- Email: [email protected]
Information priorities of Italian early-stage prostate cancer patients and of their health-care professionals
Patient Education and Counseling 56 (2005) 174–181
Information priorities of Italian early-stage prostate cancer patients and of their health-care professionals Carlo Capirci a,∗ , Deb Feldman-Stewart b , Giovanni Mandoliti a , Michael Brundage b , Giorgia Belluco c , Katia Magnani c b
a Department of Radiation Oncology, International Cancer Center, Rovigo 45100, Italy Division of Cancer Care and Epidemiology, Cancer Research Institute, Queen’s University, Kingston, Ont., Canada c International Cancer Center, Psycho-Oncological Service, Rovigo, Italy
Received 18 January 2003; received in revised form 10 November 2003; accepted 2 February 2004
Abstract The study was designed to compare the information priorities of Italian early-stage prostate cancer patients to those of their health-care professionals. Doctors (urologists and oncologists), nurses, radiation therapy technologist (RTs), and patients treated with radiotherapy in Northern Italy were surveyed. Respondents rated the importance of addressing each of 78 questions as: essential, important, or superfluous. We compared priorities between pairs of groups by correlating the percentage of each group that rated each question (a) essential and (b) superfluous. We assessed within-group agreement by identifying questions that >50% of one response. Fifty-nine doctors, 53 nurses, and 45 RTs (overall 70% response rate) and 50 patients (91% response rate) participated. All correlations between professional groups were high (all >0.74). All professional group-patient correlations were significant but much lower than those among the professionals. More importantly, there was considerable variation within each group: within each group, almost all questions were rated essential by some members but less than seven of the 78 questions were rated so by half (or more) of its members. In addition, almost all questions were also rated superfluous by some members of each profession. We concluded that the professional groups appear to agree with one another on questions essential and those superfluous to address more than they agree with a group of their patients. More importantly, there is considerable disagreement within each group. The results imply that the professionals cannot assume that their own information priorities are the same as those of their patients. © 2004 Published by Elsevier Ireland Ltd. Keywords: Prostate cancer patients; Informed consent; Professionals’ opinions
1. Introduction The Hippocratic Oath makes it clear that clinicians have always been interested in promoting patient well-being, and historically, withholding information from patients was considered by some to be beneficent [1]. More recently providing information has been viewed as beneficent and assessing the adequacy of provision of information in medical situations has been driven to a large extent by the desire to ensure that “informed consent” has been obtained. The term “informed consent” was introduced in 1957 by Paul Gebhart, a dean of medical association lawyers. In the same year, the California court adopted the expression “informed consent” in a malpractice case [2]. This was the first time a court ∗ Corresponding author. Tel.: +39-0425-393292; fax: +39-0425-393373. E-mail address: rt [email protected] (C. Capirci).
0738-3991/$ – see front matter © 2004 Published by Elsevier Ireland Ltd. doi:10.1016/j.pec.2004.02.006
had adopted “informed consent” to represent an obligation of physicians and other health-care professionals to provide patients with sufficient information about recommended diagnostic and treatment procedures, including potential risks and complications. While the legal doctrine developed for informed consent, so did its ethical framework. Three principles underlie informed consent: support for patients’ autonomy, beneficence, and justice [3]. The relative weight of the principles could differ for different cultures, as could the views of how the principles are best implemented. Such differences may be, for example, what underlies the difference between “non-disclosure dominant” cultures like Southern Europe and Japan, and “disclosure dominant” cultures like Canada, United States, and Northern Europe. The justice principle in the context of information provision means that medically similar patients should be provided with the same opportunity to receive particular
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
pieces of information, including from different physicians. Such an information core could be used as a legal professional standard for disclosure in informed consent. In the decision-making process, however, several elements that could affect the decision require an assessment by the doctor which is at least partly subjective. Such assessments include those related to quality of life, to disease aggressiveness and possible outcomes of treatment, to psychological acceptance of disease, and treatment toxicity. Although there may be subjective elements in assessments of the patient, personal values held by the physician could violate the ethical principle of justice, and also that of autonomy: patients have the right to act with intention, with understanding and without controlling influences [3]. Concern that applying the legal professional standard does not respect patient autonomy has led to its replacement by the “objective” standard: it requires physicians to provide the information that a “reasonable person” would consider material to the decision under the specified circumstances [2]. In practice the “reasonable person” has been the jury, which in turn led to concern that the jury did not represent patients’ needs well. The concern led to substituting the “objective” standard with the “subjective” standard, requiring that physicians provide the information material to the particular patient’s decision. Thus, the patient defines the information to be provided. Although physicians in most jurisdictions are legally responsible to ensure that patients are providing informed consent, they frequently are not the only health-care professionals to provide patients with information. In some places, for example, nurses are charged with the responsibility of educating the patient. In the delivery of radiotherapy, the radiation therapy technologists (RTs) see the patient on a daily basis, providing information about procedural issues in addition. Thus, to determine information priorities of those around the patient, it is important to include all health-care professionals that the patient encounters. The aim of this study was to determine the extent of agreement among Italian health-care professionals, among patients, and between the professionals and patients on what questions should be addressed with curable prostate cancer patients before treatment decisions are made. We focussed on only curable prostate cancer because the distinction between having curable versus non-curable disease by its very nature is likely to result in different issues being considered important. Although we asked participants to focus on the time before treatment decisions were made, we recognize that providing information to patients at that time can be for many reasons of which decision-making is only one [4]. The study used the methodology developed by FeldmanStewart et al. [5] to address these issues in a Canadian setting. The method was used because it included a list of relevant issues that were detailed in nature; most other studies of information priorities in prostate cancer have focussed on larger categories of information. Although Canada has been identified as a disclosure dominant culture and Italy as
175
non-disclosure dominant, we felt it appropriate to use the question list developed in Canada to study Italian priorities: if Italy really is non-disclosure dominant, then participants would identify few questions as important to address, if it is not, then the comprehensive Canadian list would provide a broad basis from which to identifying priorities.
