Patient Perceptions of Oncologist–Patient Communication About Prognosis: Changes From Initial Diagnosis to Cancer Recurrence

Abstract

Cancer recurrence presents oncologists with many challenges, including discussing prognosis. A thematic analysis of patient interviews was conducted to learn more about how this challenge is met by doctors and patients in their communication. Female patients being treated for a cancer recurrence at a comprehensive cancer center (N = 30) were asked to recall how they discussed their prognosis with oncologists when initially diagnosed and when the cancer recurred. All interviews were recorded, transcribed, and thematic analysis applied to patients’ remarks concerning prognosis. Following comparison of prognosis quotes identified in the transcript, two primary themes were drawn from the data: communicating optimism at the initial diagnosis; and communicating chronic disease management at recurrence. Three subthemes were also identified in the recurrence communication: managing the new information; ambiguous communication about the prognosis; and the information seeking paradox. Problematic integration theory is offered as an important explanatory mechanism for understanding oncologist-patient communication about prognosis during a cancer recurrence.

Although cancer survivorship has gained great attention in the United States in recent years, the fact remains that many cancer survivors will re-assume the patient role in the event of a cancer recurrence. The National Cancer Institute-sponsored Surveillance Epidemiology and End Results (SEER) data provide evidence that more than one million adults will be diagnosed with cancer each year, with 65% surviving for five years (Ries, Melbert, et al., 2007). These survivors face an increased risk of experiencing a second cancer, perhaps as many as 10% (Ries, Young, et al., 2007). If this second cancer is a recurrence, that is, the return of cancer following a disease-free period of time (ASCO, 2009), then the prognosis is likely to be poor, particularly for those with metastatic¹ disease.

The recurrence diagnosis represents a loss of control and changed identity for the sick person and close caregivers (Howell, Fitch, & Deane, 2003; Lamerato, Havstad, Gandhi, Jones, & Nathanson, 2006; Mahon & Casperson, 1995; Newsom, Knapp, & Schulz, 1996; Schulz et al., 1995). Cancer recurrence also disrupts patient support systems and forces a recalculation of hope for survival (Beadle et al., 2004; Cella, Mahon, & Donovan, 1990). These psychosocial challenges may present significant implications for physician–patient communication, particularly discussions about the patient’s new and worsened prognosis.

Although the delivery of bad news is difficult in any health context, recurrent cancer carries the added burden of a previous experience with cancer that may influence patients’ expectations for a second cure. Unfortunately, recurrent disease does not typically progress or resolve (i.e., curable) in the same way as the initial disease experience. Although some experiential aspects of recurrence have been addressed by research, doctor–patient communication at this complex time has not. Learning more from patients about how recurrent cancer is discussed in the initial phase of the disease will illuminate those aspects of recurrence that make it a unique patient experience, and ultimately inform interventions that could facilitate communication at this time. In-depth study of communication at recurrence will also shed light on those non-routinized moments of care that enable oncologists and patients to build therapeutic partnerships. Thus, the purpose of this study was to explore patients’ perceptions of what their oncologists communicated when patients were first diagnosed and then when diagnosed with a cancer recurrence.

LITERATURE REVIEW

Cancer Recurrence

In light of cancer recurrence, patients experience a wide variety of psychosocial outcomes, with distress figuring prominently in their responses. Generally, recurrence can impair the activities of daily living (e.g., completing chores, caring for children) and can produce decreases in mood and quality of life (Camileleri-Brennan & Steele, 2001; Munkres, Obsert, & Hughes, 1992; Okamura, Yamawaki, Akechi, Taniguchi, & Uchitomi, 2005). There is some disagreement across studies about general levels of distress during initial and recurrent cancer experiences. Some researchers describe no change in self-reported distress from the first to second diagnosis (Howell et al., 2003; Worden, 1989). However, Andersen and colleagues (2005) recently stipulated that cancer-specific, rather than general, distress is greater at recurrence than at the first diagnosis. Others provide evidence for significantly higher distress at recurrence than the first cancer experience (Cella et al., 1990; Mahon & Casperson, 1995; Weisman & Worden, 1986).

Although distress frequently accompanies recurrence, the feature that most differentiates it from the initial cancer experience is the need to readjust hope for survival (Mahon & Casperson, 1995; Sarenmalm, Ohlen, Oden, & Gaston-Johansson, 2008; Worden, 1989). There is some evidence that patients shift their primary source of hope from medical providers and treatment during the initial cancer, to family support and faith during recurrence (Ballard, Green, McCaa, & Logsdon, 1997). Other studies have reported that at diagnosis, recurrent patients looked to treatment staff for hope first, but perceived a changed attitude in their providers as evidenced from their communication (Cella et al., 1990; Howell et al., 2003). More specifically, recurrent patients from three study samples described changes in communication with their providers as less attentive, more hurried, and more depersonalized (Cella et al., 1990; Howell et al., 2003; Mahon & Casperson, 1995). Mahon and Casperson (1995) interviewed 20 recurrent patients, all but one of whom felt they received less information about the disease and treatment than at their first cancer experience. Similarly, a larger sample of patients interviewed within 1 month of diagnosis reported that medical staff assumed they were coping well, even though all agreed the recurrence was more upsetting than the initial diagnosis (Cella et al., 1990). Another set of women with recurrent ovarian cancer also noted changes in communication, such as less confirmation of their ideas and feelings, less information overall, and less partnership with treatment staff (Howell et al., 2003). These women interpreted this change as being without hope and subsequently felt hopeless.

