Introduction to clinical trials
Despite recent advances, the five-year survival rates for many cancers remain low, and there is a continued need for research to improve cancer outcomes. Clinical trials are research studies designed to improve cancer prevention, diagnosis, treatment, and survivorship. This research base necessarily involves enrolling cancer patients and others (e.g. family members for genetic linkage studies, healthy community volunteers to serve as matched controls) into clinical trials. Clinical trials demonstrate the efficacy of new therapies and are the mechanism through which research is translated into standards of care. The effectiveness of this translational process is greatly dependent on the number and representativeness of participants enrolled in trials, yet less than 5% of all adult cancer patients enter clinical trials. Despite a nearly 20-year effort by the National Institutes of Health (NIH) to enhance clinical trial accrual, these rates are not improving and even lower participation rates are reported in minority populations, including African Americans (AA).
The goal of this chapter is to outline issues involved in recruitment to clinical trials, to describe the ethical principles underlying informed consent and provide suggested strategies to aid communication between healthcare providers and patients about clinical trials.
Types of cancer clinical trials
According to the US National Cancer Institute, cancer clinical trials are generally categorized into one of the following phases of research: Phase I trials are initial studies with humans that usually enrol limited numbers of people. Their main purpose is to evaluate dosage safety and the frequency and method by which new drugs should be administered (e.g. either orally, or by injection into the bloodstream or muscle). Phase II trials are designed to further evaluate drug and dosage safety and to begin assessing the impact of drugs in treating specific types of cancer. Phase III trials test new drugs, new drug combinations, or new surgical procedures by comparing them against current standards of care. A participant will usually be randomly assigned to the standard (control) group or the new treatment group. Phase III trials often require, by design, large numbers of enrolees and data may be collected at multiple clinical sites across the United States and abroad.
Accrual to clinical trials
Patients are typically offered the opportunity to enrol in a clinical trial as a treatment option by their oncologists. The base rate of accrual at a cancer centre depends on the number of trials available to eligible patients at a given point in time. As noted above, low accrual rates have been reported in the literature and can be attributed to many factors, including the communication process that occurs when oncologists talk to patients (and families or companions, if present) about joining clinical trials (Fallowfield et al. 1997; Albrecht et al. 2003; Brown et al. 2007). Albrecht and colleagues suggest from their data at two cancer centres that low rates may be partially due to the extent to which physicians do and do not explicitly offer trials to their patients, and, in turn, are partially due to the extent to which patients understand that they have, or have not, been offered enrolment in a trial.
The challenge to physicians lies in the multiple and sometimes conflicting communication goals they face in communicating with patients and their families/companions in the outpatient clinic setting. Physicians must establish relational trust in the encounter, provide high quality care, ensure that patients are sufficiently informed to authentically provide ‘informed consent’ or ‘informed refusal’ in making treatment choices and when enrolling in clinical trials (accrual rates are actually performance measures for physicians at some institutions). Through this process, physicians are mandated to honour ethical and scientific principles of neutrality and full disclosure to protect their patients. This is a tall order; the full range of ethical concerns associated with clinical trials are described below.
Beneficence and the move from paternalism
These two concepts of beneficence and paternalism became linked in the Corpus Hippocraticum, with the doctors of ancient Greece undertaking to:
‘come for the benefit of the sick, remaining free of all intentional injustice and of all mischief’, while at the same time withholding the patient’s diagnosis and prognosis and diverting the patient’s attention away from the illness and treatment. (Hippocrates 1986) The notion of beneficent paternalism persisted in the medical tradition through to the 18th century when philosopher/physicians attempted to regulate professional conduct by formulating and publishing codes of conduct aimed at establishing medicine in an ethical framework (Lord 1995).
Beneficence remains a fundamental ethical principle guiding medical practice; however, the traditional paternalistic role of the doctor has become increasingly unacceptable in the light of changes in patient attitudes towards medical practice in the late twentieth century. Individuals no longer presume that the doctor knows best and many prefer an approach that involves greater patient involvement in decision-making and respect for individual autonomy. An examination of trends in physician behaviour over four decades suggests that physicians have shifted away from paternalistic styles, characterized by withholding information from patients about their prognosis, diagnosis, and treatment options in the belief that such information would be beyond patients’ comprehension and cause them excessive fear, anxiety, and loss of hope, thus worsening patient outcomes (Novack et al. 1979). Contemporary medical practitioners acknowledge the importance of providing accurate information to patients. Various reasons have been offered to account for this shift, including an increased fear of litigation among physicians, legal requirements, the publication of guidelines for the disclosure of diagnoses, and an improvement in therapies for cancer patients through technological advancement. Importantly, the law has been vital in promoting the concept of informed consent to standard and experimental treatments.