2. Methods 2.1. Design Four groups participated in the study: three groups of professionals, doctors , nurses, and radiation therapy technologists, and patients. All professionals were involved in treating patients with prostate cancer at the time of the study and were recruited from eight clinics in Triveneto of Northern Italy. The doctors included 10 urologists, 11 medical oncologists, and 38 radiation oncologists. Because the numbers of urologists and of medical oncologists were so few, their responses were added the radiation oncologists to create the doctor group. Patients were also recruited from the Triveneto region and all had received radiotherapy, some received their radiotherapy after surgery, and some received hormones after their radiotherapy. The data were gathered through a survey. For the professionals, the survey included a scenario describing a man presenting with curable prostate cancer. The man’s age and state of health made him eligible for any of three potential medical treatments: radiation therapy, surgery or watchful waiting [6]. The patients were already treated patients who were asked to consider their judgments as a recommendation for questions that should be addressed with a newly diagnosed patient in a situation like their own. We felt that newly diagnosed patients could benefit from the insights of the patients have been through the treatment process. The respondent was asked to judge the importance of addressing each of 78 questions before the patient’s treatment decision was made. Response options were “essenziale” (essential), “importante” (important), or “superflua” (superfluous). The instructions asked the professional respondents to focus on which questions should be addressed regardless of patient queries. After judging the importance of addressing the various questions, the respondent filled out a form providing personal information that included sex (for the professionals) and age. 2.2. Statistical analyses To compare relative information priorities between pairs of groups, we used Pearson product moment correlation coefficients to assess the associations between the percentages of each group that rated each question using a particular response. To assess the extent of agreement between pairs of groups (e.g. do doctors agree with nurses more than they agree with patients), we used Hartley’s Fmax [7] to compare
176
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
the proportion of accounted-for variance in one situation to that of the second. We compared the absolute amount of information deemed essential between pairs of groups using t-tests to compare the mean number of questions each rated essential. We also compared agreement within the groups. Within each group, we identified the questions that at least 50% of respondents rated essential as questions they “agreed” were essential to address. Correlations were used to determine if there were associations between the number of essential responses and demographic/professional characteristics. For the continuous variable age, the correlations were Pearson correlation coefficients. Correlations with the dichotomous variable sex were point biserial correlations. In evaluating the predictive ability of the correlations, we squared the correlation to obtain the amount of accounted-for variance.
3. Results 3.1. Respondents A description of respondents in each of the groups is shown in Table 1. As the table shows, the most notable difference among the groups is their age, with the average age of the patients almost twice that of the professionals. 3.2. Between-group comparisons 3.2.1. Comparisons on content Table 2 shows the Pearson correlation coefficients comparing the percentage of particular responses on each quesTable 1 Description of respondents in each group
Doctors Nurses RTs Patients
N
Male:female
Mean age
Response rate
59 53 45 50
40:19 12:41 22:23 50:0
43 38 37 70
69% 69% 78% 92%
(59/85) (53/77) (45/58) (56/61)
Table 2 Pearson correlation coefficients comparing overall priorities between pairs of groups Nurses
RTs
Patients
Essential questions Doctors Nurses RTs
0.82 – –
0.86 0.75 –
0.63 0.64 0.68
Superfluous questions Doctors Nurses RTs
0.84 – –
0.84 0.87 –
0.48 0.47 0.40
Table 3 For each professional group, Fmax comparing their agreement with the other professional groups to their agreement with the patients Professional agreement-essential questions Doctor–Nurse
Doctor–RT
Patient agreement with professionals Patient–Doctor 1.69∗ 1.86∗ Patient–Nurse 1.34 Patient–RT 1.60∗
Nurse–RT
1.37 1.22
Professional agreement-superfluous questions Doctor–Nurse
Doctor–RT
Patient agreement with professionals Patient–Doctor 3.06∗ 3.06∗ Patient–Nurse 3.19∗ Patient–RT 4.41∗ ∗
Nurse–RT
3.43∗ 4.73∗
P < 0.05.
tion, between pairs of groups; the top half of the table shows the comparisons of essential responses and the bottom half the comparisons of superfluous responses. As can be seen in the table, all of the correlations comparing pairs of professional groups are relatively high and account for at least 50% of the variance. Thus, the questions that most people within one profession rated essential tended to be similar to the questions that most people in each of the other professions rated essential; and similarly for superfluous questions. Table 2 also shows that the correlation coefficients drop across the board when comparing the patients to each professional group as compared to the correlations between pairs of professional groups, both when considered essential questions and when considering superfluous questions. Table 3 shows a comparison of the extent of agreement between pairs of professional groups to their agreement with the patients. As the table shows when considering essential questions, the low doctor–patient agreement on essential questions means that doctors agree significantly more with each of the professional groups than they do with the patients. The table also shows that, although patients’ essential priorities were most similar to RTs’ priorities, there was still less agreement between the RTs and patients than between the RTs and doctors. As for superfluous questions: the table shows that each professional group agrees with the other professional groups significantly more than they agree with the patients. The questions were then sorted into one of four categories according to their focus; 36 focussed on the disease, 20 on treatment details, 10 on side effects, and 12 questions on treatment alternatives. Fig. 1 shows the correlation coefficients comparing relative essential priorities of each professional group with those of the patients in each of the categories and overall. Although the low numbers of questions in each category means the confidence intervals around the correlations are large, the figure suggests that patient-professional agreement is highest overall with RTs and that highest agreement is consistent over three categories
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
177
Fig. 1. Peason correlation coefficients comparing each professional group to patients. The correlations compare the percentage of essential responses to each question between the patients and the particular professional group labeled. The bars on the left show the overall comparison that included all questions. The disease category included 36 questions, the treatment details category 20 questions, the side effects category 10 questions and the alternatives category 12 questions.
of the four categories; patient priorities around side effects appear most similar to the priorities of the nurses.
identified significantly fewer than the doctors and the nurses [F(1, 195) = 2.0, P < 0.05].
3.2.2. Comparisons on amount The average number of questions rated essential by individual participants was similar across groups: 17, 18, 17, and 15, for doctors, nurses, RTs, and patients, respectively (all t < 1). The average number of questions rated superfluous by individual participants was 20 by doctors and by nurses, 15 by RTs, and 25 by patients, which differed significantly from each other [F(3, 195) = 4.3, P < 0.01]; orthogonal contrasts showed that the patients identified significantly more superfluous questions than all of the professionals [F(1, 195) = 3.07, P < 0.01] and that the RTs
3.3. Within-group agreement 3.3.1. Agreement on content Agreement on content was investigated from two perspectives: on essential questions as content to include and on superfluous questions as content to exclude. Using a simple majority to define “agreement” within groups, each group agreed on few questions as essential: six for the doctors, two for the nurses, five for the RTs, and two for the patients. Table 4 lists the questions in their order of percentage essential responses (highest to lowest) within each
Table 4 For each group, questionsa rated essential by at least 50% of members Doctors
Nurses
RTs
Patients
3. Will the treatment cure my disease?
3. Will the treatment cure my disease? 2. Will the treatment make my disease go into remission?
3. Will the treatment cure my disease? 2. Will the treatment cause my disease to go into remission? 4. Do I have the right to refuse the treatment? 11. If my prostate cancer is not treated, will I die from it? 22. If my disease is not cured, how long will I live?