Findings concerning cancer recurrence depict an illness experience characterized by great personal and medical uncertainty. This is contrasted with the fact that recurrent patients often retain deep familiarity (i.e., reduced uncertainty) with treatment procedures and routines, wait times, clinic personnel, settings, and side effects (Worden, 1989). In fact, as the National Cancer Institute (NCI) suggests, patients may find this familiarity a resource in the face of a second diagnosis, and use their former experiences as a tool for coping (NCI, 2005). However, this familiarity may disguise the fact that the prognosis associated with a metastatic recurrence is typically not as favorable as for the initial cancer (Holland & Lewis, 2000).

Discussing the Prognosis

Prognosis discussions are unique encounters that are deeply memorable to patients (Gillotti, Thompson, & McNeilis, 2002; Mager & Andrykowski, 2002). A growing literature focuses on cancer patients’ preferences for prognostic information and provides some empirical evidence of how oncologists are managing these difficult conversations (Hagerty, Butow, Ellis, Dimitry, & Tattersall, 2005). Most advanced cancer patients want to discuss survival duration at some point, but vary widely about when and how to discuss it (Hagerty, Butow, Ellis, Lobb, et al., 2005; Hagerty et al., 2004). Hagerty et al. (2005) surveyed a sample of people recently diagnosed with metastatic cancer about their preferences for discussing prognosis, finding that most (87%) wanted to discuss treatment, side effects, and symptoms at diagnosis, but fewer (59%) wanted to discuss survival estimates. Almost half of the respondents who wanted to discuss life expectancy also wanted the oncologist to bring it up. Most respondents endorsed a realistic approach for discussion, one that is both direct and tailored to their own case.

However, oncologists and patients hold divergent views about who should introduce prognosis discussions. Generally, findings support the fact that oncologists prefer to let the patient determine when to bring up prognosis, particularly when patients are transitioning from a curative to noncurative (i.e., terminal) stage of disease (Baile, Lenzi, Parker, Buchman, & Cohen, 2002; Elger & Harding, 2002; Miyaji, 1993). Further, when physicians do give an estimate of a patient’s expected life span, they tend to overestimate time (Lamont & Christakis, 2001). Few communication tools and little advice are available to physicians to facilitate their prognosis discussions. Guidelines tend to adopt a general approach that often does not fit the unique scenarios that accompany this emotionally laden event (Wittenberg-Lyles, Goldsmith, Sanchez-Reilly, & Ragan, 2008). This appears particularly relevant to physicians and patients with recurrent cancer, who have a strong reason to address prognosis differences.

Cancer recurrence signifies a major turning point in a person’s cancer prognosis such that the patient’s expected survival is drastically shortened. As described earlier, patient preferences for prognosis discussions are quite variable, leaving oncologists hesitant to bring this issue up at all. Although patients may be aided by the recollection of a previous cancer experience, the recurrence ushers in new unknowns that may not be manageable with past strategies. How oncologists and patients approach and eventually discuss prognosis in the early phase of recurrence discovery is likely to play an important role in patients’ and caregivers’ adjustment to recurrent disease. However, although the prognosis literature suggests physicians take a passive approach to discussing prognosis, research in health communication has not examined this specific communication issue in terms of strategies, approaches, or potential outcomes within the recurrence context. Additional insight about how patients approach the prognosis question is crucial for developing tools and approaches that may facilitate difficult communication between health care providers and their patients.

Our overall goal is to examine patients’ perceptions of the messages they received from their oncologists when initially diagnosed and when their cancer recurred. Although prospective (i.e., longitudinal or cohort) study methods may provide more timely self-reports of patients’ perceptions, the exploratory nature of this study restricted us to retrospective reports about receiving both diagnoses from patients currently experiencing their recurrence. Receiving a cancer diagnosis is a highly salient and memorable event. We hope to learn something about the communication surrounding the initial diagnosis because these experiences are likely sharply recalled and considered by the recurrent patient. Recalled communication at the first diagnosis provides a baseline on which to compare conversations about the cancer recurrence. We asked the following research questions:

RQ 1: What are patients’ recollections of their oncologists’ communication about their prognosis when first diagnosed with cancer?

RQ 2: What are patients’ recollections of their oncologists’ communication about their prognosis when told of a cancer recurrence?