Active participation in the consultation requires negotiation between the physician and the patient, which is discouraged by the traditional paternalistic model. Patients are now seeking information to enable them to make decisions about treatment options, to understand prognostic issues, and to be clear about treatment side effects.
Autonomy in general refers to the individual’s right to self-determination. According to Faden and Beauchamp (Hippocrates 1986), an individual acts autonomously if three conditions are satisfied. That is, the individual acts (a) intentionally, (in accordance with a plan or one’s inner knowledge), (b) with understanding, and (c) free from controlling influences.
The extreme view of patient autonomy suggests that patients make their own decisions about treatment, while the doctor adopts a passive role. A more reasonable view argues in favour of the physician who inquires about the patient’s preferences and values, thus developing an understanding of the patient as a person before a treatment decision is reached, which maximizes the patient’s treatment goals.
In the experimental context, particularly the case of the randomized clinical trial, Kodish et al. argue that patient autonomy can only be ensured if the patient is ‘free to choose any therapy which they might have received by participating in the RCT and is equally free to choose the randomization alternative’ (Kodish et al. 1990). Moreover, Kodish et al. emphasize the right of a patient to make the choice to refuse any treatment, even when a treatment is proven, as an essential component of autonomous decision-making (Kodish et al. 1990).
Equipoise is defined as the point at which a rational and well-informed person has no preference between two (or more) available treatment options (Lilford and Jackson 1995). Thus, a physician who is convinced that one treatment option offers a better possibility of benefit for his/her patients than another, cannot ethically recommend random allocation as a means of making a treatment choice. The potential benefit to the patient must be the paramount consideration in the treatment decision. On the other hand, if the physician is uncertain about the difference in potential benefit between two (or more) treatments offered in a clinical trial, it is ethically acceptable to defer control of the treatment decision to the randomization process.
However, the practical application of equipoise as a means of justifying the selection of randomization as an ethical means of making treatment decisions remains controversial. Equipoise has also been named the ‘uncertainty principle’, reflecting the prominence of the physician’s inability to choose (based on lack of evidence) between comparative treatment benefits. Critics of equipoise question the degree of uncertainty physicians apply to the process of choosing between known and experimental treatments, and recent articles have added qualifiers such as ‘reasonably’, ‘substantially’, and ‘genuinely’ to uncertainty to try and further describe the physician’s belief about the treatment options. However, this raises the question of who decides what counts as reasonable or substantial uncertainty. This seems largely left to the conscience of the physician.
Clearly, from the physician’s perspective, reaching individual equipoise, weighing uncertainty, is a difficult process. However, another level of complexity is added by the introduction of the concept of collective equipoise. According to Chard et al. (Chard and Lilford 1998) collective equipoise relates to the uncertainty of a profession as a whole about a particular treatment modality. While individual equipoise may not be achieved (i.e. there is a preference for a particular treatment), this is balanced by others in the profession holding the opposing view (a preference for the alternate treatment). Thus, overall the profession is in collective or clinical equipoise. Clinical equipoise recognizes that it is the community of physicians that establishes best practice standards, and not the individual physician.
Supporters of the primacy of collective equipoise have suggested that in a case where clinical equipoise exists, a randomized controlled trial (RCT) is an ethical imperative to avoid retaining ineffective modes of treatment. In this situation, collective equipoise should override the physician’s individual equipoise; thus even if the physician has a preference for a particular treatment (of those being compared), s/he would be expected to recruit patients to the trial (Freedman 1987). Conversely, while clinical equipoise appears to offer a neat solution for the physician committed to research but conflicted by degrees of clinical uncertainty, there are a number of compelling arguments suggesting that equipoise is inherently unethical as a justification for randomized trials. Enkin (Enkin 2000) and others point out that if moral authority is granted to the medical community as a whole, the individual responsibility of physicians is devalued and the needs of the patient for guidance are overlooked (Hellman 1979). While the medical community may be certain about the effectiveness of a treatment at one time, this certainty can change, and the preferences of individuals do count. In addition, clinical equipoise may pose a threat to the transparency of the doctor–patient relationship. If the physician is participating in a trial justified by collective equipoise and does not disclose a particular treatment preference, then a basic ethical tenet has been violated. Again, this poses a dilemma for the physician who must balance his/her clinical opinion, enthusiasm for research, the weight of clinical uncertainty, and the best interests of the patient in order to make an ethical treatment recommendation.