3. Will the treatment cure my disease? 11. If my disease is not treated, will I die from it?
2. Will the treatment cause my disease to go into remission? 4. Do I have the right to refuse the treatment? 10. Are there other treatment possibilities not mentioned? 11. If my prostate cancer is not treated, will I die from it? 23. Could the treatment lead to my death? a
Questions ordered by their percentage of essential responses (highest to lowest).
178
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
Table 5 For each group, questionsa rated superfluous by at least 50% of members Doctors
Nurses
RTs
Patients
75. Did I get my disease from what I eat?
75. Did I get my disease from what I eat?
72. What is the cost of the treatment? 57. How common is the disease I have?
72. What is the cost of the treatment? 74. How many prostate cancer patients has my physician treated? 63. Can I spread prostate cancer to other people? 69. If my disease is not treated, will it make me sick to my stomach? 73. Did I get my disease from my work?
76. How many prostate cancer patients have been cured by my physician? 72. What is the cost of the treatment? 75. Did I get my disease from what I eat?
46. Why is it sometimes said that patients don’t live any longer with early diagnosis? 72. What is the cost of the treatment? 60. Will the treatment change the way I look? If so, for how long?
58. Am I different from the usual patient with prostate cancer? 76. How many prostate cancer patients have been cured by my physician? 69. If my disease is not treated, will it make me sick to my stomach? 70. How long have I had prostate cancer? 73. Did I get my disease from my work? 61. If my disease is not treated, will it make me tired? 66. If my disease is not treated, will it change the way I look? 46. Why is it sometimes said that patients don’t live any longer with early diagnosis? 59. How many patients like me have decided to have this treatment? 71. Can the treatment stimulate my disease to grow or spread? a
73. Did I get my disease from my work? 77. What special qualifications does my doctor have to treat my disease? 56. Will the treatment cause my hair to fall out? If so, for how long?
76. How many prostate cancer patients have been cured by my physician?
Questions ordered by their percentage of superfluous responses (highest to lowest).
group. As the table shows, although there are very few questions that members agreed were essential, three of the questions were identified by three of the four groups. We note that we compared responses of the three doctor groups and, acknowledging that the medical oncologist and urologist numbers are too small to conclude with certainty, but the results of three groups appear to show very similar patterns. While, within each group, there were few questions that a simple majority of members rated essential, almost every question was rated so by some its members. Every question was rated essential by some doctors and by some nurses. Only one question was not rated essential by any RTs, and two not rated so by any patients. On average each question was rated essential by 21, 23, 22, and 19% by the doctors, nurses, RTs, and patients, respectively. Interestingly, while there were a few questions not considered superfluous by each of the professional groups (one, one, and three by doctors, nurses, and RTs, respectively), every question was considered so by some patients. In addition, on average, more patients rated each question as superfluous than did the professional groups: the average per
question was 21, 26, 19, and 32% by the doctors, nurses, RTs, and patients, respectively. In spite of the higher average patient percentage, the doctors had the most questions rated as superfluous by at least 50% of its members; there were 13, 7, 3, and 6 questions that reached that threshold by the doctors, nurses, RTs, and patients, respectively. The particular questions rated superfluous by at least 50% of its members are listed in Table 5. 3.3.2. Agreement on amount The extent of disagreement on the number of questions that individual members within each profession rated essential can be seen in Fig. 2. It shows the number of questions deemed essential by individual participants within each group, each panel showing one group. As can be seen in each panel, there was wide variation in the numbers of questions deemed essential by members within each group: while 14% of doctors thought ≤5 questions were essential to address, another 27% thought ≥20 were essential. Those two extreme proportions for each of the other groups were 21 and 34% for the nurses, 9 and 27% for RTs, and 34 and 16% for the patients.
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
179
Fig. 2. Number of essential questions per participant in each group.
3.4. Associations with demographic characteristics We expected that the number of questions considered essential would lower with age and with male gender. There was, however, no association between age and the number of questions rated essential among those who reported their age of the doctors, nurses, or patients (r = −0.01, 0.15, −0.03, respectively, all P > 0.2); among the RTs, there was a trend for fewer questions to be rated essential as RTs got older (r = −0.26, P = 0.1). There was no association between sex and number of essential questions among doctors or nurses (r = −0.06, 0.00, respectively, both P > 0.7). Among RTs, however, males identified fewer essential questions than females (r = −0.43, P < 0.01). When controlling for sex, among RTs, there was absolutely no association between age and the number of essential questions (r = 0.02, P = 0.74); when controlling for age, there was still a trend for males to identify fewer essential questions than the females (r = −0.14, P = 0.1).