METHOD

Sampling

These data are part of an exploratory study aimed toward generating a detailed description of the patient recurrence experience, particularly surrounding such communication tasks as discussing prognosis and decision making. Descriptive data were generated from intensive interviews with a cross section of patients referred to the study by their medical oncologists. All patient participants were diagnosed with distant (i.e., metastatic) solid tumor (i.e., breast; head and neck; lung; colon; gynecologic) recurrent cancer and were seeking treatment at a large, comprehensive cancer center. Because there are many unique types of cancer, we limited study participation to patients with tumor types that (a) typically recur metastatically, (b) are fairly prevalent, and (c) are not quickly debilitating. All potential participants were diagnosed with their first recurrence of cancer.² Other eligibility criteria required patients to be at least 18 years old, cognitively unimpaired, and having a previous diagnosis of stage I–III solid tumor cancer, followed by at least a 6-month disease-free interval. Further, the patients’ recurrence diagnosis was required to be within 3 months prior to enrollment. This study was approved by the institutional review board (IRB) committee associated with the cancer center.

Patient identification.

Cooperating oncologists (n = 30) on staff at an NCI-designated comprehensive cancer center identified recurrent patients from their own daily appointments, and provided the first author a referral to complete the eligibility assessment. Most of the oncologists are female (73%), and slightly less than half had treated the patient for the initial cancer episode. Forty-four (44) eligible patients (i.e., 3 males; 41 females) were called by the first author to invite participation in the study. All eligible patients were telephoned at least three times. There was no telephone contact established for 10 of this group. Four more who were reached by telephone either declined participation (one male), withdrew from the study prior to the interview (two females), or withdrew after the interview was completed (one female). The final sample included 30 female participants experiencing a solid tumor cancer recurrence. Although the study was not designed to focus on a female sample, the heavy prevalence of breast cancer recurrence and exclusion of prostate cancer recurrence resulted in just that.

Participants ages range from 42 to 84 years (M = 63.0 SD = 11.5). Most were experiencing a distant recurrence of breast cancer (23), and others reported diagnoses of recurrent lung (4), colon (1), gynecologic (1), or head and neck (1) cancer. Patient responses represent encounters with 11 medical oncologists, 73% of whom are female. Participants reported an average amount of disease-free time (i.e., neutral time) of 7.3 years (SD = 7.4), although 75% of the sample reported neutral time between 1 and 10 years.

Data Collection and Analysis

Following identification and eligibility confirmation, participants were interviewed by the first author in a private location of their choosing. Consent was obtained at the scheduled appointment, followed by a recorded interview. Interviews ranged in duration from 30 to 76 minutes (M = 48; SD = 12). The interview protocol consisted of patients being asked to “tell me the story of how you found out about your first cancer,” followed by “now can you tell me the story of how you found out about your recurrence?” This technique allowed participants to create a mental chronological framework for referencing events. We also found that the narrative task emphasized the patient’s informant role in the interview and encouraged patients’ active involvement, resulting in easy rapport. These cancer discovery narratives were followed by several open-ended questions about the recurrence experience, and communication with the patient’s primary medical oncologist. Interview questions were formulated in advance and informed by topics identified in recurrent cancer (e.g., comparison, prognosis) and health services literature (e.g., hope, decision making). Interview questions are displayed in Table 1.

TABLE 1.

General Interview Questions

1. Can you tell me the story of how you received the news of your cancer diagnosis the first time?*

2. So now you’ve been told that your cancer has returned. Can you tell me how you found out about your recurrence and what has happened since then?*

3. Did you learn anything in your first treatment experience that you are putting to use in this one?

4. Is the communication you are having with your doctor now what you expected?

5. How does your doctor talk to you about your prognosis?

6. How does your doctor help you stay hopeful?

7. Can you give me an example of how you and your doctor make decisions?

8. How did you build trust with your doctor? Are there specific things he or she does that makes you feel confident?

^*These questions are intended to prompt narrative accounts. Participants were also asked to elaborate on topics within the narratives that matched other domains shown here.

The interviews were digitally recorded and transcribed verbatim by a professional medical transcriptionist. An average-length interview (i.e., 48 minutes) produced 30 pages of transcribed text. We removed all personal identifiers and place names during transcription, and used pseudonyms to report individual participants’ responses. Coding was accomplished in two phases. First, an independent coder identified and compiled all references to prognosis in the interview. Second, the authors reviewed these quotes with an eye toward exploring potential themes. Each phase is explained in more detail in the following

An independent undergraduate student coder was trained to use the Atlas-ti (Atlas-ti, 1993–2006) program to identify transcribed text relevant to the research questions. Atlas-ti enables creation of a user-defined, personal computer (PC)-based coding system that organizes all the transcripts within one unit and allows easy text searching, tagging, and comparison. In this study, the coder’s task was to identify those segments of the transcribed interview that referenced prognosis discussions. Data reported here are generated from the interview question “How does your doctor talk to you about your prognosis” and comments about prognosis expressed in participants’ cancer discovery narratives. This phase of coding entailed listening to the audio while following along with the transcript, pausing to highlight (i.e., tag) potential interview segments for analysis. The first author worked closely with the coder to confirm the described cancer episode as referencing the initial or recurrent phase of disease. These segments of text (and surrounding contextual information) were extracted from the transcript and compiled within quotation files. Each quotation was automatically accompanied by additional data that included the source case number (later converted to pseudonym), coding date, and transcript line number. Prognosis quotation files were then analyzed by both authors who identified similar and contrasting themes. The authors engaged in an iterative analysis of each case, identifying shared themes, and categorizing data within their appropriate theme. The results and interpretation of these data are presented next, followed by discussion of the conclusions and implications of this study.

