Skip to main content
Free access
Art of Oncology
May 01, 2005

Necessary Collusion: Prognostic Communication With Advanced Cancer Patients

Publication: Journal of Clinical Oncology


As a young oncologist, I find myself wrestling with many of the most difficult issues in the practice of oncology: making difficult trade-offs between risk and benefits of arduous treatments, finding sources of strength to deal with the devastating realities of patients' situations, and determining how to communicate important information to patients in ways that give them honest and useful information, but still allow them to preserve some hope for the future. It is this last issue on which I offer my thoughts here.
I have noticed that although common wisdom holds that we should give accurate and honest prognostic information to patients at all times, in practice this is hard to do. So, I find myself engaging in a process of communication about prognosis with patients which, in a sense, allows the cold, hard facts to come out over time, as opposed to presenting them up front. I have found other oncologists who seem to use a similar strategy. I do this because it is a style of communication which, I think, “works” for me and my patients: patients come to deal with the painful knowledge of their ultimate outcomes over time in ways that I think allow them to cope along the way. But I have continuously asked myself whether this is the “right” way to communicate. Is such a strategy of communication ethical?
Looking to the literature about prognostic communication with patients with advanced cancer, two lines of thought dominate. The first suggests that patients want accurate estimates of their prognosis and that accurate estimates benefit patients by allowing them to make end-of-life plans consistent with their values.1-3 The second suggests that patients desire, above all, to maintain hope for their situations, and that the “management of hope”4 permits the prognosticating oncologist to take some liberties with prognostic estimates in order to preserve the patient's ability to maintain hope.
These two lines of thinking appear to be at odds with one another. Is there, however, an ethical way of resolving the tensions between these two views? In the end, I believe that except in instances where stark frankness is openly requested, a style of communication that allows patients to dictate most of the flow of prognostic information or to avoid it is an ethical strategy of prognostic communication. I will refer to this strategy as “necessary collusion.” By collusion I mean the spoken or unspoken agreement we sometimes enter into with some patients to avoid or delay discussing a definitive, numerical prognostic estimate of life expectancy. This article is an examination of whether such collusion is ethical. I am going to argue that such collusion can be ethical, because of certain special characteristics of the situation of patients with advanced cancer and their relationships with their oncologists.

Collusion in Doctor-Patient Communication About Prognosis

One group5 has attempted to evaluate what the authors called “false optimism about recovery” among patients with small-cell lung cancer, by following 35 patients intensively through the course of their illnesses. The authors report that in discussing clinical care, including prognosis, doctors and patients communicated in ways that the investigators characterized as collusive. According to the authors, this was done so that optimism could be retained through much of the treatment course until patients began to deteriorate physically themselves or saw other patients around them do so. The collusion they describe usually centered on shifts in the conversations between oncologists and patients, from the bad news of diagnosis and prognosis to the good news of treatment planning. Because much of their course of illness was defined by the treatment calendar, patients infrequently asked about prognosis and physicians rarely offered such information, assuming that when patients did not ask, they did not want to know. The authors of this study portray collusion as a strategy by which oncologists and patients avoid the truth about prognosis. The truth about prognosis is seen in this study as being “swept under the rug”—just beneath the surface, but not overtly acknowledged. Other studies have noted similar collusive tendencies in doctor-patient communication in oncology.6-8
Some might argue that such collusion on the part of the oncologist is not honest, and thus not ethical. I believe, however, that this kind of collusion can be justified ethically. In fact, for many patients I would argue that it is necessary. This “necessary collusion” stems from three assertions: (1) it is consistent with many patients' stated and unstated wishes, (2) it preserves hope by acknowledging the initial uncertainty surrounding prognosis in patients with advanced cancer, and (3) it allows prognostic information to emerge over time in ways that take advantage both of the progressive degree of prognostic certainty that emerges over time and of the increasing receptiveness to prognostic information that patients experience as their physical situation worsens. I will examine each of these assertions in turn.

