review on trust between cancer patients and clinicians found that patient trust in oncologists was generally high. 21 Patient factors associated with higher trust across multiple studies included older age, female sex, and White, non-Hispanic race/ethnicity. 21 In more recent qualitative work, individuals with breast and prostate cancers described oncologists' empathetic communication, thoroughness, and knowledgeability as enhancing trust. 22 However, an ethnographic study among women with cancer found that trust was a dynamic across the trajectory of illness. 23 While clinician communication did influence trust, the relational aspect of trust tended not to play a role until later in treatment. In contrast, on initial diagnosis trust was more "calculative," derived from an individual's assessment of websites, family and friend recommendations, availability or accessibility of the clinician, and urgency of commencing treatment. Relatively little is known about how individuals with MM, specifically, develop trust with clinicians.
The purpose of this study is to explore how individuals with MM describe the development of trust in clinicians and how they view trust as shaping treatment decisions or experiences. Results of this work can illuminate ways in which effective trust is established between patients and clinicians in the context of MM. These insights may help to inform future clinical and research efforts to strengthen trust and improve shared decision-making among individuals with MM.

| METHODS
This qualitative study was part of a larger project exploring patient and clinician experiences with shared decision-making in MM. Individuals with MM and clinicians who care for individuals with MM were interviewed. In this paper, we report only findings from individuals with MM. Because of the rarity of MM, we used convenience sampling to recruit individuals who had been newly diagnosed or relapsed within the past 2 years. We recruited participants from both large, academic medical centers and smaller community-based centers primarily in Northern and Central California, working with oncology staff to distribute pamphlets to eligible individuals, posting on multiple medical centers' study pages, and disseminating recruitment flyers at MM support groups. Individuals received a $40 Amazon gift card in appreciation for their participation. Individual 1-h semi-structured phone interviews were conducted between October 2018 and February 2019. Two interviews were conducted inperson at the patients' requests. All interviews were recorded and transcribed verbatim. The UC Davis Medical Center Institutional Review Board approved this study.
This interview-based study adhered to accepted criteria for reporting qualitative research (COREQ; Appendix A). 24 The interview guide consisted of 11 openended questions and was refined during data collection with suggested prompts (Appendix B). Before the interviews began, patients were told, "The purpose of this study is to understand the decision-making processes for treatment of multiple myeloma from both the patient and clinician perspectives," and that they had been chosen to participate due to their recent diagnosis or relapse with MM. Patients were asked to describe their experience being diagnosed and making decisions for initial and (if applicable) subsequent treatments. We asked about how treatment choices were presented and discussed, what factors were most important in making treatment decisions, and what role was played by the oncologist and family or friends during decision-making. Patients were not presented with anything specifically about trust and/ or shared decision-making unless to probe further into a comment made by the participant. The decision to focus our analysis on trust and shared decision-making was an inductively driven decision decided upon after all the interviews were conducted, and this analysis pertains only to portions of the interviews related to those topics.
The initial coding scheme was developed inductively. 25 Two researchers (R.L.W. and A.E.C.W.) initially independently reviewed two randomly chosen transcripts and documented emerging themes. They then met to discuss, compare, and develop these emerging themes. They repeated this process three times, further developing their coding schema in Nvivo with parent and child subthemes. The coding scheme and any differences of opinion were reviewed with the PI. Upon consensus, they went back and independently recoded these initial six transcripts with the more finalized codebook and met again to compare codes. They then tested their codebook on an additional four transcripts to compare their coding and concluded that they were reaching consensus. The remaining nine interviews were coded independently by R.L.W or A.E.C.W. We noted that after 12 patient interviews we were seeing many similarities in experience and considered that we may be reaching theoretical saturation. 26 Any disagreement in the application of the coding scheme was negotiated and a conclusion was achieved by consensus. The researchers and the PI met frequently to discuss the coding process. The three themes discussed in this manuscript are a subset of the themes inductively found in the data. These themes were selected by the entire research team as the focus for this manuscript.

| RESULTS
We interviewed a total of 19 individuals with MM. Most of our sample identified as non-Hispanic White (78.9%), between the ages of 55-64 (36.8%) or 65-74 (31.6%), and married (57.9%). Our participants were generally welleducated, with 42.1% reporting a 2-or 4-year college degree and 26.3% reporting a post-baccalaureate degree. Approximately two thirds of participants identified as newly diagnosed with MM (63.2%). The sample was evenly divided between men (n = 9) and women (n = 10). Details of participant characteristics can be found in Table 1.