4. Discussion and conclusions The results of this study suggest that the information priorities of health-care professionals treating early-stage prostate cancer patients who receive radiotherapy in Italy are more similar to one another than they are to the priorities of
their patients. We observed, for example, significantly more agreement on what questions are essential between the doctors and both professional groups than between the doctors and the patients. When considering the relative amount of agreement between the patients and each professional group in each sub-category of questions, it appears that questions about side effects stand out as anomalous: the priorities seem most similar between the patients and RTs in all subcategories except side effects where patient priorities appear to agree most with the nurse priorities. It is interesting to wonder if the apparent higher level of agreement between the RTs and patients is related to the fact that of the three professional groups, the RTs spend most time with the patients. The higher agreement about side-effect priorities between the nurses and patients may reflect the fact that side effects are most often treated by the nurses. These results raise the possibility that different professional groups might specialize in the categories of information that they provide to patients. Higher agreement among the professional groups than between the professionals and the patients was also observed in the questions considered superfluous. The patients consistently identified more questions as being superfluous than did any of the professional groups. The results suggest that in determining information to be shared with patients, one needs to pay attention not only to information that should
180
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
be shared but also to information that need not be shared. It is difficult to know if the superfluous information is, in fact, information that would offend patients if provided or that it is just not of interest; however, providing the unwanted information risks at the minimum, overwhelming patients which, in turn, can cause anxiety and frustration [8]. More important than the extent of agreement at the group level, however, is the fact that there is a large amount of variability within each of the groups surveyed. Within each group, there were very few questions that even just a simple majority of members thought were essential to address. However, everyone surveyed indicated that some questions were essential. And almost every question was considered essential by some members of every group. Similarly, almost every question was considered superfluous by some members of every group. The implication of these observations are discussed below. The extent of variation within the doctors suggests that developing a professional standard for informed consent would require much effort in order to arrive at a consensus. The high variability in priorities among the doctors also raises concern about violations to the justice principle. To the extent that information provided to patients is a function of the individual doctor’s information priorities, these results suggest that all patients would not be provided with access to the same information. In addition, high variability within all three professional groups means that an individual patient who sees multiple professionals, from either the same or different professions, is likely to face different information priorities which, in turn, may contribute to patient confusion that has been reported [9]. Although no doubt beneficial for patients if they didn’t have to face confusion generated by disagreement amongst the professionals, using a professional standard for disclosure is questionable: providing patients with the information that they want has been associated with improved medical outcomes and quality of life [10]. The high variability within patient priorities raises concern about the viability of an objective standard for informed consent. These results suggest that meeting any one standard for informing patients would likely not address many of the concerns of most individual patients. Thus, the only viable standard for informed consent is the subjective standard. Meeting the subjective standard, however, will demand both extra time and effort in order to determine what is important to each individual patient. We acknowledge that our study included only patients who had received radiotherapy, and that patients with early-stage prostate cancer can be treated with other treatments. Adding more patients to our sample, would not change our observation that each question is considered essential by some patients, and that each question is also considered superfluous by some patients. Addressing the limitation, therefore, is not likely to alter our main conclusion that patients’ information priorities vary considerably
from one another and that a subjective standard is the only viable approach to providing information. Consistent with the notion that Italy has a non-disclosing approach to providing information to patients, most members of each group identified few questions as being essential to address. Interestingly, the information desired by patients was generally less than the amount that the professionals thought should be provided: patients trended toward rating the lowest average number of questions as essential, and rated the highest average number as superfluous. This balance appears different from other cultures (e.g. Canadian results [4,5]) and it is difficult to know why the difference exists. One possibility is that, in this time of increasing globalization of information and sharing, there is a cultural shift occurring in Italy (as elsewhere [11]) regarding information sharing. The differences we found in attitudes between professionals and patients, might therefore, reflect the cultural shift as the patients’ average age was approximately three decades older than that of the professionals. 4.1. Limitations We acknowledge that our speculation about the impact of globalization as reflected in these results raises questions about a similar age difference that exists between Canadian professionals and patients but that the difference in their information priorities is in the opposite direction from that of the Italian professional-patient difference. We suggest that the difference between the countries is the result of a rather complex interaction of globalization, culture, age, and an awareness of time constraints in busy medical practices. As in our understanding of the differences between the Italian professionals and patients, we suggest that globalization has a greater impact on younger people than on those older. The patients of Canada and of Italy, being older, therefore, reflect their respective cultures with less influence of the world around, and the globalization influences on the younger professionals reduces cultural differences between the countries. In addition, the professionals of both countries are cognizant of time constraints of busy medical practices, thus, limit the amount of information they deem essential to address. The net result is that Canadian patients, reflecting their disclosure dominant culture, think more information should be shared than their professionals think, while the Italian patients from their non-disclosure dominant culture think less should be shared than their professionals think. Globalization of information and sharing has further implications for informing patients. The large variation among patient priorities that we found clearly suggests that the only viable standard for informed consent is the subjective standard. However, data suggest that current informing practice in Italy is not to use the subjective standard [12–14]. We suggest that current practice in Italy frequently is to use an objective standard but to base it on North American priorities rather than to develop an Italian standard. Globalization of information makes it quite easy to obtain such
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
information from distant locations. Our patient data suggest, however, that adopting an information sharing standard from another culture without critical appraisal could be risky, especially if it comes from a culture with a different disclosure approach. The complication of globalization does not alter the conclusion of the study that an objective standard, from any source, is not viable in Italy. These results make clear that the only viable standard for providing information to patients in this setting is a subjective standard. 4.2. Practice implications The use of a subjective informing standard means that information provided to each individual patient must be determined with the patient. There are a number of different strategies that can be used. For example, one could identify a core of information to provide, then lead the patient through a process to get him to identify the additional information that he would like to know, perhaps by identifying categories of information and encourage the patient to indicate if they would like to know more. References [1] Novack DH, Plumer R, Smith RL, Ochitill H, Murrow GR, Bennett JM. Changes in physicians’ attitudes toward telling the cancer patient. J Am Med Assoc 1979;241:897–900. [2] Appelbaum PS, Lidz CW, Meisel A. Informed consent: legal theory and clinical practice. New York: Oxford University Press; 1987.
181
[3] Faden RR, Beauchamp TL. A history and theory of informed consent. New York: Oxford University Press; 1986. [4] Feldman-Stewart D, Brundage MD, Hayter C, Groome P, Nickel JC, Downes H, Mackillop W. What questions do patients with curable prostate cancer want answered? Med Decis Making 2000;20:7–19. [5] Feldman-Stewart D, Brundage MD, Hayter C, Groome P, Nickel JC, Downes H, Mackillop W. What prostate cancer patients should know: variation in professionals’ opinions. Radiother Oncol 1998;49:111– 23. [6] Fleming C, Wasson JH, Albertsen PC, et al. A decision analysis of alternative treatment strategies for clinically localized prostate cancer. J Am Med Assoc 1993;269:2650–8. [7] Church JD, Wilke EL. The robustness of homogeneity of variance tests of asymmetric distributions: a monte carlo study. Bull Psychonom Soc 1976;7:417–20. [8] Tattersall MH, Gattellari M, Voigt K, Butow PN. When the treatment goal is not cure: are patients informed adequately? Supp Care Cancer 2002;10:314–21. [9] Grove A. Taking on prostate cancer. Fortune 1996; 55–71. [10] Annunziata MA, Foladore S, Magri MD, Crivellari D, Feltrin A, Bidoli E, Veronesi A. Does the information level of cancer patients correlate with quality of life? A prospective study. Tumori 1998;84:619–23. [11] Horikawa N, Yamazaki T, Sagawa M, Nagata T. Changes in disclosure of information to cancer patients in a general hospital in Japan. Gen Hosp Psychiatry 2000;22:37–42. [12] Annunziata MA, Foladore S, Magri MD, et al. Does the information level of cancer patients correlate with quality of life? A prospective study. Tumori 1998;84:619–23. [13] Mosconi P, Meyerowitz BE, Liberati MC, Liberati A. Disclosure of breast cancer diagnosis: patient and physician reports. GIVIO (Interdisciplinary Group for Cancer Care Evaluation, Italy). Ann Oncol 1991;2:273–80. [14] Pronzato P, Bertelli G, Losardo P, Landucci M. What do advanced cancer patients know of their disease? A report from Italy. Supp Care Cancer 1994;2:242–4.
