Communication Strategies in a Code Status Conversation.

Despite the emphasis on engaging in shared decision-making for decisions involving life-prolonging interventions, there remains uncertainty about which communication strategies are best to achieve shared decision-making. In this paper, we present the communication strategies used in a code status discussion in a single case audio recorded as part of a research study of how patients and physicians make decisions about the plan of care during daily rounds. When presenting this case at various forums to demonstrate our findings, we found that some clinicians viewed the communication strategies used in the case as an exemplar of shared decision-making, whereas other clinicians viewed them as perpetuating paternalism. Given this polarized reaction, the purpose of this perspective paper is to examine the communication strategies used in the code status discussion and compare those strategies with our current conceptualization of shared decision-making and communication best practices.

Most hospitalized patients have a defined code status that indicates whether the patients will be resuscitated if they became pulseless; as a result, conversations between physicians and patients on the topic of resuscitation are relatively common. Because code status decisions involve potentially limiting a life-prolonging intervention, clinicians should engage in a shared decision-making process (1). Nevertheless, such conversations can be quite difficult, particularly when the physician perceives that a resuscitation attempt would be unsuccessful or could lead to a decreased quality of life. A policy statement issued by the American College of Critical Care Medicine and the American Thoracic Society emphasized that clinicians should receive communication skills training to encourage patient involvement in the decision-making process, but the statement conceded that there remains uncertainty about which communication strategies are best to achieve shared decision-making (1).

The following case of a code status discussion exemplifies the uncertainty and debate surrounding communication during shared decision-making. This code status discussion was audio recorded as part of a research study of how patients and physicians make decisions about the plan of care during daily rounds. Forty conversations during daily rounds were audio recorded in a small teaching hospital and analyzed using conversation analysis, a qualitative method of the turn-by-turn analysis of the interaction (2). When presenting this case at various forums to demonstrate our findings, we were intrigued by the reaction we received from clinicians. To some, this was an exemplar of shared decision-making; to others, this was an exemplar of paternalism. Given this polarized reaction, the purpose of this perspective paper is to examine the communication strategies used in the code status discussion and compare those strategies with our current conceptualization of shared decision-making and communication best practices.

Case

This case involved a conversation between the medical team and a 73-year-old white male patient with a high school education. The patient had been admitted the previous night, complaining of weakness in his legs leading to falls in the previous week. The patient had a past medical history of multiple chronic illnesses that limited the probability of surviving resuscitation without resulting in significant physical or cognitive changes. The team had gone through an assessment, asked the patient questions about his living situation, and discussed the plan for his hospital stay. Then, a physician on the team began a sequence on the patient’s code status. (“Pr1” is the physician, and “Pt1” is the patient.)

This code status discussion has a structure known as a “perspective-display sequence” (3, 4). In the context of end-of-life treatment decision-making, the perspective-display sequence involves three turns: 1) the physician eliciting the patient’s perspective, 2) the patient sharing their perspective, and 3) the physician incorporating that perspective into a treatment recommendation (5).

Pr1: You know I know overnight they talked about your code status. Do you remember talking about that at all? Is that phrase ringing a bell? A lot of times we ask people you know what they would want us to do if we came here that you had—if you had died or passed away suddenly on us.

Pt1: Yeah.

Pr1: What’re your thoughts on that?

Pt1: Well I guess if you can bring me back to life, I’d rather live than die.

Pr1: Yeah and that’s what most people say. I’ll tell you though in your situation with everything you got going on the chance of you surviving, that’s probably around like three to four percent. Pretty poor odds.

Pt1: Yeah.

Pr1: And if you did survive it, almost certainly you’d be in a nursing home for the rest of your life. I just tell you that ’cause we don’t do a good job of tellin’ you. Most people say I wanna live, but we don’t do a good job of tellin’ ya what livin’ looks like after that.

Pt1: Yeah.

Pr1: Which is probably with a lot of tubes in you, livin’ in a nursing home.

Pt1: Yeah. I really don’t want that.

Pr1: And that’s what a lot of folks say. If you feel pretty strongly about avoiding that quality of life, you know, I would recommend we make you what’s called a DNR. A do not resuscitate. That just means in that sort of extreme situation if you were to die on us, we would allow a natural death. We’d let you pass peacefully.

Pt1: Yeah.

