In localized prostate cancer, definitive therapy includes radical prostatectomy, external beam radiation, brachytherapy, and cryotherapy. Optimal treatment, however, is unknown because there is a lack of convincing evidence regarding which form of therapy offers the best long-term survival. Moreover, all forms of therapy have unique and potentially serious side effects. Indeed, despite at least one recent report,1 questions remain about whether the morbidities associated with radical prostatectomy are worth the small potential benefits in overall survival when compared with watchful waiting, especially for low- and intermediate-grade disease.2, 3 Ultimately, decisions about how to treat localized prostate cancer are highly personal. Ideally, each patient would base his decision on accurate information, weighing all options in terms of the likelihood for tumor recurrence, personal preferences for avoiding specific treatment-related side effects, and pragmatic considerations such as cost, convenience, and requirements for follow-up care.
Little is known about how men diagnosed with localized prostate carcinoma choose their treatment. Studies using hypothetical patients, retrospective data, or non-U.S. populations have examined discrete influences on and correlates of decision-making, such as social supports, informational sources, demographic characteristics, and concerns about side effects and posttreatment quality of life.4–19 Some of these studies have utilized focus group and individual patient interviews to aid in the development of surveys and psychometric instruments but most have asserted conclusions without reporting the content of the actual interviews. There is also a small literature based on testimonials of predominately white, socially privileged men that recount odysseys of receiving a prostate cancer diagnosis, agonizing over what to do, and traversing perplexing health care systems.20–22 Because the topic has not been adequately addressed in the prior literature this study aims to describe, from the perspective of working- and middle-class patients themselves, how they make treatment decisions in real time following the diagnosis of localized prostate carcinoma.
- Top of page
- MATERIALS AND METHODS
We carried out detailed interviews of 20 patients with localized prostate cancer while they were still contemplating treatment but before they had initiated it. We recorded salient aspects of their decision-making experiences in real time, thereby avoiding recall bias. To our knowledge, no previously-published studies have used this methodology. On the basis of thematic analysis, we elucidated three factors that exerted particularly strong influences on initial treatment preferences. The first was profound fear and uncertainty, often corresponding with a desire to receive treatment as quickly as possible. In fact, a majority of patients were uninterested in obtaining a second opinion, typically because of concerns that this would delay treatment and increase uncertainty. Also, several patients erroneously declared that the only purpose of a second opinion is to confirm a diagnosis. Second, most patients had influential misconceptions about treatment, especially prostatectomy. Their beliefs were highly polarized: one group of patients avowed that prostatectomy is the best way of guaranteeing cure while another group asserted that it is very drastic or dangerous. Finally, almost all patients relied on anecdotes—stories about other people's cancer experiences—to make sense of their own diagnosis and treatment choices. For the most part, these stories did not accurately match patients' own clinical circumstances.
Chart reviews and follow-up interviews revealed that most patients received treatments that matched their initial preferences. In addition, at the time of their follow-up interviews most patients invoked the same flawed ideas about prostate cancer and the same rationale for treatment that they had articulated during their initial interviews 6-8 months earlier. Although we were unable to directly determine the factors that actually influenced their final treatment choices, the persistence of these ideas bolsters our concern that many patients made early and quick decisions on the basis of erroneous and incomplete information, and then never changed their minds about treatment or altered the reasons for their choices. Altogether, these findings raise doubts about the quality of the decision-making process. Although the quality of an actual treatment decision in localized prostate carcinoma cannot, in general, be assessed in relation to an optimal form of treatment, the quality of the decision-making process is compromised when decisions are precipitous, and when they are based on significant misunderstandings and a lack of relevant knowledge.
Many of the patients we interviewed seemed as interested (and sometimes more so) in eliminating fear and minimizing uncertainty over the short term as they were in realizing cure over the longer-term. Even though most patients volunteered that prostate cancer is “slow growing,” this abstract knowledge did little to dispel the vividly frightening, yet unlikely prospect of prostate cancer suddenly “exploding.” Patients remained anxious and wanted to proceed with treatment as quickly as possible. Fear of death was much more prominent than fear of treatment side effects—to such an extent, in fact, that almost half the patients discounted concerns about treatment side effects altogether. In comparison with the possibility of death (which was greatly exaggerated), side effects were simply irrelevant. Another source of fear, again among nearly half the patients, was lack of definitive knowledge about the nature of the tumor. These patients believed—erroneously—that surgery furnishes the greatest amount of clinically useful knowledge and allows one to know quickly and with certainty whether cure has been achieved. Among a separate group of patients, on the other hand, prostatectomy induced fear because its dangers were magnified out of proportion to its real risks. In summary, although fear is to be expected and may be necessary for motivating serious deliberation and action, it overpowered reason in ways that were not rectified through the simple provision of accurate health information.