Information priorities of Italian early-stage prostate cancer patients and of their health-care professionals Carlo Capirci a,∗ , Deb Feldman-Stewart b , Giovanni Mandoliti a , Michael Brundage b , Giorgia Belluco c , Katia Magnani c b
a Department of Radiation Oncology, International Cancer Center, Rovigo 45100, Italy Division of Cancer Care and Epidemiology, Cancer Research Institute, Queen’s University, Kingston, Ont., Canada c International Cancer Center, Psycho-Oncological Service, Rovigo, Italy
Received 18 January 2003; received in revised form 10 November 2003; accepted 2 February 2004
Abstract The study was designed to compare the information priorities of Italian early-stage prostate cancer patients to those of their health-care professionals. Doctors (urologists and oncologists), nurses, radiation therapy technologist (RTs), and patients treated with radiotherapy in Northern Italy were surveyed. Respondents rated the importance of addressing each of 78 questions as: essential, important, or superfluous. We compared priorities between pairs of groups by correlating the percentage of each group that rated each question (a) essential and (b) superfluous. We assessed within-group agreement by identifying questions that >50% of one response. Fifty-nine doctors, 53 nurses, and 45 RTs (overall 70% response rate) and 50 patients (91% response rate) participated. All correlations between professional groups were high (all >0.74). All professional group-patient correlations were significant but much lower than those among the professionals. More importantly, there was considerable variation within each group: within each group, almost all questions were rated essential by some members but less than seven of the 78 questions were rated so by half (or more) of its members. In addition, almost all questions were also rated superfluous by some members of each profession. We concluded that the professional groups appear to agree with one another on questions essential and those superfluous to address more than they agree with a group of their patients. More importantly, there is considerable disagreement within each group. The results imply that the professionals cannot assume that their own information priorities are the same as those of their patients. © 2004 Published by Elsevier Ireland Ltd. Keywords: Prostate cancer patients; Informed consent; Professionals’ opinions
1. Introduction The Hippocratic Oath makes it clear that clinicians have always been interested in promoting patient well-being, and historically, withholding information from patients was considered by some to be beneficent [1]. More recently providing information has been viewed as beneficent and assessing the adequacy of provision of information in medical situations has been driven to a large extent by the desire to ensure that “informed consent” has been obtained. The term “informed consent” was introduced in 1957 by Paul Gebhart, a dean of medical association lawyers. In the same year, the California court adopted the expression “informed consent” in a malpractice case [2]. This was the first time a court ∗ Corresponding author. Tel.: +39-0425-393292; fax: +39-0425-393373. E-mail address: rt [email protected] (C. Capirci).
0738-3991/$ – see front matter © 2004 Published by Elsevier Ireland Ltd. doi:10.1016/j.pec.2004.02.006
had adopted “informed consent” to represent an obligation of physicians and other health-care professionals to provide patients with sufficient information about recommended diagnostic and treatment procedures, including potential risks and complications. While the legal doctrine developed for informed consent, so did its ethical framework. Three principles underlie informed consent: support for patients’ autonomy, beneficence, and justice [3]. The relative weight of the principles could differ for different cultures, as could the views of how the principles are best implemented. Such differences may be, for example, what underlies the difference between “non-disclosure dominant” cultures like Southern Europe and Japan, and “disclosure dominant” cultures like Canada, United States, and Northern Europe. The justice principle in the context of information provision means that medically similar patients should be provided with the same opportunity to receive particular
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
pieces of information, including from different physicians. Such an information core could be used as a legal professional standard for disclosure in informed consent. In the decision-making process, however, several elements that could affect the decision require an assessment by the doctor which is at least partly subjective. Such assessments include those related to quality of life, to disease aggressiveness and possible outcomes of treatment, to psychological acceptance of disease, and treatment toxicity. Although there may be subjective elements in assessments of the patient, personal values held by the physician could violate the ethical principle of justice, and also that of autonomy: patients have the right to act with intention, with understanding and without controlling influences [3]. Concern that applying the legal professional standard does not respect patient autonomy has led to its replacement by the “objective” standard: it requires physicians to provide the information that a “reasonable person” would consider material to the decision under the specified circumstances [2]. In practice the “reasonable person” has been the jury, which in turn led to concern that the jury did not represent patients’ needs well. The concern led to substituting the “objective” standard with the “subjective” standard, requiring that physicians provide the information material to the particular patient’s decision. Thus, the patient defines the information to be provided. Although physicians in most jurisdictions are legally responsible to ensure that patients are providing informed consent, they frequently are not the only health-care professionals to provide patients with information. In some places, for example, nurses are charged with the responsibility of educating the patient. In the delivery of radiotherapy, the radiation therapy technologists (RTs) see the patient on a daily basis, providing information about procedural issues in addition. Thus, to determine information priorities of those around the patient, it is important to include all health-care professionals that the patient encounters. The aim of this study was to determine the extent of agreement among Italian health-care professionals, among patients, and between the professionals and patients on what questions should be addressed with curable prostate cancer patients before treatment decisions are made. We focussed on only curable prostate cancer because the distinction between having curable versus non-curable disease by its very nature is likely to result in different issues being considered important. Although we asked participants to focus on the time before treatment decisions were made, we recognize that providing information to patients at that time can be for many reasons of which decision-making is only one [4]. The study used the methodology developed by FeldmanStewart et al. [5] to address these issues in a Canadian setting. The method was used because it included a list of relevant issues that were detailed in nature; most other studies of information priorities in prostate cancer have focussed on larger categories of information. Although Canada has been identified as a disclosure dominant culture and Italy as
175
non-disclosure dominant, we felt it appropriate to use the question list developed in Canada to study Italian priorities: if Italy really is non-disclosure dominant, then participants would identify few questions as important to address, if it is not, then the comprehensive Canadian list would provide a broad basis from which to identifying priorities.