Pr1: And we wouldn’t do any of that stuff that would put ya in a nursing home for the rest of your life.

Pt1: Sounds good.

Pr1: Okay. What that means is we were going to come in and put a purple bracelet on you that says DNR. Okay? It’s not going [to] change any of the care we give you. We’re still going to do everything we can to get you stronger and get you feeling better but again just in that one situation if you were to die on us we wouldn’t try to bring you back to a quality of life that would be no good for you.

Pt1: Yeah.

Pr1: Make sense?

Pt1: Okay. Yeah.

Pr1: Alright, thanks for chatting with us about that. Okay?

Pt1: Okay.

Pr1: Otherwise we’ll get to work. Alright. Really nice meetin’ you.

Examining the Communication Strategies

We will discuss how each part of the perspective-display sequence was employed in this conversation: eliciting the patient’s perspective, patient sharing their perspective, and the recommendation.

Eliciting the Patient’s Perspective

The physician framed resuscitation as what the team would do if the patient were to die. Scholars have advocated for using words such as “allow natural death” over “do not resuscitate” to frame the code status discussion (6), but, in this example, the physician initially discussed death without suggesting that they would be “allowing” the death or that doing so would be “natural.” However, we argue that initially avoiding these terms is useful, because both “allowing” and “natural” are not impartial terms: They imply a polarity between the “right” and the “wrong” thing to do. This is important because the physician then elicits the patient’s perspective from the patient. Eliciting the patient’s perspective is a critical component of shared decision-making (1), and doing so in a way that allows the patient to share an unfiltered perspective is important for open dialogue.

Jacobsen and colleagues highlighted shared decision-making as first evaluating the prognosis and treatment options and then understanding the range of priorities that are important to the patient, given the prognosis (followed by a recommendation) (7). The physician in this case deviated from this sequence in one important way: He did not discuss the prognosis or treatment options before asking for the patient’s perspective. This omission may have been helpful in assuring that the patient was able to share his preferences in response to the neutral presentation, “What are your thoughts on that?” Alternatively, asking for the patient’s perspective after discussing the prognosis may limit the patient’s ability to share their perspective, particularly in situations of a poor prognosis when the patient’s preference for resuscitation could be seen as a move against the physician’s authority. In addition, previous work has shown that emphasizing that the treatment would not work or would cause harm instead of asking for the patient’s perspective leads to surrogate resistance; patients and surrogates react strongly when they perceive they are not being offered a choice (8). Asking patients to share their perspective in neutral language before discussing prognosis provides the opportunity for an honest discussion of the patient’s perspective.

Patients Sharing Their Perspective

The patient’s shared perspective, “I’d rather live than die,” is perhaps not surprising, given that patients who choose “full code” tend to conceptualize resuscitation as the restoration of life (9), and the presentation of the choice seemed to be one of life versus death. However, instead of moving to the third part of the sequence—giving a recommendation based on this shared perspective—the physician presented additional information about the patient’s choice using a type of “incomplete syllogism” (10) that has also been noted in surgeon recommendations against doing surgery (11). An incomplete syllogism starts like a syllogism with a general premise for all people and then a particular premise for the individual’s situation, but instead of stating the conclusion, the conclusion is merely implied and left for the individual determine (10). Starting with “that’s what most people say,” the physician suggested a general premise that most people “would rather live than die.” The second part of the syllogism is the particular premise for the patient’s situation: The physician described the poor chance of the patient surviving resuscitation and the expected quality of life if survival happened. In this case, the patient deduced the conclusion to the syllogism by declaring that he did not want that quality of life, to which the physician reaffirmed this conclusion that most people would not want to have the described quality of life (“And that’s what a lot of folks say”).

According to Bernacki and colleagues, giving a direct, honest prognosis and focusing on the patient’s quality of life are both considered best practices for communication about serious illness care goals (12). In addition, expert practitioners in doctor–patient communication have shared that patients need an adequate understanding of the outcomes, including the effectiveness of treatment options, to avoid unrealistic expectations (13). For code status decisions in particular, patients with more knowledge of the outcomes of resuscitation are less likely to choose resuscitation (14). One could argue that, by sharing the expected poor prognosis quantitatively as well as describing qualitatively what the patient’s quality of life might be like if his choice were followed, the physician was providing the patient important information to better understand his choice.