| Qualitative themes
We identified several themes from participants' descriptions of establishing trust with clinicians. While most individuals described some degree of trust, there were different ways in which individuals chose to trust. We organized themes into three broad categories-externally validated trust, internally validated trust, and trust in relation to shared decision-making-and identified subthemes within each category.
Externally validated trust describes patients' predisposition to trust or distrust their clinicians. Externally validated trust may reflect an overarching belief in the trustworthiness of medicine, or beliefs about the trustworthiness of specific institutions or clinicians, but relies on external factors or characteristics rather on interactions between patient and clinician. Internally validated trust, in contrast, describes how individuals develop trust in a specific clinician. Internally validated trust develops through interpersonal interactions between patients and clinicians when patients observe behaviors that validate trustworthiness. Trust in relation to shared decision-making describes how patients relate feelings of trust, or lack thereof, to the process of decision-making. Each theme is explained below with illustrative quotes. See Appendix C, Tables C1-C3, where additional supporting quotes are presented for each theme. Participant quotes are lightly edited for clarity (e.g., removing "um," "er").

| Externally validated trust
Within externally validated, we identified several subthemes: trust in clinicians, trust in institutions, and personal beliefs or preferences. Example quotes illustrating these themes are listed in Appendix Table C1.

Trust in clinicians
Several patients expressed a high level of trust in clinicians generally, particularly in MM specialists. These feelings of trust were often enhanced if trusted others (e.g., family, friends, or other patients) had endorsed the particular clinician as having a good reputation: I started researching what multiple myeloma was. And I remember reading a statement, I don't remember by whom, but saying that it's very important to find a doctor who specializes in multiple myeloma. So, what I did was one of the times that the group was in [referring to the assigned group of clinicians on the inpatient oncology rotation]…And I said, Do you have a doctor on staff that specializes in MM? And they mentioned [

Trust in institutions
Many patients described particular institutions and clinicians within them as trustworthy, not because of first-hand experiences, but because of their reputation or prestige. Often, affiliation with an academic center or particular specialization in MM inspired trust. One patient described why he was excited to change from his small-town practice to a large, academic cancer center, because of access to clinical trials and new treatments, like chimeric antigen receptor (CAR) T-cell therapy: You know when you get passed on from a small town or an outlying city somewhere, right, and you get passed onto something like this [large academic medical center], like you just hit the major league. You know? (Pt08) Conversely, a few individuals had pre-existing perceptions of healthcare in general that made them hesitant to trust clinicians: Everything I read about the medical system it just doesn't inspire confidence, and it's like, what's the third leading cause of death in the US, you know? It's people seeking out medical care. (Pt19)

Personal beliefs or preferences
Due to personal beliefs or preferences, patients also sometimes identified external characteristics that they associated with trustworthiness. For example, a female patient preferred finding a female physician because of her belief that "female doctors are more caring and more receptive to women's opinions." (Pt16) Preferences that patients had could be related to stereotypes or biases. For example, one patient expressed difficulty connecting with his primary care provider due to his accent: And he -his English is fine, you could understand him but he's got, you know he's got an accent and what have you. And I'm not saying that you know I can't have a good conversation in English with him but sometimes I felt like I wasn't getting my point across as well as um, as I thought as well as I wanted it to or as well as I felt he ought to -he should understand it. (Pt12) Another was excited to learn that his oncologist was of Indian descent, because of his belief that they are more competent clinicians: And so we get in the office and [my new clinician] walks in and he's Indian. And not that I'm a racist sort of thing but he walked in I thought, yes! I get the smart [expletive]. I'm in there, I'm gonna be around for a while! Yeah! You know? (Pt08)

| Internally validated trust
Most patients relied to some degree on internally validated trust, finding signs to validate trustworthiness in the actions and behaviors of clinicians that patients personally observed. Subthemes related to internally validated trust were often related to communication skills, and included: competence, responsiveness, taking time & listening, honesty, and empathy. Example quotes illustrating these themes are listed in Appendix Table C2.