In the decision science literature, an early, primitive reaction to almost any personally relevant object or event is a dichotomous good-bad evaluative assessment. Such a reaction may help to winnow down large choice sets into smaller numbers of options that are more easily contemplated.34 Polarized attitudes for and against surgery may exemplify this phenomenon. Patients' first and most important step in weighing treatment alternatives seemed to depend on their beliefs about the intrinsic merits and disadvantages of prostatectomy, specifically. Assessments of prostatectomy, however, were often based on powerful misconceptions (see Table 2). Misconceptions about cancer are common and often go unrecognized even though they may exert a profound effect on patient behavior.35, 36 Although urologists (as in this study) typically present patients with factual information that, in principle, might correct their misconceptions, it is likely that patients often fail to recognize that such information is inconsistent with their preexisting beliefs. In this way, their misconceptions persist. Alternatively, they may simply choose to discount such information.37
When making treatment recommendations, physicians often rely on personal opinion and anecdote more than on high-quality evidence.38 It is unsurprising that patients also rely frequently on anecdotal reasoning rather than statistical information.39, 40 In this study, the vast majority of patients reported the influence of anecdote when forming treatment preferences. In adversity, patients seek advice from other people who are known or trusted and who have lived through seemingly similar situations. This is the basis of support groups and is a key way of garnering emotional and practical assistance from others. Anecdotal information is especially important when it provides appropriate reassurance about particular forms of treatment (e.g. that most patients, in fact, do well with prostatectomy). Conversely, anecdotal information impairs decision-making quality when it leads patients to ignore treatment alternatives or make decisions based on inaccurate appraisals of their own clinical circumstances.41
In recent years, the literature on patient-centered decision-making has emphasized the importance of informed decision-making (IDM) and, to a somewhat lesser degree, shared decision-making (SDM).42–45 IDM in prostate cancer occurs when a patient understands the nature of the disease, the risks and benefits of various treatment alternatives, and makes a decision consistent with his values and outcome preferences. SDM is an extension of IDM in which the clinician assumes an active role in helping the patient to clarify and express his values and preferences, proposes a form of treatment that seems to be most consistent with the patient's values and preferences, and collaborates with the patient to arrive at a mutually satisfactory treatment decision. Both IDM and SDM emphasize the centrality of communicating medical information to patients and encouraging patients to clarify their values and preferences in ways that can be represented as declarative facts. Also, IDM and SDM tend to emphasize the presentation and framing of numerical risk information in ways that patients can easily and correctly interpret.46, 47
This study illustrates that although attention to health information, outcome preferences, and the framing of numerical risk is necessary, it is hardly sufficient for achieving quality in patient-centered decision-making. In particular, traditional models of IDM and SDM often give short shrift to the three powerful influences on decision-making that we identified here. First, emotions such as fear and anxiety engender a profound desire for rapid treatment that is not alleviated by intellectual knowledge that prostate cancer, for example, is “slow growing.” Second, influential misconceptions about cancer and treatment are commonplace and are not easily redressed by simply imparting accurate health information. Finally, there is a natural propensity to base personal decisions on anecdotal information, even when anecdotes are not germane to patients' own circumstances. In emphasizing the importance of rational decision-making, IDM and SDM approaches implicitly stress how patients “should” behave while often overlooking how they, in fact, do behave.
To address these challenges, clinicians should strive to meet patients closer to where they actually begin the decision-making process. Explicitly describing common misconceptions may allow patients to consciously recognize and correct them. Sources of patient fear can be elucidated and, where possible, reassurance can be offered. Patients can be asked to recount stories about other people that influence their attitudes and perceptions about prostate cancer, or can be connected to other individuals who have faced similar decisions in the face of comparable risk. Then, they can be helped to understand key differences between these stories and their own circumstances. Finally, the potential value of obtaining second opinions can be made clearer.
In future work, we plan to incorporate these recommendations into treatment decision-aids. Because we found that patients tended to be either passive information recipients or active information seekers,48 we plan to tailor decision-aid content not only to a patient's clinical characteristics but to his preferences for handling and processing information. In comparison with usual care, we expect that such a decision-aid will lead to less precipitous and more personally satisfactory decisions, fewer treatment alternatives excluded on the basis of erroneous information, greater mutability in treatment preferences over time, and an enhanced ability to critically assess the relevance of other people's experiences with prostate cancer.
This was a cross-sectional study of a small number of patients in a single Veterans Affairs Medical Center where, compared with the general population, patients tend to have poorer health and fewer socioeconomic resources.49 In addition, patients in this study consulted with resident rather than attending urologists. Conversely, we included an age- and educationally diverse patient sample (including college graduates) and, for the first time, carried out comprehensive interviews during the actual decision-making period as well as 6 to 8 months following the treatment. Ultimately, we believe that the decision-making themes we have identified are likely to have universal applicability.