Collusion Is Consistent With Patients' Wishes

Many published studies9-11 suggest that the overwhelming majority of cancer patients want to know everything about their disease, including prognosis. These studies have been used to justify the belief that patients with advanced cancer all want accurate prognostic information all of the time. The most important problem with many of these studies is that they conflate the information needs of patients treated initially with curative intent (“curable patients”) and those initially treated without curative intent (“incurable patients”). The conceptual flaw in such a study design is that curable patients receive and perceive prognostic information (including the chances that they will be cured) in totally different ways from those who have incurable disease. The conclusions from such studies of information preferences among cancer patients must therefore be interpreted with caution.
By contrast, a different picture emerges from published studies that focus on terminally ill cancer patients. In Cassileth et al's12 early study of information and decision-making preferences among metastatic cancer patients, fewer than 60% of patients wanted to know what the likelihood of cure was. In a study of information needs in terminal illness (60% of study subjects had cancer), Kutner et al13 found that, although 100% of patients stated they wanted honesty from their doctors, 71% of patients wanted to know the likelihood of cure, and only 60% of patients wanted to know their overall life expectancy, fewer than the number of patients who wanted to know if they would be able to eat later in their illness. Kaplowitz et al's study14 of both curable and incurable cancer patients found that a large majority of the patients interviewed wanted a qualitative prognosis, but that fewer than half wanted a quantitative estimate, and that those with poorer prognoses were less likely to request prognostic information. Hagerty et al15 found that only 59% of responding patients wanted to be told how long they had to live when first diagnosed. Eleven percent of patients in this study preferred never to discuss dying and palliative care, and 21% indicated that they only wanted the information if they asked directly.
Although still representing a majority of patients, these studies of terminally ill patients suggest that many patients may prefer not to know, or to delay knowing, difficult-to-hear information, and that complete respect for patients' abilities to make their own decisions dictates that an assessment of patients' actual and current desires (which may change over time) be attempted. Freedman16 has even suggested that an ethical approach to prognostic disclosure when the wishes of a patient are not known is to “offer truth,” and then to respect the answer a patient gives. Collusion thus allows patients' wishes, not physicians’ preferences, to dictate the flow of prognostic information.

Collusion Preserves Hope by Acknowledging Uncertainty

I think everyone would agree that the preservation of hope is a vital goal for patients with advanced cancer. Many studies confirm this,4,17-21 and several studies demonstrate that the ability to maintain hope through avoidance of prognostic information may be important.22-24 But one question is, how can oncologists communicate in a way that is honest and yet preserves hope? Oncologists can always be honest, for example, by responding to direct questions as honestly and specifically as possible. However, I also think that an honest answer requires that we take the extreme variability of cancer biology and of cancer patients into account. When a patient asks, “Doc, how long have I got? I want the truth,” we should answer with as much specificity as one can honestly communicate, recognizing that specificity is not always possible. But when patients do not ask for specifics or request not to be told a specific numerical prognosis, it seems wiser to avoid devastating (and often inaccurate) time frames until there are clear indications that a patient is ready to hear them. Although it may sound as if I am advocating a controlled withholding of information from patients, I am really arguing for allowing the honest uncertainty that surrounds disease outcome and prognosis to be marshaled into the service of a patient's sense of hope and open-endedness.21,25 Allowing patients to come to terms with their terminal prognosis over time respects the tremendous emotional pain that accompanies the knowledge of a terminal prognosis.

Collusion Preserves Hope by Allowing Information to Emerge Over Time

In thinking about this issue, it has also occurred to me that several features of the situation of patients with advanced cancer, and of their relationship with their oncologists, may be used to facilitate the preservation of hope in an honest way. First, prognostic certainty for most patients is nearly impossible. However, an individual's prognosis may be refined over time with observation of the natural history of the cancer. Second, with time often comes an increasing receptiveness on the part of patients to devastating prognostic information when their disease has made them progressively more ill. Finally, patients who may have a difficult time believing their prognosis accept such devastating information better from a physician whom they trust and with whom they have built a relationship. All three of these factors support prognostic disclosure over time as opposed to handing down a specific numerical prognosis from the outset.
But how can oncologists help to guide patients through the emotional pain of terminal illness without misleading them or hiding the truth? I have found an answer in adopting a strategy of communication known as “forecasting” instead of a strategy of offering specific numerical estimates of life expectancy.