Competence
Patients who sought out their own information about diagnosis and treatment options often used this knowledge as a way to triangulate their trust in the oncologist. Patients were reassured in their oncologists' recommendations when they reflected knowledge of current evidence and new developments in treatment. Some patients expected detailed demonstrations of knowledge. For example, one patient who had performed independent research about options expected not only to hear that her oncologist was familiar with these options, but also that a detailed rationale would be provided to justify treatment recommendations: I want to hear their rationale, like I'm moderately knowledgeable about the different protocols and I know there's different mechanisms on different fields, like 3 main classes of drugs. And so I want to hear their rational. I want to hear their explanation about, okay, why pick this? Why pick this regimen over that regimen? Like, what's your rationale for that? And I would want them to explain their rationale to me. (Pt11)

Responsiveness
Many individuals described confidence that the care team would be there when needed as key to promoting trust. Often, this was about the reliability of the care team in following up with abnormal results, or returning patient phone calls in a timely manner: You can call over there and I get a call back or get an email right away. That's probably a big thing for me too. (Pt10) Conversely, feeling that they were being discounted or forgotten eroded trust: "My primary care physician when he came back from vacation he never followed up with me and I was very angry about that. He never called me, he never emailed, nothin'. It was almost like, well, it just sucks to be you." (Pt09).

Taking time and listening
Overwhelmingly, patients expressed that oncologists who were willing to spend time listening to concerns without distraction inspired trust. One patient, for example, appreciated his oncologist's willingness to disengage from the computer and make the effort to make complex information understandable to him.
He set back from the computer and just looked at me and started talking to me. And started talking to me in layman's terms just so I could understand. I don't understand all that blood work and stuff so he explained it. He explained what was going on. And he explained it in detail and with compassion. And then he showed me in papers, which I liked that, he just handed me papers. Because I'm not real computer literate. So he handed me paperwork and we read over it together. And so I could see that he had done his homework basically. And by the time we had that conversation I was ready to trust him, I was ready to trust him, um-hum. (Pt05) In contrast, many patients who did not feel trust described experiences where they perceived the oncologist was not willing to spend the time to ask questions and provide explanations. One patient who had gone through several oncologists before finding one that he trusted described always feeling rushed: It's your life and you know my oncologist spends 15 minutes with me and he's just a whole lineup all day of people who are in and out of his office… And honestly I think the best way to put this is that I just felt like I was a cog in a wheel. (Pt09)

Honesty
Honest and transparent communication were also important in establishing trust between patients and oncologists. One patient described his clinician's direct explanation of the disease and its trajectory as helping to lay the foundation for a positive relationship: He explained the whole thing about multiple myeloma. He explained where it was, where it's going, what they do. He just flat out laid it on the line. From then on me and him had like, we had the greatest relationship. (Pt08) On the other hand, trust was eroded when patients perceived that the communication was not transparent. A patient who was initially treated in the hospital by a team of oncologists and residents described her perception that nobody wanted to be the one to deliver honest information about prognosis or negative effects of treatment. This led her to do her own research rather than just accepting treatment recommendations: And I don't know if that's with all cancer patients but they just didn't know what to say. They would kind of just sugar coat when they'd come in and talk to me. So no one really educated me on what myeloma was. I kinda had to self-educate myself. (Pt10)

Empathy
Patients described oncologists' positive attitudes as helping them to cope with uncertainty and eventual treatment failure: But at the end of the day he always told me not to worry at all because if this particular one doesn't work there's another one. And I'm a very good candidate to be on another med or clinical trial. He's always really so positive with me. (Pt01) Communicating warmth and empathy was also important in developing trust. One patient described a physician's ability to communicate empathetically as key in his decision to be treated there: Similarly, one or two patients who were less trusting of their oncologists described them as competent, but not caring, or suspected them of having ulterior financial motives for their treatment recommendations: Because I've met doctors [who] were so educated, you know or you'll think oh you want them because he's educated, he has experience. But they're not caring doctors. They just know their stuff but they don't really mean to help patients you know they just mean to help themselves. (Pt16)