2. Methods 2.1. Design Four groups participated in the study: three groups of professionals, doctors , nurses, and radiation therapy technologists, and patients. All professionals were involved in treating patients with prostate cancer at the time of the study and were recruited from eight clinics in Triveneto of Northern Italy. The doctors included 10 urologists, 11 medical oncologists, and 38 radiation oncologists. Because the numbers of urologists and of medical oncologists were so few, their responses were added the radiation oncologists to create the doctor group. Patients were also recruited from the Triveneto region and all had received radiotherapy, some received their radiotherapy after surgery, and some received hormones after their radiotherapy. The data were gathered through a survey. For the professionals, the survey included a scenario describing a man presenting with curable prostate cancer. The man’s age and state of health made him eligible for any of three potential medical treatments: radiation therapy, surgery or watchful waiting [6]. The patients were already treated patients who were asked to consider their judgments as a recommendation for questions that should be addressed with a newly diagnosed patient in a situation like their own. We felt that newly diagnosed patients could benefit from the insights of the patients have been through the treatment process. The respondent was asked to judge the importance of addressing each of 78 questions before the patient’s treatment decision was made. Response options were “essenziale” (essential), “importante” (important), or “superflua” (superfluous). The instructions asked the professional respondents to focus on which questions should be addressed regardless of patient queries. After judging the importance of addressing the various questions, the respondent filled out a form providing personal information that included sex (for the professionals) and age. 2.2. Statistical analyses To compare relative information priorities between pairs of groups, we used Pearson product moment correlation coefficients to assess the associations between the percentages of each group that rated each question using a particular response. To assess the extent of agreement between pairs of groups (e.g. do doctors agree with nurses more than they agree with patients), we used Hartley’s Fmax [7] to compare
176
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
the proportion of accounted-for variance in one situation to that of the second. We compared the absolute amount of information deemed essential between pairs of groups using t-tests to compare the mean number of questions each rated essential. We also compared agreement within the groups. Within each group, we identified the questions that at least 50% of respondents rated essential as questions they “agreed” were essential to address. Correlations were used to determine if there were associations between the number of essential responses and demographic/professional characteristics. For the continuous variable age, the correlations were Pearson correlation coefficients. Correlations with the dichotomous variable sex were point biserial correlations. In evaluating the predictive ability of the correlations, we squared the correlation to obtain the amount of accounted-for variance.
3. Results 3.1. Respondents A description of respondents in each of the groups is shown in Table 1. As the table shows, the most notable difference among the groups is their age, with the average age of the patients almost twice that of the professionals. 3.2. Between-group comparisons 3.2.1. Comparisons on content Table 2 shows the Pearson correlation coefficients comparing the percentage of particular responses on each quesTable 1 Description of respondents in each group
Doctors Nurses RTs Patients
N
Male:female
Mean age
Response rate
59 53 45 50
40:19 12:41 22:23 50:0
43 38 37 70
69% 69% 78% 92%
(59/85) (53/77) (45/58) (56/61)
Table 2 Pearson correlation coefficients comparing overall priorities between pairs of groups Nurses
RTs
Patients
Essential questions Doctors Nurses RTs
0.82 – –
0.86 0.75 –
0.63 0.64 0.68
Superfluous questions Doctors Nurses RTs
0.84 – –
0.84 0.87 –
0.48 0.47 0.40
Table 3 For each professional group, Fmax comparing their agreement with the other professional groups to their agreement with the patients Professional agreement-essential questions Doctor–Nurse
Doctor–RT
Patient agreement with professionals Patient–Doctor 1.69∗ 1.86∗ Patient–Nurse 1.34 Patient–RT 1.60∗
Nurse–RT
1.37 1.22
Professional agreement-superfluous questions Doctor–Nurse
Doctor–RT
Patient agreement with professionals Patient–Doctor 3.06∗ 3.06∗ Patient–Nurse 3.19∗ Patient–RT 4.41∗ ∗
Nurse–RT
3.43∗ 4.73∗
P < 0.05.
tion, between pairs of groups; the top half of the table shows the comparisons of essential responses and the bottom half the comparisons of superfluous responses. As can be seen in the table, all of the correlations comparing pairs of professional groups are relatively high and account for at least 50% of the variance. Thus, the questions that most people within one profession rated essential tended to be similar to the questions that most people in each of the other professions rated essential; and similarly for superfluous questions. Table 2 also shows that the correlation coefficients drop across the board when comparing the patients to each professional group as compared to the correlations between pairs of professional groups, both when considered essential questions and when considering superfluous questions. Table 3 shows a comparison of the extent of agreement between pairs of professional groups to their agreement with the patients. As the table shows when considering essential questions, the low doctor–patient agreement on essential questions means that doctors agree significantly more with each of the professional groups than they do with the patients. The table also shows that, although patients’ essential priorities were most similar to RTs’ priorities, there was still less agreement between the RTs and patients than between the RTs and doctors. As for superfluous questions: the table shows that each professional group agrees with the other professional groups significantly more than they agree with the patients. The questions were then sorted into one of four categories according to their focus; 36 focussed on the disease, 20 on treatment details, 10 on side effects, and 12 questions on treatment alternatives. Fig. 1 shows the correlation coefficients comparing relative essential priorities of each professional group with those of the patients in each of the categories and overall. Although the low numbers of questions in each category means the confidence intervals around the correlations are large, the figure suggests that patient-professional agreement is highest overall with RTs and that highest agreement is consistent over three categories
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
177
Fig. 1. Peason correlation coefficients comparing each professional group to patients. The correlations compare the percentage of essential responses to each question between the patients and the particular professional group labeled. The bars on the left show the overall comparison that included all questions. The disease category included 36 questions, the treatment details category 20 questions, the side effects category 10 questions and the alternatives category 12 questions.
of the four categories; patient priorities around side effects appear most similar to the priorities of the nurses.
identified significantly fewer than the doctors and the nurses [F(1, 195) = 2.0, P < 0.05].
3.2.2. Comparisons on amount The average number of questions rated essential by individual participants was similar across groups: 17, 18, 17, and 15, for doctors, nurses, RTs, and patients, respectively (all t < 1). The average number of questions rated superfluous by individual participants was 20 by doctors and by nurses, 15 by RTs, and 25 by patients, which differed significantly from each other [F(3, 195) = 4.3, P < 0.01]; orthogonal contrasts showed that the patients identified significantly more superfluous questions than all of the professionals [F(1, 195) = 3.07, P < 0.01] and that the RTs
3.3. Within-group agreement 3.3.1. Agreement on content Agreement on content was investigated from two perspectives: on essential questions as content to include and on superfluous questions as content to exclude. Using a simple majority to define “agreement” within groups, each group agreed on few questions as essential: six for the doctors, two for the nurses, five for the RTs, and two for the patients. Table 4 lists the questions in their order of percentage essential responses (highest to lowest) within each
Table 4 For each group, questionsa rated essential by at least 50% of members Doctors
Nurses
RTs
Patients
3. Will the treatment cure my disease?
3. Will the treatment cure my disease? 2. Will the treatment make my disease go into remission?
3. Will the treatment cure my disease? 2. Will the treatment cause my disease to go into remission? 4. Do I have the right to refuse the treatment? 11. If my prostate cancer is not treated, will I die from it? 22. If my disease is not cured, how long will I live?