Forecasting in Incurable Cancer Patients

The sociologist Douglas Maynard26 wrote about a concept he terms “forecasting” in the delivery of bad news. This concept, which derives from empirical studies of social interactions surrounding the delivery of bad news, involves the provision of “some warning that bad news is forthcoming without keeping the recipient in a state of indefinite suspense (stalling) or conveying the bad news abruptly (being blunt).” As Maynard explains, the collusion that develops through forecasting between the deliverer and the receiver leads to a “deeply collaborative, orderly achievement,”26 which makes the bad news more tolerable to the receiver.
Kodish and Post4 present a useful discussion of such a strategy focusing on balancing hope and realism differently at different times during the clinical course of a patient's care. I would agree and argue that understood more broadly Maynard's concept of forecasting should be the dominant pattern of prognostic communication with cancer patients facing life-ending diagnoses. Prognostication emphasizes numerical estimates of a patient's life expectancy, a process which includes uncertainty of estimate, but no uncertainty of outcome. Forecasting, by contrast, is a strategy in which the future is predicted, though without absolute certainty, leaving room for the uncertainty that surrounds every patient's eventual outcome after diagnosis. As successive events are realized (“The chemo will eventually stop working” or “A cancer's response to a second or third line of treatment is usually shorter than to the first line”), patients are in Maynard's sense led to their own understanding of their prognosis, anticipating negative events in advance. Forecasting allows both the oncologist and the patient to face and prepare for likely outcomes, and to maintain hope about unlikely ones.


One of the strengths of using forecasting instead of a more narrow view of prognostication for those patients who do not desire a precise, numerical estimate is that it preserves a sense of open-endedness about eventual outcomes. Some will argue that leaving such uncertainty about eventual outcomes can lead to false hopes about the possibility of being cured. However, as Slevin et al27 have argued, the hope that patients with terminal cancer experience is often not related to false hopes of cure, but rather to the need to feel that life still retains some open-endedness, and that it is this open-endedness that gives people meaning in life.
Fulfilling the physician's obligations to tell the truth under such circumstances seems impossible, on the surface. How does one simultaneously leave a patient with a sense of open-endedness about their terminal cancer and inform them about their life-ending disease in a way that is useful for deciding how best to spend the time they have remaining? One solution lies in presenting a clear-eyed view of what is likely, unlikely, and possible. Because of the inherent uncertainty and biologic variability of cancer and of individual cancer patients, we can honestly tell our patients that a disease is curable or incurable, provide a range of prognoses, including outlying examples, and make certain that patients understand what the goals of therapy are. This kind of hope, hope that recognizes the obstacles and odds, is what Groopman28 calls “true hope,” which he distinguishes from false hope. In his view, false hope is the pervasive belief that everything will turn out all right.

Prognostic Communication Within the Doctor-Patient Relationship

I have tried to argue that prognostic communication with patients with advanced cancer must, at times, privilege open-endedness and uncertainty over certainty in prognosis because of the crucial role hope plays in the lives of patients with advanced cancer. This strategy ultimately relies on a metered series of conversations during which the technique of forecasting is used to bring about incremental awareness of the ultimate course and outcomes of the disease, and on careful and honest maintenance of hope through the preservation of open-endedness. But how can such collusion be necessary, by which I mean ethically imperative? The answer lies in an examination of the healing relationship between oncologists and cancer patients.
The critical but challenging process of communication about prognosis to patients with life-ending diagnoses is intimately linked to the central process of doctoring: the act of healing. But how can patients with life-ending cancer diagnoses be healed since, by definition, their diseases are incurable? A distinction between health and disease proposed by Kass29 has helped me to think about the goals of medicine and resolve this apparent contradiction. In short, Kass argues that the proper goal of medicine is health, and not cure of disease. Although we clearly cannot cure advanced cancer patients' disease, we can fundamentally contribute to their healing through the care we provide and the relationship we sustain with patients facing life-ending diagnoses. This important effect, though difficult to characterize (Kass calls it “magical and shrouded in mystery”) is real. All of us who look after patients have witnessed the relief and security that come on patients when they offer themselves to our care. This is what they mean when they tell us, “I'm in your hands.” Oncologists contribute specifically to their incurable patients' health by serving as a source of hope. Although not commonly recognized, oncologists become a locus—sometimes the locus—of hope for their patients.20,21 This is an awesome responsibility but I have come to believe that it is one of the core acts of healing we perform. For us as oncologists, therefore, the preservation of hope is an ethical obligation, one that follows from the nature of what we do.