RESULTS AND INTERPRETATIONS

The first research question focused on patients’ recollections of their oncologists’ communication about prognosis when they were first diagnosed with cancer. Given the wide variability in neutral time across this group of patients, these events may have occurred several years prior to the interview date. Examination of the interview data revealed optimism about treatment and eventual cure as a recurrent theme in the initial cancer episode.

Communicating Optimism for Cure

When diagnosed with the initial cancer, many patients reported that doctors’ messages conveyed optimism that they would be cured. The participants’ comments describe an acute approach to treatment that focused on cure and a cancer-free future.³ These initial discussions about treatment can be characterized as confident, positive, and hopeful. The following comment exemplifies this finding:

Kathy: Well the first time I remember she just said “Well I’m going to cure you,” and she was like, “If you want stats and that, there are reputable places you can look it up online,” she’s like, “but everybody is different.”

During the initial cancer diagnosis, several patients relied on a 5- or 10-year timeline as a reliable rate for attaining cure. These familiar time spans appear to hold very strong significance to patients experiencing their initial cancer, and give them a tool for managing their uncertainty. For example, Michelle remarked, “You’re basically almost sort of told at five years that you’re cured, and you worry.”

Michelle went on to emphasize the importance of the five-year survival mark:

Two years ago I know I went in and I go “It’s five years. It’s five years. I’m so excited!” She [the doctor] goes, “You made it to five years.” I don’t know if she ever … I can’t remember if she said to me, “You are cured.” I don’t remember her saying that.

Some patients perceived an optimistic prognosis from the continued reassurance from their oncologists. This perception was bolstered by commonly known survival markers, such as the five-year timeline just described, and early detection of the initial tumor. The remarks of two different patients demonstrate this view:

Susan: … he came and said he felt he got it all and I should be okay. “Well, we’ll see what happens in the first year,” and then as each year went past we thought “I’m going to be fine.”

Lisa: The prognosis was fine, as far as they were concerned, because in both cases they had caught it very early and it was in an early stage and that, so … Back then I guess I just, even though my mother [who did not survive cancer] had been diagnosed very early, I just wasn’t thinking of death.

In fact, some patients’ expectations were so established that they were sometimes surprised to encounter evidence contradictory to their beliefs about cancer, including rapid deterioration and death. Sarah describes these expectations well, and spotlights the particular credibility other survivors hold:

Now I have to back up, ‘cause lung cancer in my mind was associated with death. It was a disease that was incurable. But because I’ve been around for a while, I’ve known people who’ve had it and I’ve talked to people who’ve had it, and then it had came up in casual conversations. Somebody said to me, “Well I’m a survivor,” and I said, “Really?” And then she told me that she had had it a year or two or three, whatever, and that Dr. C had done surgery, so at that point I already had some knowledge of lung cancer as not death, not equal to death.

In general, the initial cancer diagnosis introduced tremendous uncertainty into the women’s lives. This uncertainty was met by confidence for a cure, and focus on incremental survival goals. However, the longer they went cancer-free after treatment, the more likely that the uncertainty, while never absent, moved from the forefront of their consciousness to the background (duPre & Ray, 2008). This state was dramatically altered by the return of disease, as demonstrated in the communication reported by the women when told of cancer recurrence.

The second research question solicited patients’ recollections about their oncologist’s communication about their prognosis when their cancer recurred. Our analysis identified a primary theme reflecting a shift toward communicating chronic disease management.

Communicating Chronic Disease Management

Once they were told of the cancer recurrence, patients reported a shift from optimism about cure (i.e., acute model) to a more chronic disease management approach. Following recurrence, the disease was described as “manageable” instead of “curable.” Although patients knew that the recurrence significantly lessened the chance for a cure, many did not appear to fully realize the terminal implications of metastatic disease. Within this overall chronic disease management approach theme, we identified subthemes of: (1) disease discovery, (2) ambiguous communication, and (3) the information seeking paradox. These subthemes characterized changes in participants’ perceptions of communication with or from their doctors.

Disease discovery.