The ethical principle of justice refers, in the current context, to the application of rules of fairness and equality to the clinical trial process. This can be realized as: (a) fairness in the distribution of the harm and benefit of trial treatments; and (b) equitable criteria for the inclusion of potential trial participants.
In the first instance, it is argued that injustices occur when individuals are advantaged through medical research at the expense of others. Marquis (Marquis 1983) clarifies this point in noting that few would condone a society in which people are sacrificed for their functioning organs in order to benefit the needs of the society in general. However, ethical conflicts can arise and ethical principles surrounding individual versus social benefit need to be recognized and balanced. Trial patients (particularly those participating in Phase I studies) are commonly treated with promising new treatments that are not guaranteed to provide any personal benefit but which may benefit others in the future. The crucial difference between Marquis’ example and the plight of trial patients is that trial patients are routinely informed of the uncertainty of treatment benefit prior to trial entry. Thus, gaining informed consent guards against such an ethical problem. However, as the quality of information provision about trials is variable and as patients can misunderstand this information, it is possible to question the validity of the safeguard of informed consent.
Low representation of minority patients in clinical trials results in inequity in access to the latest technologies and cancer treatments, compromises the generalizability and external validity of trial results, and may fail to identify important positive or negative treatment effects among underrepresented populations. Inequitable access to state-of-the-art cancer care contributes to health disparities in cancer mortality and survival.
Informed consent can be defined as an autonomous action taken by a patient giving permission for doctor to undertake a medical plan. Informed consent became part of US law in 1914. However, while this legal instruction instituted patient authorization as part of the treatment process, it did not define the nature of the information that should be provided to the patient about their illness or possible treatments.
Gert et al. (Gert et al. 1997) provide a considered view of the bioethics of the consent process (both for clinical trials and standard treatments). They differentiate between the moral rules governing this situation that are more or less compulsory and governable by law (including provision of adequate information, lack of coercion, and assessment of patients’ competence to make a choice) and moral ideals, to which doctors aspire but cannot necessarily fulfil in all instances; for example, providing information about alternative treatments in a way which does not overemphasize the attractiveness of one, or belittle another. The consent process involves the doctor presenting information to the patient about their illness and the options for treatment, and making an appraisal about the patient’s response to the information, including the degree to which the patient understands the information provided (Gert et al. 1997). The rationale underlying the doctrine of informed consent is to protect patient autonomy and to ensure that patients have an active role in making barriers to recruitment.
Gaining informed consent to clinical trials is problematic for doctors. Many doctors experience problems initiating clinical trial discussions and find the dual roles of caring physician and experimenter difficult to resolve (Fallowfield 1995). Prospective studies have reported that 70–80% of non-accrual is attributable to the doctor. Doctors’ reasons for not accruing patients to trials include concerns over (a) damaging the doctor—patient relationship; (b) acknowledging the uncertainty of treatment benefits; and (c) practical issues such as rigid protocol designs, patient inconvenience, and extra work for physician. These results suggest that efforts to improve doctors’ participation in clinical trials need to address communication difficulties experienced by doctors when recruiting patients to trials. Communication difficulties with patients are evident in three key areas: (i) oncologists omit critical information or the information is poorly presented leading to patient misunderstanding and poor recall of information; (ii) oncologists underestimate their patients’ information needs and overestimate the amount of information they give; and (iii) in spite of evidence-based calls to routinely involve patients in decision-making, patients are often not involved and their decision-making preferences are not being met. These difficulties are compounded with minority patients, as physicians use less supportive and positive talk, and for example, are less patient-focused with African American patients than white patients. Moreover, AA patients have been shown to be less active communicators than white patients (Street et al. 2005). Such racial disparities in physician–patient communication could lead to less exchange of information and less patient involvement, in turn leading to less informed decisions (Gordon et al. 2006) and lower trust in physicians.