I began with the question of whether the strategy that I, and probably many other oncologists, use to communicate about prognosis can be justified ethically. I have argued that patients with advanced cancer, like all patients with cancer, have a vital and inalienable right to maintain hope. I have also argued that oncologists, because of the unique position and function they serve in the lives of their patients, represent hope to their patients, and that this fact obligates them morally to find ways to help patients preserve hope. Nevertheless, they must also be honest.
Faced with such a situation, can we ethically come to a tacit or explicit agreement with patients (which I have called “collusion”) to avoid dire and specific numerical prognostic estimates until patients ask for them or appear ready to hear them? Recognizing that the natural history of any individual's cancer is extremely variable, I think we can justifiably and honestly outline a range of prognostic estimates and include descriptions of outcomes that are unlikely—even highly unlikely—as long as such outcomes are within the range of clinical possibility. This honest prognostic picture, which forecasts the future while including some open-endedness about the outcome, is the means by which hope may be maintained for patients.
Over time, we should refine these prognostic estimates, delivering information through a measured series of “forecasts” about the future. As each successive forecast is realized, patients gain confidence in their oncologist, and the inevitable outcome of their disease is revealed to them. Through the course of this, patients become engaged in a beneficial, healing relationship that such efforts to preserve hope engender in patients. This kind of “collusion” is necessary for some patients because it serves a vital function (the preservation of hope), and because the preservation of hope is at the core of oncology.
Editor's Note: Two commentaries follow on pages 3151-3152 and 3153-3154.

Author's Disclosures of Potential Conflicts of Interest

The author indicated no potential conflicts of interest.


I thank Melissa Cavaghan, MD, for her thoughtful comments on earlier versions of this manuscript.
Author's disclosures of potential conflicts of interest are found at the end of this article.