Some research suggests that patients who accept the possibility of recurrence following their treatment are less distressed and adjust more quickly at second diagnosis than those who believe recurrence is inevitable or will never occur (Weisman & Worden, 1986). Reaction to news of the recurrence varied. Some were prepared for a possible recurrence, like Ellen: “I sort of think I always thought it would happen, so I wasn’t at all surprised.” Others, however, seemed stunned by the news, like Lisa: “The rug was pulled out from under me, you know. When I went home, I felt like I was going home to get ready to die.” Tracy offered a more tempered reaction: “I was just shocked that I was already … I mean it’s not [like] I never thought it could come back.” These varied perceptions point to the need to examine various individual and situational factors within the recurrent disease context that may contribute to patients’ distress. In each case, disease discovery was memorable and appeared to trigger a familiar coping script that was quickly met by a new approach from providers toward the disease.

Shortly after receiving the recurrence diagnosis, patients found out that they were facing a different set of circumstances. Cure had now taken a back seat to ongoing “disease management.” This is evidenced in discussions of ongoing treatment strategies, alarming terminology, a sense of urgency, and a new model of living with cancer. Patients reported that at recurrence, doctors’ messages became less hopeful and more pragmatic.

Tracy: Manageable, that it’s not curable anymore, ‘cause it’s in the bone, but at least you can manage it for a long time…. Well they said that it should be manageable. He’s starting me on this treatment, and you know as long as I would react to it, it should … You know he’s going to give me this thing for the bone and hopefully it would shrink everything up, and this other medicine should heal my bone eventually.

The following comments reflect the challenge of this changing worldview. The recurrence discovery jolts the patient from a wide open world of many possibilities to a smaller, more finite place, bounded by disease characteristics and treatments, but, nonetheless, moving forward in time.

Ellen: You know I want to plan this. You know how do I deal with this thing here that is the rest of my life? And you know unfortunately, I think, and luckily, I think breast cancer bony metastasis has you know pushing the edge on eight, five, ten years I mean. So that’s great, but it’s still pretty finite for me. So the way she couches it, she kind of, we work in bits I think. We work in bits and bites, and it’s you know “Try a drug two to four years until you need to do something different.”

The striking challenge of adjusting to a chronic model of disease management is captured by these comments about the future. Fran considered her terminal status in light of her previously hoped-for timeline.

Fran: Now it’s not considered curable. It’s considered a chronic disease that, like, you know high blood pressure or diabetes, which are also not curable. So it’s a different … Now I can’t hope for a cure like I did. You know I was like rooting for reaching my 10 years, and every time I had my appointment it was like, Okay, I get to live another year. Now that’s out of the picture.

Margaret recalled being told the specific goal of treatment for her recurrence, but expressed less specificity toward the outcome.

“This is what we’re going to put you on, and our main goal is to hold this stable and allow it hopefully to shrink what’s there and not allow it to go into any of your major organs, because that could be a problem.”

Similarly, Donna expressed a view of the future that featured continuous treatment, but stopped short at the outcome.

You know like I said, take the treatments for the rest of my life, and they might be able to space them. I might not have to get them every month. She said [the treatments] could be every two months, but definitely she said I would have to take them you know the rest of my life, unless something new comes out.

Indeed, these perceptions suggest that at recurrence, the notion of “the rest of my life” is no longer defined by the patient’s age, but rather by the success of an ongoing treatment plan that offers only “stability” at best. Implications of this recalibrated time frame remain shrouded in uncertainty and may present a significant challenge to communication.

Ambiguous communication.

Ultimately, many recurrent patients come to the conclusion that managing cancer as a chronic disease may also suggest a worsened prognosis. Yet without much explicit discussion about their prognosis, these patients are left with a deeply uncomfortable ambiguity. Consider, for example, the clinical convention of staging a cancer based on its severity. This is a familiar concept to cancer patients and closely tied to prognosis during the initial cancer episode. However, because solid tumor cancers most often recur in a new location, recurrent disease is considered either “regional” or “distant,” rather than the more familiar stages I–III. Distant cancer is metastatic, known more familiarly as stage IV. The following remarks convey some of the personal meaning associated with cancer staging for recurrent disease. Although Sarah knew that stage IV indicated terminal status, she also realized the valuable time span for living that preceded death.

Sarah: They stage it at Stage IV. To me that’s equal to death. I mean I’m going to die tomorrow … But they stage it differently than I thought. I was familiar with this disease from people saying, “What stage is it?” Each stage means closer to death…. I think at [the cancer facility] it was more clearer to me. It was clearer to me that it was going to be slower. It was going to be death at the end, but much slower and I’d have more time in between for a normal life. It seemed that that’s the way they said it.

Linda was more pragmatic, and pointed out perhaps one of very few certainties, in her remark, “there is no stage V in this ballgame.” Ellen considered the personal meaning of both terms, recurrence and metastatic, drawing a less desirable outcome from the more clinically oriented word.

Ellen: And you know I think about recurrence and I think about metastatic disease, and to me those two things aren’t necessarily the same. So when I thought about having a recurrence, you know I guess I think about … Well that’s not really true I guess. I think about metastatic disease as being worse than a recurrence.