Many eligible patients who are invited to participate in a trial, decline (though estimates widely vary from 23–50%). Reasons for trial refusal by eligible patients include concerns regarding experimentation, and uncertainty and loss of control over treatment decisions. Many patients and the general community do not understand the role of randomization in avoiding bias in treatment selection (Ellis et al. 1999). Other barriers have been identified such as race, gender, and lack of knowledge of the requirements of trial participation, and the possibility of receiving a placebo. Studies suggest that racial differences in patient barriers to clinical trial participation are due in part to non-clinical factors related to: (a) a paucity of culturally relevant information that is evident in minority patients’ lack of trial knowledge and understanding of important trial procedures such as dose escalation and randomization; (b) higher mistrust of the research enterprise and the medical system among AA patients, which is reflected in concerns regarding experimentation and loss of control over treatment decisions; (c) factors related to the patient–physician relationship; and (d) family pressures. Patients who actively participate in their healthcare by asking questions and involving themselves in making treatment decisions have improved outcomes such as lower anxiety and increased perceived control over their disease as compared to patients who are passive. Cancer patients vary in their ability to be active communicators. Previous studies have found that physicians are more informative, accommodative, and supportive with more actively communicating patients and that, with appropriate support, the participatory level of patients can be increased.
Patients’ decisions about enrolling in trials are affected by what physicians tell them about the clinical trial and how physicians tell them the information. Content messages are what physicians tell patients about the trial. These include legally proscribed aspects of the study protocol (essentially the information on the consent document), and the potential adverse effects (side effects) that the patient is likely to experience from the drug therapies. Albrecht and colleagues have added three additional types of content messages that they have found important for patient understanding of clinical trials. These include messages of reassurance and support, specifically regarding the patient’s experience of each potential side effect, reassurance and support regarding the patient’s decision to enrol in the clinical trial (whether s/he decides for or against enrolment) and discussion regarding the benefits and drawbacks of clinical trial participation.
Finally, perhaps the most important content message from the physician is to recommend to a patient that s/he enrol in a trial. Such recommendations do influence patients’ decisions (Eggly et al. 2008). In observing clinical offers, Eggly and colleagues have shown that most physicians do recommend the trials that they are offering to their patients. Indeed, in contrast to the equipoise principle, many do so in a more directive, not general manner, such as saying, ‘I recommend this trial for you’, as opposed to saying ‘I recommend this trial’.
The patient’s perspective: Improving the decision process
From the patient’s perspective, the decision of whether to enrol in a clinical trial is complicated by the reasons used to arrive at the conclusion and the affective and cognitive aspects of the decision as it is made and afterwards. Reasons for the decision made vary widely and include personal factors (perceived quality of life, length of survival), family members, and significant others’ opinions, perceptions of the potential side effects, and perceptions of the financial costs involved in enrolling in the trial. Physician communication behaviours also factor in to patients’ judgements, especially how well they seemed to listen and answer questions, how the patient perceived the way the physician interacted with his/her family or companions, and how well explanations were given and the level of empathic support provided.
Cognitive and affective aspects of the decision involve the degree to which the patient is confident in the decision s/he is making, the extent of agreement shared with the physician and family/companions regarding the nature of the decision, and the level of positive relational effect perceived with the physician, and the family/companion as they face the decision and the treatment process together (Albrecht et al. 2008).
Communication about clinical trials
The development of communication skills training has been suggested as a promising way forward to aid clinicians in the difficult task of clinical trial recruitment. Brown et al. in a series of articles have developed and pilot tested an informed consent communication skills workshop. The results of this programme of research revealed four areas where communication training could aid physician–patient communication (Brown et al. 2004b, c; Brown et al. 2007). These included: (a) shared decision-making strategies; (b) the sequence of moves in the consultation; (c) the type and clarity of the information provided; and (d) disclosure of controversial information and coercion. These themes reflect the clinical judgement and theoretical perspectives of linguists, psycho-oncologists, ethicists, and oncologists involved in the analysis.
Shared decision-making strategies
Participation in treatment decision-making, at the patient’s preferred level of involvement, was identified as an essential component of seeking informed consent to the clinical trial. Fourteen strategies contributing to a collaborative decision-making framework were identified. They are summarized in Box 32.1.