Haidet P, Hamel MB, Davis RB, et al: Outcomes, preferences for resuscitation, and physician-patient communication among patients with metastatic colorectal cancer: SUPPORT Investigators—Study to understand prognoses and preferences for outcomes and risks of treatments. Am J Med 105::222,1998-229,
Lamont EB, Christakis NA: Complexities in prognostication in advanced cancer: “To help them live their lives the way they want to. ” JAMA 290::98,2003-104,
Weeks JC, Cook EF, O'Day SJ, et al: Relationship between cancer patients' predictions of prognosis and their treatment preferences. JAMA 279::1709,1998-1714,
Kodish E, Post SG: Oncology and hope. J Clin Oncol 13::1817,1995,
The AM, Hak T, Koeter G, et al: Collusion in doctor-patient communication about imminent death: An ethnographic study. BMJ 321::1376,2000-1381,
Kirwan JM, Tincello DG, Lavender T, et al: How doctors record breaking bad news in ovarian cancer. Br J Cancer 88::839,2003-842,
Reich M, Mekaoui L: La conspiration du silence en cancerologie: Une situation a ne pas negliger. Bull Cancer 90::181,2003-184,
Ruddick W: Hope and deception. Bioethics 13::343,1999-357,
Davidson JR, Brundage MD, Feldman-Stewart D: Lung cancer treatment decisions: Patients' desires for participation and information. Psychooncology 8::511,1999-520,
Fallowfield L, Ford S, Lewis S: No news is not good news: Information preferences of patients with cancer. Psychooncology 4::197,1995-202,
Meredith C, Symonds P, Webster L, et al: Information needs of cancer patients in west Scotland: Cross sectional survey of patients' views. BMJ 313::724,1996-726,
Cassileth BR, Zupkis RV, Sutton-Smith K, et al: Information and participation preferences among cancer patients. Ann Intern Med 92::832,1980-836,
Kutner JS, Steiner JF, Corbett KK, et al: Information needs in terminal illness. Soc Sci Med 48::1341,1999-1352,
Kaplowitz SA, Campo S, Chiu WT: Cancer patients' desires for communication of prognosis information. Health Commun 14::221,2002-241,
Hagerty RG, Butow PN, Ellis PA, et al: Cancer patient preferences for communication of prognosis in the metastatic setting. J Clin Oncol 22::1721,2004-1730,
Freedman B: Offering truth: One ethical approach to the uninformed cancer patient. Arch Intern Med 153::572,1993-576,
Back AL, Arnold RM, Quill TE: Hope for the best, and prepare for the worst. Ann Intern Med 138::439,2003-443,
Benzein E, Saveman BI: Nurses' perception of hope in patients with cancer: A palliative care perspective. Cancer Nurs 21::10,1998-16,
Brockopp DY, Hayko D, Davenport W, et al: Personal control and the needs for hope and information among adults diagnosed with cancer. Cancer Nurs 12::112,1989-116,
Delvecchio Good MJ, Good BJ, Schaffer C, et al: American oncology and the discourse on hope. Cult Med Psychiatry 14::59,1990-79,
Miyaji NT: The power of compassion: Truth-telling among American doctors in the care of dying patients. Soc Sci Med 36::249,1993-264,
Butow PN, Maclean M, Dunn SM, et al: The dynamics of change: Cancer patients' preferences for information, involvement and support. Ann Oncol 8::857,1997-863,
Leydon GM, Boulton M, Moynihan C, et al: Cancer patients' information needs and information seeking behaviour: In depth interview study. BMJ 320::909,2000-913,
Wenrich MD, Curtis JR, Ambrozy DA, et al: Dying patients' need for emotional support and personalized care from physicians: Perspectives of patients with terminal illness, families, and health care providers. J Pain Symptom Manage 25::236,2003-246,
Waitzkin H, Stoeckle JD: The communication of information about illness: Clinical, sociological, and methodological considerations. Adv Psychosom Med 8::180,1972-215,
Maynard DW: On “realization” in everyday life: The forecasting of bad news as a social relation. Am Sociological Rev 61::109,1996-131,
Slevin ML, Stubbs L, Plant HJ, et al: Attitudes to chemotherapy: Comparing views of patients with cancer with those of doctors, nurses, and general public. BMJ 300::1458,1990-1460,
Groopman JE: The anatomy of hope: How people prevail in the face of illness (ed 1) . New York, NY, Random House, 2004
Kass L: Toward a more natural science: Biology and human affairs . New York, NY, Free Press, 1985

Information & Authors


Published In

Journal of Clinical Oncology
Pages: 3146 - 3150
PubMed: 15860876


Published in print: May 01, 2005
Published online: September 21, 2016


Request permissions for this article.



Paul R. Helft
From the Department of Medicine, Division of Hematology/Oncology, the Indiana University Center for Bioethics, and the Regenstrief Institute for Health Care, Indiana University School of Medicine, Indianapolis, IN


Address reprint requests to Paul R. Helft, Division of Hematology/Oncology, 535 Barnhill Dr, RT 473, Indianapolis, IN 46202; e-mail: [email protected]

Metrics & Citations




Article Citation

Download Citation

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

For more information or tips please see 'Downloading to a citation manager' in the Help menu.


Download article citation data for:
Paul R. Helft
Journal of Clinical Oncology 2005 23:13, 3146-3150

View Options

View options


View PDF

Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Personal login Institutional Login

Purchase Options

Purchase this article to get full access to it.

Purchase this Article


Subscribe to this Journal
Renew Your Subscription
Become a Member







Share article link