These comments suggest that an important task for people with recurrence is to consider the meanings they held for many of these terms, such as staging, recurrence, or metastasis. How patients consider and change these meanings may hold important implications for the patient’s adjustment to the new diagnosis and subsequent quality of life. These challenges may be rooted in divergent meanings held by providers and patients regarding such terms as “metastasis” or “manageable” (Dougherty, Kramer, Klatzke, & Rogers, 2009). The meaning of medical jargon is at the core of sensemaking processes in illness contexts (Castro, Wilson, Wang, & Schillinger, 2007; Safeer & Keenan, 2005), and appears to be salient during recurrent cancer as well.

The shift from acute curable status to chronic terminal disease also has strong implications for information sharing between doctors and patients. Generally, advanced cancer is characterized by great uncertainty surrounding the location of the illness, efficacy of the treatment, and expectancies for survival (Mishel & Braden, 1988; Stiegelis et al., 2004). This kind of uncertainty can have negative repercussions for both biological (e.g., blood pressure) and psychosocial (e.g., well-being) outcomes (Clayton, Mishel, & Belyea, 2006; Greco & Roger, 2003; Padilla, Mishel, & Grant, 1992). As these data below show, illness uncertainty may also have substantial effects on prognosis communication.

The information-seeking paradox.

Considering recurrent cancer as an incurable chronic disease presents patients with the dilemma of working with oncologists “bit by bit,” all the while facing their own mortality. This present/future paradox has implications for patient information seeking, particularly regarding their own prognosis. These data appear to present a difficult tension for patients between wanting to be involved in their care and knowing what to expect, versus not wanting to know the body of evidence associated with their condition. Clearly, patients must find a way to manage the uncertainties inherent in a recurrence diagnosis.

However, the strategy of working in “bits and bites” ensures a steady focus in the present, offering some measure of certainty and control. Alternatively, other patients may actively and selectively avoid information, to maintain uncertainty about their future (Brashers, Goldsmith, & Hsieh, 2002). Some research shows that people with high anxiety may be motivated to seek information but fail to retain it over time (Turner, Rimal, Morrison, & Kim, 2006). Although potential reasons for various information seeking strategies are numerous, it is clear from the data that managing these information preferences varies among patients. For example, some patients in this sample wanted to know as much as possible.

Carla: Me, personally, I’d like to know, ‘cause I have my daughter, my grandchildren, my husband. I want to … Yeah, I want to take care of things.

Others are more cautious, balancing their need to know with realization of the consequences of information gain.

Susan: You know like some people want to know how long they have to live and all that kind of stuff, and I don’t really want to know, you know. I mean am I going to die in six months? Well I don’t want to know that.

Tracy: Like if it’s this stage, what’s the five-year survival rate? I don’t want to look at that stuff, ‘cause I don’t want to, at this moment I don’t want someone telling me it’s limited.

The “prognosis dance”.

The differences in how patients seek or avoid information create a tension between physician and patient when discussing prognosis. Changing treatment and prognosis expectations require patients to reconstruct their vision of the future, yet there are few cues to signal or guide this shift. Importantly, there appears to be awareness on the part of several patients regarding what they want to know and when. For Ellen, this translated to controlling the information exchange process rather than the outcome: “It’s a dance; and she’ll tell me if I ask her, but I don’t really ask her. I did ask her once and I didn’t like the answer, so I don’t ask her anymore [laughter].” Similarly, Michelle tested the information seeking process and found what she needed to hear to anchor her hopes for the future.

Michelle: She basically, I think she wanted to tell me as little as possible to get me through to the next step, and I pushed a little. She said, “The first thing you’re going to do is get your ovaries out and go on this medication.” I’m sitting there going, okay, she doesn’t want to give away too much, I don’t think, but basically then she did say to me, “You have about a 60% chance that taking all the estrogen away will shrink and kill the cancer cells.” So that’s what I needed to hear. I needed to hear her say to me I had a better than 50% chance that this was going to die on its own and I would never have to go through chemotherapy again.

These comments suggest that knowing what “you need to hear” represents a meaningful certainty that can support hope. Identification of those bits of information that hold promise is an important process in a shifting situation. This activity may reflect patients’ transitioning the object of their hope from medical to more existential sources (Ballard et al., 1997).

An information-seeking tension is also evident in the idea that although every person’s cancer experience is unique, there is a large body of scientific evidence that describes and predicts how a given cancer will behave. Ellen and Beth’s remarks underscore their desire and belief to fall outside the norms of this evidence base.

Ellen: I know those numbers are not hard and fast, so of course I want to be the outlier like everybody else.

Beth: I don’t get real caught up in it. I mean you know I said [to my doctor], “You know you told me [initially that] 75% of people who have cancer at the stage I was initially diagnosed survived to the five-year mark.”

This tension leaves patients to negotiate the degree to which they will live their experience within the realm of individual possibility or the likelihoods presented by a body of scientific evidence. Beth describes her choice in spiritual terms:

Beth: Well what does it lead me to expect for the future? And he [the oncologist] said, “You can pretty much expect that you probably will have a recurrence about every 12 months,” … but you know I also look at I have people praying for me all over the country, and I have people saying to me, “You are going to be the miracle person,” and you know it’s hard to think that about yourself, but given my faith, I do believe in miracles, so I’ll take it.