3. Will the treatment cure my disease? 11. If my disease is not treated, will I die from it?
2. Will the treatment cause my disease to go into remission? 4. Do I have the right to refuse the treatment? 10. Are there other treatment possibilities not mentioned? 11. If my prostate cancer is not treated, will I die from it? 23. Could the treatment lead to my death? a
Questions ordered by their percentage of essential responses (highest to lowest).
178
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
Table 5 For each group, questionsa rated superfluous by at least 50% of members Doctors
Nurses
RTs
Patients
75. Did I get my disease from what I eat?
75. Did I get my disease from what I eat?
72. What is the cost of the treatment? 57. How common is the disease I have?
72. What is the cost of the treatment? 74. How many prostate cancer patients has my physician treated? 63. Can I spread prostate cancer to other people? 69. If my disease is not treated, will it make me sick to my stomach? 73. Did I get my disease from my work?
76. How many prostate cancer patients have been cured by my physician? 72. What is the cost of the treatment? 75. Did I get my disease from what I eat?
46. Why is it sometimes said that patients don’t live any longer with early diagnosis? 72. What is the cost of the treatment? 60. Will the treatment change the way I look? If so, for how long?
58. Am I different from the usual patient with prostate cancer? 76. How many prostate cancer patients have been cured by my physician? 69. If my disease is not treated, will it make me sick to my stomach? 70. How long have I had prostate cancer? 73. Did I get my disease from my work? 61. If my disease is not treated, will it make me tired? 66. If my disease is not treated, will it change the way I look? 46. Why is it sometimes said that patients don’t live any longer with early diagnosis? 59. How many patients like me have decided to have this treatment? 71. Can the treatment stimulate my disease to grow or spread? a
73. Did I get my disease from my work? 77. What special qualifications does my doctor have to treat my disease? 56. Will the treatment cause my hair to fall out? If so, for how long?
76. How many prostate cancer patients have been cured by my physician?
Questions ordered by their percentage of superfluous responses (highest to lowest).
group. As the table shows, although there are very few questions that members agreed were essential, three of the questions were identified by three of the four groups. We note that we compared responses of the three doctor groups and, acknowledging that the medical oncologist and urologist numbers are too small to conclude with certainty, but the results of three groups appear to show very similar patterns. While, within each group, there were few questions that a simple majority of members rated essential, almost every question was rated so by some its members. Every question was rated essential by some doctors and by some nurses. Only one question was not rated essential by any RTs, and two not rated so by any patients. On average each question was rated essential by 21, 23, 22, and 19% by the doctors, nurses, RTs, and patients, respectively. Interestingly, while there were a few questions not considered superfluous by each of the professional groups (one, one, and three by doctors, nurses, and RTs, respectively), every question was considered so by some patients. In addition, on average, more patients rated each question as superfluous than did the professional groups: the average per
question was 21, 26, 19, and 32% by the doctors, nurses, RTs, and patients, respectively. In spite of the higher average patient percentage, the doctors had the most questions rated as superfluous by at least 50% of its members; there were 13, 7, 3, and 6 questions that reached that threshold by the doctors, nurses, RTs, and patients, respectively. The particular questions rated superfluous by at least 50% of its members are listed in Table 5. 3.3.2. Agreement on amount The extent of disagreement on the number of questions that individual members within each profession rated essential can be seen in Fig. 2. It shows the number of questions deemed essential by individual participants within each group, each panel showing one group. As can be seen in each panel, there was wide variation in the numbers of questions deemed essential by members within each group: while 14% of doctors thought ≤5 questions were essential to address, another 27% thought ≥20 were essential. Those two extreme proportions for each of the other groups were 21 and 34% for the nurses, 9 and 27% for RTs, and 34 and 16% for the patients.
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
179
Fig. 2. Number of essential questions per participant in each group.
3.4. Associations with demographic characteristics We expected that the number of questions considered essential would lower with age and with male gender. There was, however, no association between age and the number of questions rated essential among those who reported their age of the doctors, nurses, or patients (r = −0.01, 0.15, −0.03, respectively, all P > 0.2); among the RTs, there was a trend for fewer questions to be rated essential as RTs got older (r = −0.26, P = 0.1). There was no association between sex and number of essential questions among doctors or nurses (r = −0.06, 0.00, respectively, both P > 0.7). Among RTs, however, males identified fewer essential questions than females (r = −0.43, P < 0.01). When controlling for sex, among RTs, there was absolutely no association between age and the number of essential questions (r = 0.02, P = 0.74); when controlling for age, there was still a trend for males to identify fewer essential questions than the females (r = −0.14, P = 0.1).