| Trust in relation to decisionmaking: concurrence with professional recommendations
Often for individuals with MM, shared decision-making is not as straightforward as selecting between treatment options. Rather, patients choose a clinician and trust them to curate informed recommendations based on their specialized knowledge and attention to their patients' values. When patients trusted that their clinicians understood their needs or had previous positive experiences relying on treatment recommendations it instilled confidence in treatment choices. However, patients who were more generally distrustful or who had prior negative experiences sometimes expressed a desire to conduct their own research or to switch care teams entirely. Example quotes illustrating this theme are listed in Appendix Table C3.
Many patients expressed confidence relying on decisions made by their clinicians when a trusting relationship had been established. One patient, for example, described her process deciding to participate in a clinical trial. She was eligible for several trials and also had the option not to participate, but ultimately deferred to her clinician's recommendation because of her faith in him: Really to be honest with you…I wasn't weighing anything [in terms of treatment decisions]. I was just grateful to get into this trial. I believe in my doctor... [he] wanted me in this one, I got in it. He was happy; I was happy. So that was just my attitude about that. (Pt02) Sometimes patients who tended to defer to the clinician's recommendations explained that they did not have the time or capacity to question the clinician's judgment or engage in a more complex decision-making process: It was pretty obvious that I couldn't negotiate the intellectual portfolio that was gonna be required of me [to make a treatment decision]. So, plus I was really getting kind of looped out by all the chemo. So, at a certain point I just had to say, you know I trust this guy, I have to listen to what he says, because I don't think I can make this decision myself. (Pt15) However, some individuals use their previous positive experiences with the clinician's treatment recommendations to validate their decision to trust in subsequent treatment recommendations: Yes, honestly [I listen to whatever treatment my doctor recommends], because I didn't think I was gonna live and since they saved me like they did I trusted them so extensively and I still do. Yes, because they know exactly every inch of me, every single cell. They've taken cells out of my body now and put cells in. I trust them extensively. (Pt01) In contrast, patients who expressed lower trust in their physician were more apt to seek second opinions and do their own research rather than relying on their physician's recommendation. One patient described a lack of trust in his first oncologist, resulting from his perception that the oncologist was always rushed, and gave generic "cookiecutter" treatment recommendations. Ultimately, this lack of trust led him to refuse treatment and switch clinicians.
They had this protocol that they used for everyone. He did a bone marrow biopsy in my hip and said, yeah, he confirmed it was myeloma. This is the treatment we're gonna put you on. And I just did not have a rapport with him. And at that point I really started thinking about getting out of [my cancer center]. [Describing the new cancer center after switching, and why he trusted their recommendations more] He talked to me a lot about multiple myeloma, what it was. One of the things he said that was particularly reassuring was that this is something that's very manageable. (PT09) Another patient, similarly, sought out second opinions and ultimately decided against her oncologist's recommended treatment because of her suspicion that they would receive a "commission" and therefore had an incentive to prescribe unnecessary treatments: It's supposed to be a patient/doctor decision. It's not just the doctor telling the patient you're gonna be treated…patients have the right to say no and to do their own research and be a part of the decision making. But this doctor was so adamant at me going on treat-ment…Well why would you treat a patient when they don't need it unless you needed the extra money? So that's why I want people to be aware that some doctors are like that.
A few patients were reassured by doing their own research, even as they generally accepted treatment recommendations from clinicians.
[My clinician] hasn't ever suggested anything that's so outlandish that [I would question his treatment recommendation]. And believe me I do get on the internet and you know research some of this stuff. But I -the thing I don't do is read the horror stories [about different treatments]. Because I think that just kind of puts another nail in the coffin so to speak, because then you do start to worry if you read -if you read the negative stuff. (Pt14) Some patients who went with clinician recommendations for chemotherapy researched and pursued concurrent integrative treatments. They wanted to discuss these with their clinicians even though they were choices that were not offered within the oncology care setting.
[My doctor] basically said it's my decision [if I want to use naturopathic treatments]. He said if it was anybody else I would say, no way, but you know he knows how strong I am and I how much research I've done on this and he said, well as long as we continue to monitor your blood results go for it. And so I did. (Pt17) After the initial treatment phase, patients often sought out information about potential upcoming treatment options. In this phase of treatment, patients typically understand that they will experience a relapse sooner or later and begin gathering information before needing to make a treatment decision.
[Explaining that even though he primarily relies on his clinician's recommendations, he still wants to research different options in advance because he knows he will eventually relapse] Yeah, there are two clinical trials that I asked about because I'm pretty much at the end of my protocols.