These tensions can also be uncomfortable for clinicians who may resist making definitive claims.

Kathy: They’re really not saying a whole lot. They’re like, “We really don’t know. Everybody is different. It spreads differently.” And I remember the first time when we thought it was just in my liver and in the spot in my chest I said, “Okay, so tell me when … ” I said, “You know five years from now, ten years from now am I going to be here?” And I think just because I put it point blank like that, she said, “Five years from now, maybe. I mean there’s a chance you could be here.” She’s like, “Ten years from now, there’s no way.” So and since then, now that it’s spread more they’re being even … Everybody actually is being really vague. I mean anybody that I ask has not given me a time-frame. So I mean I think when I was in the hospital, I think I said to her and she just said, “There’s no way of knowing.”

The dance around prognosis can be described as a dance no one wants to lead. Patients and doctors consider each other carefully, searching for cues about how to proceed. Information seeking presents a potential risk to the patient’s self-concept, so these conversations appear to focus narrowly on specific concerns but remain colored with hopeful pledges and sentiments. At its core, discussing prognosis at recurrence requires negotiating multiple tensions between the threat of knowing more and the security of knowing less, the promise of medical treatment and suggestion of alternative sources of hope, and the idea that each cancer is unique, yet predictable.

CONCLUSIONS AND IMPLICATIONS

Recurrence is a unique point on the cancer trajectory that requires patients to reopen a difficult chapter of life they would most likely prefer to leave behind. In this sense the person with recurrent cancer is required to re-establish depersonalizing illness roles and routines (Radley, 1994). As such, it is quite possible that recurrent patients are operating under a set of expectations grounded in prior experience. Because recurrence is different from the initial disease, in terms of pathology, treatment, and prognosis, these expectations may present a significant challenge to oncologist–patient communication.

Theoretical Implications

Problematic integration (PI) theory is a useful theory for understanding the dynamics of uncertainty in prognosis communication (Babrow, 2001; Bradac, 2001; Brashers, 2001; Hines, Babrow, Badzek, & Moss, 2001). PI theory counters the idea that uncertainty is a target state that can be resolved with sufficient information, or that uncertainty is necessarily negative or, in fact, resolvable (Babrow, 2007). Instead, PI theory focuses on the association between the probability of an outcome and the person’s evaluation of that outcome. In this case, cure is a highly valued outcome during the initial cancer episode, and, as these data show, much oncologist–patient communication creates a high positive probability of this outcome. Conversely, at recurrence, although the resumption of treatment is familiar, the likely disease outcome is very different, essentially terminal. This new dilemma presents a distinct challenge to treating oncologists, who practice within an evidence-based medical paradigm (Bensing, 2000). Oncologists must balance empirical evidence with providing reassurance and support to the patient. Oncologists’ prognosis talk may communicate messages that actually increase patient uncertainty, and uncertainty may be preferred by patients when told of a recurrence (Ford, Babrow, & Stohl, 1996). For some, uncertainty may function as a coping mechanism, while for others it exacerbates their stress (Babrow, Hines, & Kasch, 2000).

Physician–patient communication serves as a primary mode for shaping and modifying both probabilistic and evaluative orientations toward recurrent cancer. Perhaps the core task of communication at cancer recurrence requires this process of reintegrating changing probabilities and values (Babrow, Hines, & Kasch, 2000; Hines et al., 2001). For example, a person with recurrent cancer, in adjusting his or her perceived probability of disease free survival, may begin to devalue aggressive treatment in exchange for more highly valued quality of life. Communication with the oncologist offers the ultimate reality check regarding a patient’s state of health. Along these lines, the information-seeking paradox will remain an awkward dance unless patients and physicians uncover a communicative path for expressing their concerns and making informed decisions.

An additional consideration should be given to language convergence/meaning divergence (LC/MD) theory (Dougherty et al., 2009). This theory offers a framework for predicting potentially adverse effects from meaning divergence in oncologist–patient communication. Dougherty et al. describe the “illusion of shared meaning” (p. 39) that may result in the event that communicators perceive different meanings in the face of shared vocabulary or language. This is particularly relevant to recurrent cancer patients who refer to a prior cancer experience for sensemaking but find original meaning clusters may be different because of changes in the disease. Unless oncologists and patients make a clear effort to address potential changes in meaning from the initial to recurrent episode, care goals and decisions may be problematic.

Practical Implications

This research presents several practical implications for both research and clinical practices. First, these results suggest that breaking bad news is an interactive process (Gillotti et al., 2002). Much bad news research focuses on the delivery mechanics of bad news, and the impact of illness stage, rapidity of progression, etc. Researchers have not carefully examined patients’ perceptions of changes in their doctor’s communication that may occur as a prognosis changes. Now researchers need to design longitudinal studies that capture the process of illness adjustment, and related changes in communication.