4. Discussion and conclusions The results of this study suggest that the information priorities of health-care professionals treating early-stage prostate cancer patients who receive radiotherapy in Italy are more similar to one another than they are to the priorities of
their patients. We observed, for example, significantly more agreement on what questions are essential between the doctors and both professional groups than between the doctors and the patients. When considering the relative amount of agreement between the patients and each professional group in each sub-category of questions, it appears that questions about side effects stand out as anomalous: the priorities seem most similar between the patients and RTs in all subcategories except side effects where patient priorities appear to agree most with the nurse priorities. It is interesting to wonder if the apparent higher level of agreement between the RTs and patients is related to the fact that of the three professional groups, the RTs spend most time with the patients. The higher agreement about side-effect priorities between the nurses and patients may reflect the fact that side effects are most often treated by the nurses. These results raise the possibility that different professional groups might specialize in the categories of information that they provide to patients. Higher agreement among the professional groups than between the professionals and the patients was also observed in the questions considered superfluous. The patients consistently identified more questions as being superfluous than did any of the professional groups. The results suggest that in determining information to be shared with patients, one needs to pay attention not only to information that should
180
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
be shared but also to information that need not be shared. It is difficult to know if the superfluous information is, in fact, information that would offend patients if provided or that it is just not of interest; however, providing the unwanted information risks at the minimum, overwhelming patients which, in turn, can cause anxiety and frustration [8]. More important than the extent of agreement at the group level, however, is the fact that there is a large amount of variability within each of the groups surveyed. Within each group, there were very few questions that even just a simple majority of members thought were essential to address. However, everyone surveyed indicated that some questions were essential. And almost every question was considered essential by some members of every group. Similarly, almost every question was considered superfluous by some members of every group. The implication of these observations are discussed below. The extent of variation within the doctors suggests that developing a professional standard for informed consent would require much effort in order to arrive at a consensus. The high variability in priorities among the doctors also raises concern about violations to the justice principle. To the extent that information provided to patients is a function of the individual doctor’s information priorities, these results suggest that all patients would not be provided with access to the same information. In addition, high variability within all three professional groups means that an individual patient who sees multiple professionals, from either the same or different professions, is likely to face different information priorities which, in turn, may contribute to patient confusion that has been reported [9]. Although no doubt beneficial for patients if they didn’t have to face confusion generated by disagreement amongst the professionals, using a professional standard for disclosure is questionable: providing patients with the information that they want has been associated with improved medical outcomes and quality of life [10]. The high variability within patient priorities raises concern about the viability of an objective standard for informed consent. These results suggest that meeting any one standard for informing patients would likely not address many of the concerns of most individual patients. Thus, the only viable standard for informed consent is the subjective standard. Meeting the subjective standard, however, will demand both extra time and effort in order to determine what is important to each individual patient. We acknowledge that our study included only patients who had received radiotherapy, and that patients with early-stage prostate cancer can be treated with other treatments. Adding more patients to our sample, would not change our observation that each question is considered essential by some patients, and that each question is also considered superfluous by some patients. Addressing the limitation, therefore, is not likely to alter our main conclusion that patients’ information priorities vary considerably
from one another and that a subjective standard is the only viable approach to providing information. Consistent with the notion that Italy has a non-disclosing approach to providing information to patients, most members of each group identified few questions as being essential to address. Interestingly, the information desired by patients was generally less than the amount that the professionals thought should be provided: patients trended toward rating the lowest average number of questions as essential, and rated the highest average number as superfluous. This balance appears different from other cultures (e.g. Canadian results [4,5]) and it is difficult to know why the difference exists. One possibility is that, in this time of increasing globalization of information and sharing, there is a cultural shift occurring in Italy (as elsewhere [11]) regarding information sharing. The differences we found in attitudes between professionals and patients, might therefore, reflect the cultural shift as the patients’ average age was approximately three decades older than that of the professionals. 4.1. Limitations We acknowledge that our speculation about the impact of globalization as reflected in these results raises questions about a similar age difference that exists between Canadian professionals and patients but that the difference in their information priorities is in the opposite direction from that of the Italian professional-patient difference. We suggest that the difference between the countries is the result of a rather complex interaction of globalization, culture, age, and an awareness of time constraints in busy medical practices. As in our understanding of the differences between the Italian professionals and patients, we suggest that globalization has a greater impact on younger people than on those older. The patients of Canada and of Italy, being older, therefore, reflect their respective cultures with less influence of the world around, and the globalization influences on the younger professionals reduces cultural differences between the countries. In addition, the professionals of both countries are cognizant of time constraints of busy medical practices, thus, limit the amount of information they deem essential to address. The net result is that Canadian patients, reflecting their disclosure dominant culture, think more information should be shared than their professionals think, while the Italian patients from their non-disclosure dominant culture think less should be shared than their professionals think. Globalization of information and sharing has further implications for informing patients. The large variation among patient priorities that we found clearly suggests that the only viable standard for informed consent is the subjective standard. However, data suggest that current informing practice in Italy is not to use the subjective standard [12–14]. We suggest that current practice in Italy frequently is to use an objective standard but to base it on North American priorities rather than to develop an Italian standard. Globalization of information makes it quite easy to obtain such
C. Capirci et al. / Patient Education and Counseling 56 (2005) 174–181
information from distant locations. Our patient data suggest, however, that adopting an information sharing standard from another culture without critical appraisal could be risky, especially if it comes from a culture with a different disclosure approach. The complication of globalization does not alter the conclusion of the study that an objective standard, from any source, is not viable in Italy. These results make clear that the only viable standard for providing information to patients in this setting is a subjective standard. 4.2. Practice implications The use of a subjective informing standard means that information provided to each individual patient must be determined with the patient. There are a number of different strategies that can be used. For example, one could identify a core of information to provide, then lead the patient through a process to get him to identify the additional information that he would like to know, perhaps by identifying categories of information and encourage the patient to indicate if they would like to know more. References [1] Novack DH, Plumer R, Smith RL, Ochitill H, Murrow GR, Bennett JM. Changes in physicians’ attitudes toward telling the cancer patient. J Am Med Assoc 1979;241:897–900. [2] Appelbaum PS, Lidz CW, Meisel A. Informed consent: legal theory and clinical practice. New York: Oxford University Press; 1987.
181
[3] Faden RR, Beauchamp TL. A history and theory of informed consent. New York: Oxford University Press; 1986. [4] Feldman-Stewart D, Brundage MD, Hayter C, Groome P, Nickel JC, Downes H, Mackillop W. What questions do patients with curable prostate cancer want answered? Med Decis Making 2000;20:7–19. [5] Feldman-Stewart D, Brundage MD, Hayter C, Groome P, Nickel JC, Downes H, Mackillop W. What prostate cancer patients should know: variation in professionals’ opinions. Radiother Oncol 1998;49:111– 23. [6] Fleming C, Wasson JH, Albertsen PC, et al. A decision analysis of alternative treatment strategies for clinically localized prostate cancer. J Am Med Assoc 1993;269:2650–8. [7] Church JD, Wilke EL. The robustness of homogeneity of variance tests of asymmetric distributions: a monte carlo study. Bull Psychonom Soc 1976;7:417–20. [8] Tattersall MH, Gattellari M, Voigt K, Butow PN. When the treatment goal is not cure: are patients informed adequately? Supp Care Cancer 2002;10:314–21. [9] Grove A. Taking on prostate cancer. Fortune 1996; 55–71. [10] Annunziata MA, Foladore S, Magri MD, Crivellari D, Feltrin A, Bidoli E, Veronesi A. Does the information level of cancer patients correlate with quality of life? A prospective study. Tumori 1998;84:619–23. [11] Horikawa N, Yamazaki T, Sagawa M, Nagata T. Changes in disclosure of information to cancer patients in a general hospital in Japan. Gen Hosp Psychiatry 2000;22:37–42. [12] Annunziata MA, Foladore S, Magri MD, et al. Does the information level of cancer patients correlate with quality of life? A prospective study. Tumori 1998;84:619–23. [13] Mosconi P, Meyerowitz BE, Liberati MC, Liberati A. Disclosure of breast cancer diagnosis: patient and physician reports. GIVIO (Interdisciplinary Group for Cancer Care Evaluation, Italy). Ann Oncol 1991;2:273–80. [14] Pronzato P, Bertelli G, Losardo P, Landucci M. What do advanced cancer patients know of their disease? A report from Italy. Supp Care Cancer 1994;2:242–4.