Reprinted from Social Science and Medicine, Volume 58, Issue 2, R.F Brownet al, ‘Developing ethical strategies to assist oncologists in seeking informed consent to cancer clinical trials’, pp. 379–390, Copyright © 2004 Elsevier Science Ltd, permission from Elsevier, http://www.sciencedirect.com/science/journal/02779536
Importantly, language that portrays the patient as an active agent in the process of deciding about and enacting their own healthcare encourages the sense of an autonomous self among patients. Grades of agency occur; the most active participant is portrayed as the doer, decider. The least active participant is portrayed as the person or object ‘done to’ (the one who is treated, told, or organized).
Sequence of moves in the consultation
The analysis led to an understanding of the importance of sequence in the interaction. Thus, the consultation data were categorized into a series of phases and an ideal sequence of these phases was identified. This model was developed to promote patient understanding of information, to ensure equal weight was given to the discussion of standard and experimental treatments, and to avoid potential coercion (see Fig. 32.1).
Within Pathways 1 and 2, a number of facts need to be communicated in order for the patients to give ethical informed consent. However, merely including these facts will not necessarily ensure understanding. The fullness and clarity of the explanation needs also to be considered.
Disclosure and coercion
Issues that could be covered include: (a) that in many instances the participating doctors may be investigators on the trial and thus have a potential conflict or duality of interest; (b) the accessibility of trial treatments after the trial has ceased; (c) the availability of other potentially suitable trials.
Words used by doctors, who may be quite unaware of their ramifications, may encourage or perhaps coerce patients into participating in clinical trials. These are outlined next.
If the doctor does not explicitly state their views on clinical trial participation, while acknowledging patient choice, patients may feel unspoken pressure to participate. Preferences may be covertly suggested in many ways; for example, by spending more time talking about the trial treatment versus standard treatment.
Doctors commonly use the term ‘you are eligible for this trial’. This phrase, however, can imply that the patient is ‘lucky’ to have been selected, or should be hopeful that their disease status allows them to participate in the trial. We suggest using the phrase ‘the trial is suitable for you’.
Appealing to altruism
A common motivation for patients to enter clinical trials is a sense of making a contribution to medical knowledge which will benefit others, or altruism. Finding the balance between recognizing and appreciating patient altruism and using it in a coercive fashion can be difficult. Once again, the terms used can make a difference. Thus the use of terms such as ‘You can benefit future generations’ is perhaps more coercive than ‘This will help us find the answer to this question’.
Coercion may also occur when the potential value of a clinical trial treatment is presented with a positive frame versus negative framing for the standard treatment (or vice versa). Research suggests that some patients prefer positively framed information (‘you have a 70% chance of cure’) as this encourages a positive outlook, while others prefer negatively framed information (‘you have a 30% chance of the cancer coming back’) as this emphasizes the importance of additional treatment.
Evaluation data and modification to programme for MSKCC
A modified version of this physician-focused communication skills training programme, based on the COMSKIL model (Brown and Bylund 2008) was implemented and evaluated as part of a larger programme of communication skills training at Memorial Sloan Kettering Cancer Center. At the completion of the communication skills training module, participants were asked to complete an evaluation. They were asked to rate aspects of their own sense of confidence and self-efficacy in dealing with the communication challenge targeted by the module. Participants were presented with a list of question and asked to indicated their response on a Likert scale from 1–5 with anchors at 1—‘Strongly disagree’ to 5—‘Strongly agree’.
We compared scores on the self-evaluation data pre- and post-training using paired sample t tests. Participants’ confidence in making shared treatment decisions, including discussing of a clinical trial, increased significantly pre- and post-training. Three other post-training questions about self-efficacy and confidence all received average scores above four, indicating that the participants strongly agreed that they would use newly acquired skills, provide better care after training, and had been prompted to critically evaluate their own communication skills (Table 32.1).
Table 32.1 Course evaluation data: Mean rating for shared treatment decision-making module
Course evaluation items
N = 101
Before this workshop I felt confident making shared treatment decisions
Now that I have attended the workshop I feel confident to make shared treatment decisions
4.04 (.66) (p < .05)
I feel confident that I will use all the skills that I learned today
The skills I learned today will allow me to provided better patient care
The workshop prompted me to critically evaluate my own communication skills
Adapted from Brown RFet al., Developing patient—centered communication skills training for oncologists: Describing the content and efficacy of training, Communication Education Special Edition, Volume 59, Issue 3, pp. 236–249, Copyright © 2010, with permission of the authors.
Seeking informed consent to cancer clinical trials presents a significant communication challenge for oncologists and patients. Improving this communication may lead to increased accrual to clinical trials. Strategies aimed to aid this communication focus on four areas designed to ensure a transparent dialogue, underpinned by ethical informed consent and free from coercion.