| DISCUSSION
Among patients with MM, trust between patient and clinicians evolves throughout the treatment trajectory, but precisely how trust impacts the shared decision-making process remains uncertain. More often than not, patients in our study deferred to clinicians' treatment recommendations. However, the patient's trust in the clinician was key in allowing them to accept the proposed treatment plan. While some prior quantitative work has found that greater trust in clinicians is associated with shared decision-making, 27,28 our findings suggested that the relationship between trust and shared decision-making may be more complex.
Our findings about how trust develops in MM are largely consistent with findings from previous qualitative work among individuals with cancer. 21,23 In particular, our themes of externally and internally validated trust are similar to the concepts of "calculative" (based on reputation) and "relational" (based on interactions) trust previously described by Yeh. 23 Calculative and relational trust are described as evolving in a linear process, with the role of calculative trust waning and the role of relational trust growing over time. We found this to be relevant for individuals with MM as well, because it takes time and repeated interactions to develop internally validated trust. However, in the context of MM, there is additional complexity because of the chronic nature of the disease and the involvement of multiple clinicians over time. For example, an individual with MM may go through the process of establishing trust with a medical oncologist, but later require a transplant specialist, a palliative care clinician, or an oncologist who participates in clinical trials. In addition, MM is more likely to be diagnosed either emergently or after a lengthy period of vague symptoms compared with other cancers, 29 especially common cancers which can be diagnosed through routine screening. In emergent cases, patients may have limited time for even calculative assessments of trustworthiness, while patients who experience lengthy workups or delayed diagnoses may be less disposed to trust clinicians.
The individual factors we identified as driving both externally and internally validated trust, such as reputation, honesty, empathy, and competence were also consistent with findings from previous work. 20,21,23 However, one important external factor mentioned by several participants was finding a clinician who specializes specifically in MM, due to the complex nature of the disease. Finding a specialist may be more challenging in MM compared with more common cancers, particularly for individuals who live in rural or medically underserved areas. Future work might explore further how access to highly specialized clinicians impacts the development of trust in MM and other relatively uncommon cancers. Several factors that we identified as relating to externally validated trust are beyond the immediate control of the healthcare team, such as the individual patient's perceptions of the oncologist or healthcare center's reputation, and their pre-existing preferences. However, these findings suggest that efforts to elevate the reputation of the clinician or organization within the community, such as through engagement with community organizations, could help to promote trust.
At the same time, the themes we identified around internally validated trust suggest that there are steps clinicians can take to improve trusting relationships. Consistent with previous work, our findings suggest that building trust involves specific patient communication approaches, such as taking adequate time for appointments, answering questions, appearing informed and knowledgeable, actively listening, and expressing empathy. 19,21,30 Relatively few studies have examined interventions to increase trust, though physician communication skills training, physician disclosure of financial incentives, and patient-physician concordance in care beliefs have been associated with significant increases in patient trust in randomized controlled trials. 31 While no patients in our study reported that their clinicians disclosed financial incentives, a few did express concerns that financial motives could underlie treatment recommendations, suggesting that transparency around such incentives may help to promote trust.
Patients' descriptions of trust and decision-making in our study also support the conceptualization of shared decision-making in MM as being both an "interactional" as well as "transactional" process. As described by Epstein and Street, 32 a transactional model of shared decisionmaking is one in which patients and clinicians exchange information, often using a quantitative approach (e.g., decision aids that emphasize average treatment risks and benefits), to come to an agreement on a treatment decision. In an interactional model, on the other hand, clinicians and patients engage relationally and affectively in a process of shared deliberation, achieving what Epstein and Street term "shared mind." 32 Shared mind occurs when clinician and patient are attuned in their relationship, such that "new ideas and perspectives emerge through the sharing of thoughts, feelings, perceptions, meanings, and intentions." 32 Whereas shared decision-making from a transactional perspective is often explicit (i.e., the patient makes a choice from among discrete options after reviewing information), interactional shared decision-making can be more implicit, in that decisional preferences are conveyed through attunement and "shared mind" rather than in a series of individual choices.
While many patients in our study described transactional decision-making interactions, the concepts of interactional decision-making and "shared mind" also resonate with our participants' descriptions of feeling known or seen by their oncologists and the confidence such feelings inspired in the oncologists' treatment recommendations. Thus, while it may appear that some patients simply defer to oncologists' recommendations out of blind trust, it is possible that for others this deference follows a process of affective engagement and information-sharing that has allowed the oncologist to present tailored information and treatment recommendations informed by an understanding of patient values. In the former scenario, trust may be viewed as a barrier to shared decision-making because it leads the patient to accept treatment recommendations without question. In the latter scenario, trust is essential to achieving the attunement between clinicians and patients that is necessary for "shared mind" to occur.
Several previous studies have pointed out that trust could have a negative impact on shared decision-making. 23,33,34 Yeh suggest that as relational trust grows, some patients cease participation in the decision-making process and give complete control to their clinicians. 23 Helmes et al. found, similarly, that higher trust was associated with preferring to leave medical decisions to the clinician, raising the concern that trust may not actually lead to this preference (consistent with an interactional shared decisionmaking model), but rather that patients might express trust as a way of justifying their lack of interest in participating actively in the decision-making process. 33 Alternatively, Kraetschmer et al. 34 observe that there may be optimal levels of trust, such that both very high and very low levels of trust may work counter to shared decision-making processes. They observed that individuals whose trust was high but not "blind" had decisional preferences most aligned with shared decision-making principles, whereas individuals with "blind" trust preferred a very passive role and individuals with low trust preferred an autonomous role. 34 We found, similarly, that our participants' descriptions of trust in relation to shared decision-making ranged from very trusting to very distrusting of treatment recommendations, with many individuals falling somewhere in between. Further supporting Kraetschmer et al.'s findings, the few individuals who were highly distrustful seemed not to engage in decision-making discussions with clinicians they distrusted at all, preferring to seek out second opinions or switch care providers altogether.
The clinician-patient relationship in a disease with a long-term treatment trajectory like MM, may lend itself particularly well to an interactional model of shared decision-making, because of the opportunity to build trust and strong patient-provider relationships over time. The complexity of decision-making in MM may nudge patients toward an interactional decision-making approach; given multiple, rapidly changing treatment options, a purely transactional decision-making approach may be too daunting for many patients. If this is the case, it may be that for many individuals with MM finding a clinician they trust is the most important choice, facilitating interactional decision-making as the clinician-patient relationship develops across the trajectory of care. This interpretation is supported by our finding that many patients sought out information about future treatment options and wanted to discuss them with clinicians, even as they reported generally deferring to treatment recommendations when it was time to decide. Indeed, this anticipatory aspect of decisionmaking may be a feature unique to MM and other incurable cancers, as the inevitability of eventual relapse leads patients to pre-emptively seek information about future treatment options. While our study suggests that most individuals with MM are generally trusting of their clinicians and their treatment recommendations, the extent to which these feelings reflect interactional shared decision-making as opposed to a "blind" trust and preference for less autonomous decision-making are not entirely clear. While our study was not designed to examine changes in trust or shared decision-making over time, studies of longitudinal clinician-patient relationships in chronic cancers could shed further light on the development of trust and its role in shared decision-making in this context.
Our study has several limitations. Our qualitative interviews were designed to elicit rich data related to decisionmaking in MM rather than to produce generalizable findings. Given that our sample largely identified as White, highly educated, and healthy enough to participate in this study, it is possible that trust and decision-making operate differently in other populations and for other conditions. In fact, differences are likely given our finding that trust was influenced by personal beliefs and preferences, and evidence from prior work that culture, language, race, and ethnicity impact trust and shared decision-making. 35,36 While our study sheds light on factors that influence trust and the relationship of trust to shared decision-making, the extent to which patient-clinician trust and shared decision-making can be improved is unclear. Future work might explore relationships of trust and shared decisionmaking among more diverse groups of MM patients, and examine the effect of communication training or similar interventions on trust and shared decision-making.
Despite these limitations, our study offers new insight into the ways in which trust develops and impacts shared decision-making in the context of patient-clinician relationships in MM. While our findings suggest there are many possible approaches for building trust in clinicianpatient relationships, establishing trust is an essential step on the path to effective shared decision-making.