However, clinicians who viewed this communication strategy as paternalistic argued that the way the physician provided the information “nudged” the patient into a different perspective and a different choice (subsequently restricting that choice). Emergency physicians discuss having an “agenda” of what they believe is the best option for the patient and engaging in “guided” shared decision-making to lead the patient to that choice (15). In addition, oncologists display behaviors of implicit persuasion during treatment decision-making, such as underreporting side effects or presenting recommendations as decisions authorized by the medical authority (16). In this case, it is possible that the physician presented a bleak outcome with unpleasant discourse of “a lot of tubes in ya, livin’ in a nursing home” with the incomplete syllogism to lead the patient to a conclusion against this option. Nevertheless, Blumenthal-Barby and colleagues argued that a neutral and balanced presentation of options may not always be appropriate, including in situations in which it is necessary to counter existing bias that patients already hold to ensure that the patient is making an informed choice (17). The physician in this case may have provided the necessary information––in language that the patient appeared to understand––to counter the patient’s bias of preferring life over death as a means to give a balanced presentation of the outcomes of the choices.

Recommendation

On the basis of the patient’s changed response, the physician provided a recommendation for a do-not-resuscitate order. Together with the recommendation, the physician stated what a do-not-resuscitate order entailed (here using the “allow natural death” language) and closed the conversation when the patient indicated agreement with the recommendation.

When and how to incorporate recommendations during conversations of shared decision-making continue to be important topics of study. Frongillo and colleagues found that providers who gave recommendations were less likely to ask for the patient’s perspective (18). Similarly, Landmark and colleagues showed that starting with a recommendation and then formulating a hypothesis of the patient’s perspective (rather than first asking for the perspective) constrains the patient’s options (19). In this example, the physician did ask for the patient’s perspective in his own words before making a recommendation, which exhibited more of a shared decision-making model. Jacobsen and colleagues suggested that recommendations are appropriate in shared decision-making when they are based on patient priorities most compatible with the likely prognosis and available treatment options (7). Although one could argue that there was nudging in the above example when describing the prognosis, it came after the patient had shared his perspective and in the context of providing the patient information to more fully understand his choice and uncover his priorities, a design that appears to be promising and needs to be studied further.

Discussion

This case demonstrates that although we understand generally what “pieces” should fit into shared decision-making (eliciting preferences, providing information, and so forth), we are still unclear about when each of these sequences should occur in the interaction and what they should look like. More research is needed to determine how to elicit patient preferences, such as how it should be phrased and whether to elicit preferences before or after providing information about prognosis, options, and outcomes. More research is also needed about how to provide such information. Was this an example of nudging or providing honest information about the outcomes of the patient’s choice? Is it acceptable to “nudge” or engage in “guided decision-making,” particularly to counter a potential bias against death? Determining what’s “best” will require not only further studies using robust methods that link communication with patient outcomes but also a continued dialogue about how much influence and “nudging” from physicians seem appropriate during shared decision-making or if any nudging is considered paternalism.

This case also shows the value of discussing real-world scenarios among colleagues. Differences in opinion of how these conversations “should” proceed can reveal the uncertainty that remains in the field and the need to regard communication skills as tools for one’s toolbox to apply when the patient’s situational context is appropriate (20, 21). Discussing real-world cases may also be helpful to include in teaching communication skills to trainees. Miller and colleagues suggested a combination of didactic teaching methods (such as lectures) with observation and practice as necessary for trainees to obtain both the explicit and tacit knowledge they argue are needed to learn clinical judgment and clinical skills, including how to communicate effectively (22). Other effective training models use a small-group approach for learners to offer suggestions to their peers for difficult communication problems as they work through them in real time with patient actors, providing an opportunity for important discussion of various communication strategies (23, 24). Incorporating real-world cases—presented word for word—could also be beneficial to start a dialogue about the implications of using various communication strategies for patients in a real-world context.

There has been a call for the study of the “basic science” of communication to better understand the nuances of how best to communicate with patients (25). In exploring this case, we identified some of the nuances that require further study. Until we have more evidence of how these nuances affect patient outcomes and a further discussion among colleagues of how these nuances fit into our model of what these conversations should look like, we will leave it up to the reader to decide if this is a case of shared decision-making or paternalism.