We have some evidence that curable early stage illness is met by optimistic confidence and the cushion of five year timelines. When cancer returns, there appears to be less doctor–patient communication about prognosis and narrow focus on potential treatments. However, patients are still processing information and concluding meaning about their condition. For example, patients’ sensitivity to subtle changes in communicated optimism or increased ambiguity may have significant implications for doctor-patient relationships. Mahon and Casperson (1995) reported that cancer patients perceived clinical staff to become more businesslike at recurrence, and that staff assumed they were emotionally “OK.” Other research underscores the importance patients place on communication with their doctors when initially diagnosed or when diagnosed with a recurrence (McWilliam, Brown, & Stewart, 2000). These perceptions may be tied to increases in communicated ambiguity and the “prognosis dance.”

Additionally, future research of communication during chronic or life threatening illness should explore potential dialectical tensions that may result as a by-product of problematic integration. Communication researchers have successfully identified several dialectical tensions relevant to interpersonal and intimate relationships (Baxter, 1990). It is unknown how the information-seeking paradox may affect the oncologist–patient relationship, but it is possible that divergent meaning and unattended tensions may have multiple negative effects, including decreased compliance, greater decision regret, and suboptimal quality of life. Further, carefully designed interventions that serve to modify patients evaluations of the recurrence prognosis or modify oncologist’s predictions of patients’ reactions to prognosis discussions may lead to an interactive model of problematic integration that accounts for the impact of shared communication on individual uncertainties. Dyadic process models, similar to those depicting deception or adaptation, may provide a useful template for developing an interactive model of problematic integration (Buller & Burgoon, 1996; Burgoon, Stern, & Dillman, 1995).

Raising awareness of these issues and identifying crucial intervention points may help doctors and patients overcome their unwillingness to discuss prognosis in concrete terms. For example, communication skills training for patients has been shown to help doctors and patients reach a shared understanding regarding desired content and amounts of information (Cegala, 2003, 2006). Future interventions could enable patients to express their preferences for discussing prognosis through tools that facilitate meta-communication about bad news.

Limitations

Though unintended, all of the study patient participants are female. This outcome reflects the over-representation of breast cancer patients in the study. This female sample represents a distinctly gendered view of communication and cancer experience, which may have influenced findings regarding information exchange, preferences for involvement, and the development of physician–patient relationships (Butow, Maclean, Dunn, Tattersall, & Boyer, 1997; Gabbard-Alley, 1995; Harrison, Maguire, & Pitceathly, 1995). The fact that the few men who were approached for the study either avoided contact or refused participation may indirectly support this claim. Our future projects are designed to integrate gendered patterns of communication into their analyses. Further, our participants volunteered to share their experiences, which may indicate a greater tendency to communicate about their health in general. This idea raises the possibility that participants’ recollections may be biased or contain other inaccuracies. Data derived from narrative methods can certainly be viewed as participants’ reconstructions of past events (Hydén, 1997). While we are confident that participants articulated their illness experiences authentically, our focus is intent on identifying important aspects of experiencing recurrence and discussing prognosis. Although participants’ memory for specific details may have deteriorated, core themes are generalized and less vulnerable to small inaccuracies (van Heugten, 2004; Woodgate, 2005). We plan to further assess construct validity with additional methods in future studies (Sharpley & Christie, 2007).

Additionally, all of these participants shared a similar prognosis and were at approximately the same point from diagnosis when they were interviewed. Although these criteria were intentionally set in order to provide some control of disease factors that potentially influence the recurrence experience, they produce a narrow slice of patient experience. We would expect to find different perceptions of physicians’ communication among those experiencing a more benign cancer recurrence, multiple recurrences, or transitioning to palliative care. Along these same lines, all participating patients received care at an NCI-designated comprehensive cancer center. Patients receiving care at private oncology practices or community hospitals may encounter different practices or styles of communication rooted in a unique organizational culture. Despite these limitations, the results of this study offer compelling evidence for the importance of patients’ perceptions of and communication about their prognosis.

Summary

This study explores the perceptions of a small group of women newly diagnosed with a cancer recurrence. The participants’ recollections of communication with their doctors offer a small glimpse into the difficulty of resuming the role of cancer patient after a period of disease-free survivorship.

Doctor–patient communication at cancer recurrence can reflect a nuanced dance involving clinical information, conveyed support, patient preferences, family caregiver dynamics, and transformed expectations for both the patient and doctor. Conversations are deeply lined with uncertainty and potential tensions that need to be successfully managed. As Babrow, Hines, and Kasch (2000) note, “Much remains to be learned about how illness-related uncertainty may be most effectively managed, but one thing is clear: To be effective, illness explanations must be designed to manage uncertainty” (p. 43). This suggests eliciting patients’ preferences about discussing prognosis soon after diagnosis. What this communication looks like, how it changes over time, and how it enables or constrains doctors’ and patients’ needs are empirical questions with important pragmatic implications. This study begins to address these issues and the crucial role of communication regarding a changed prognosis during a recurrence of life-threatening illness.