ETHICS STATEMENT
This research was reviewed and approved by the Institutional Review Board (IRB) of the University of California, Davis.  2%). The sample was evenly divided between men (n = 9) and women (n = 10).

APPENDIX A COREQ 32-ITEM CHECKLIST
Data collection 17. Interview guide Participants were not provided with an interview guide before the interview.
The interview guide was pilot tested with members of the research team. The interview guide was reviewed and updated during data collection.

Repeat interviews
No repeat interviews were conducted.
19. Audio/visual recording Audio recordings were collected.
20. Field notes Researchers wrote brief summaries and highlights of interviews immediately following the interviews.

Duration
Interviews lasted for 1 h.

Data saturation
After we had interviewed seven patients and had received their corresponding transcripts, we began to review the interviews and conduct initial analysis. We assessed that after 12 patients we had reached data saturation. If additional patients then contacted us for interviews, we still interviewed them as we wanted to include as many individual experiences as possible to illustrate the broadest range.

Transcripts returned
Transcripts were not returned to participants.

Data and findings consistent
There is consistency between the data presented and findings. Quotations were selected for their representation of the data.

Clarity of major themes
Our three major themes (and subthemes) are clearly presented in the findings along with definitions or explanations.

Clarity of minor themes
Yes, diverse cases are included in the findings. For example, in the "personal beliefs or preferences" subtheme, we provide examples how physicians' ethnicities can serve both as a positive and as a negative attribute for their patients.