The authors continue to pursue research agendas that explore gaps in communication about clinical trials and factors that affect decision-making about clinical trials. In addition, the authors are evaluating the utility of interventions that aid both physicians and patients in communicating about trials.
Albrecht TL, Eggly SS, Gleason ME, et al. (2008). Influence of clinical communication on patients’ decision making on participation in clinical trials. J Clin Oncol26, 2666–73.Find this resource:
Albrecht TL, Penner LA, Ruckdeschel JC, et al. (2003). Understanding patient decisions about clinical trials and the associated communication processes; a preliminary report. J Cancer Educ18, 210–14.Find this resource:
Brown RF, Butow PN, Boyle F, Tattersall MH (2007). Seeking informed consent to cancer clinical trials: evaluating the efficacy of communication skills training. Psychooncology16, 507–16.Find this resource:
Brown RF, Butow PN, Butt DG, Moore AR, Tattersall MH (2004a). Developing ethical strategies to assist oncologist in seeking informed consent to cancer clinical trials. Soc Sci Med58, 379–90.Find this resource:
Brown RF, Butow PN, Butt DG, Moore AR, Tattersall MH (2004b). Developing ethical strategies to assist oncologists in seeking informed consent to cancer clinical trials. Soc Sci Med58, 379–90.Find this resource:
Brown RF, Butow PN, Ellis P, Boyle F, Tattersall MH (2004c). Seeking informed consent to cancer clinical trials: describing current practice. Soc Sci Med58, 2445–57.Find this resource:
Brown RF, Bylund CL (2008). Communication skills training: describing a new conceptual model. Acad Med83, 37–44.Find this resource:
Chard JA, Lilford RJ (1998). The use of equipoise in clinical trials. Soc Sci Med47, 891–8.Find this resource:
Eggly S, Albrecht TL, Harper FW, Foster T, Franks MM, Ruckdeschel JC (2008). Oncologists’ recommendations of clinical trial participation to patients. Patient Educ Couns70, 143–8.Find this resource:
Ellis P, Dowsett SM, Butow PN, Tattersall MH (1999). Attitudes to randomised clinical trials among outpatients attending a medical oncology clinic. Health Expect2, 33–43.Find this resource:
Enkin MW (2000). Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomised controlled trial. BMJ321, 756–8.Find this resource:
Fallowfield L (1995). Can we improve the professional personal fulfilment of doctors in cancer medicine. Br J Cancer71, 1132–3.Find this resource:
Fallowfield L, Ratcliffe D, Souhami R (1997). Clinicians’ attitudes to clinical trials of cancer therapy. Eur J Cancer33, 2221–9.Find this resource:
Freedman B (1987). Equipoise and the ethics of clinical research. N Eng J Med317, 141–5.Find this resource:
Gert B, Culver CM, Clouser KD (1997). Bioethics: A Return to Fundamentals. Oxford University Press, New York, NY.Find this resource:
Gordon HS, Street RL Jr, Sharf BF, Souchek J (2006). Racial differences in doctors’ information-giving and patients’ participation. Cancer107, 1313–20.Find this resource:
Hellman S (1979). Editorial: randomised clinical trials and the doctor—patient relationship. An ethical dilemma. Cancer Clin Trials2, 189–93.Find this resource:
Hippocrates (1986). A History and Theory of Informed Consent, Oxford University Press, New York, NY.Find this resource:
Kodish M, Lantos JD, Siegler M (1990). Ethics of randomised clinical trials. Cancer65, 2400–5.Find this resource:
Lilford RJ, Jackson J (1995). Equipoise and the ethics of randomisation. J R Soc Med88, 552–9.Find this resource:
Lord RS (1995). Informed consent in Australia. Aust N Z J Surg65, 224–8.Find this resource:
Marquis D (1983). Leaving therapy to chance. Hastings Cent Rep13, 40–7.Find this resource:
Novack DH, Plumer R, Smith RL, Ochitill H, Morrow GR, Bennett JM (1979). Changes in physician attitudes towards telling the cancer patient. J Am Med Assoc241, 897–900.Find this resource:
Street RL, Gordon HS, Ward MM, Krupat E, Kravitz RL (2005). Patient participation in medical consultations: why some patients are more involved than others. Med Care43, 960–9.Find this resource: