Psychological responses of patients receiving a diagnosis of cancer

P. E. Schofield1,+, P. N. Butow2, J. F. Thompson3, M. H. N. Tattersall4, L. J. Beeney5 and S. M. Dunn6

1 Department of Haematology & Medical Oncology, Peter MacCallum Cancer Institute, Victoria; 2 Medical Psychology Research Unit, University of Sydney, NSW; 3 Department of Surgery, University of Sydney and Sydney Melanoma Unit, Gloucester House, Royal Prince Alfred Hospital, Camperdown, NSW; 4 Department of Cancer Medicine, University of Sydney, NSW; 5 Department of Psychology, University of Sydney, NSW; 6 Department of Psychological Medicine, Royal North Shore Hospital, St Leonards, NSW, Australia

Received 14 February 2002; revised 9 July 2002; accepted 18 July 2002


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background:

Current recommendations on how to break bad news are primarily based on expert opinion. Little is known about the association between communication practices and patients’ psychological response.

Patients and methods:

One-hundred and thirty-one patients with newly diagnosed melanoma were surveyed 4 months after the initial consultation at the Sydney Melanoma Unit regarding their communication experiences and their satisfaction with these experiences. They completed the Hospital Anxiety and Depression Scale (HADS) at this time, and 4 and 13 months later.

Results:

Both patients’ satisfaction with communication and their psychological morbidity were found to be associated with particular communication practices. Practices linked to lower anxiety included preparing the patient for a possible diagnosis of cancer; having the people wanted by the patient present to hear the diagnosis; giving the patient as much information about the diagnosis as desired; providing written information; presenting the information clearly; discussing the patient’s questions the same day; talking about the patient’s feelings; and being reassuring. Practices linked with lower levels of depression included using the word ‘cancer’; discussing the severity of the situation, life expectancy and how the cancer might affect other aspects of life; and encouraging the patient to be involved in treatment decisions.

Conclusions:

This study provided preliminary evidence that communication strategies recommended in the literature produce positive patient outcomes. Further studies are needed which document actual communication.

Key words: breaking bad news, communication, diagnosis, melanoma, psychological adjustment, truth telling


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Most in the medical community now agree that patients have a moral and legal right to know the truth about their illness [1]. However, the medical principle of primum non nocere may be threatened if a blanket policy of full disclosure is adopted, since not all patients wish to be given all the details of their illness [2]. In an area dominated by expert opinion, the pertinent issue is to determine which communication practices make a measurable difference to patient outcome. The present study investigates this question in a sample of patients diagnosed with a thick and therefore potentially life-threatening primary melanoma.

Several commentators have emphasised the importance of imparting individually tailored information with hope and empathy [1, 36]. The key features of this approach are flexibility according to patient needs, obtaining patient feedback, encouraging positive thinking and providing hope and reassurance. Patients have individualised requirements for information and involvement in their treatment plan, and different ways of coping with a diagnosis of cancer [2].

Specific communication practices for delivering a diagnosis of cancer have primarily derived from expert opinion, consensus and some small descriptive studies. The lack of research has led to some quite distinct variations in recommendations [79]. For example, some experts say there should be no delay in telling [1] whereas others argue it is best if the clinician selects the most appropriate time and place to tell [7, 10]. Some say that it is best to have family members present at diagnosis [1, 11], but others maintain it should be the patient’s choice [4]. Moreover, reports from physicians suggest that there are wide variations in the extent to which recommended communication practices are followed [12].

There is a need to establish which communication practices optimise patients’ experiences. Physicians and nurses believe that the way in which bad news is broken influences the patient’s subsequent psychosocial adjustment [13]. Considerable debate surrounds methods of determining which of numerous patient outcomes should be selected as the most appropriate indicators for effective communication. Patient and clinician preferences for communication are an obvious option [2].

Patient satisfaction is a popular outcome measure and has been positively related to information provision, uninterrupted time spent with the patient in an appropriate environment, doctors’ social skills and empathy including eliciting patient understanding and preferences [14, 15]. While patient satisfaction is a laudable outcome, Ong et al. [14] question whether it is sufficient, given that the purpose of medical care is to optimise the health and well-being of patients. Hence, some argue that psychological adjustment should be the ‘gold-standard’ outcome, as it is an important component of overall patient health status [7].

Patients’ assessment of their clinician’s communication skills [8, 16, 17] and being offered a choice in treatment [18] have been significantly related to patients’ subsequent psychological adjustment. However, these studies have been limited by design deficits [8, 16] or have assessed only a limited range of communication practices [18]. Very few studies on doctor–patient communication have investigated how prognosis or treatment options are communicated.

The primary aim of the present study was to determine which communication strategies recommended in the literature and endorsed by patients optimise patients’ responses to receiving a diagnosis of melanoma using two indicators: patient reported satisfaction with communication and psychological morbidity (anxiety and depression) over 17 months following diagnosis.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Setting
This study was conducted at the Sydney Melanoma Unit (SMU) at the Royal Prince Alfred Hospital in Sydney, Australia.

Sample
A sample of 194 consecutive new patients, with recently diagnosed melanoma attending the SMU for treatment by or clinical opinion from one of five surgeons, were screened for inclusion. All patients were seen at SMU within 4 weeks of their melanoma diagnosis being established by excision (or occasionally incision) biopsy. Inclusion criteria for patients were as follows: melanomas with a Breslow thickness of >1.5 mm; >18 years of age; English speaking and contactable by telephone. Nine patients were excluded because their treating surgeon thought that they would be unable to cope with the demands of the study. Of the 150 patients who were deemed eligible, three had died before initial contact and 14 refused to participate. Characteristics of the 14 who refused were compared with those who responded, using available data including age, gender, site and thickness of melanoma. No differences were found, except that those who refused were more likely to be female [{chi}2 (1) = 4.90; P = 0.03]. The remaining 133 patients completed the questionnaire; however, the responses of two patients were lost.

Design and procedure
This study was carried out with ethics committee approval. The participation of patients was sought via a letter signed by their surgeon or by a researcher 1–2 months after their initial consultation at the SMU. Patients were then contacted by telephone. Those who consented were sent a standardised consent form to complete. Consenting patients were posted the baseline questionnaire to be completed at home and returned by mail. The mean interval from the initial consultation at the SMU to questionnaire completion was 3.8 months. At this time, information about the extent of the disease and treatment options will have been disclosed and the high initial anxiety generated by the diagnosis dissipated [19] while recall will still be vivid [20]. The baseline questionnaire contained items about the patients’ recollections of their communication experiences and preferences; their satisfaction with the consultation and their psychological morbidity. Four months after the baseline questionnaire was returned, the second questionnaire was posted to patients at home and 119 questionnaires were completed and returned. The third and final questionnaire was posted to patients 13 months after baseline and 109 were returned completed. Only psychological morbidity was assessed in the two follow-up questionnaires.

Patient characteristics
Patients’ SMU records provided information about demographics and disease characteristics, including melanoma site and thickness, and cancer stage.

Communication experiences and preferences
Patients were asked to recall the time when they were first told that they had cancer and to report their experiences. The sections relevant to this paper dealt with three aspects of the consultation: diagnosis, treatment options and implications of the diagnosis (including prognosis). Items covered whether the patient was prepared for the diagnosis, whether there was a delay in telling, who told the diagnosis, how it was told and in what circumstances (environment and presence of kin), how much information was given about diagnosis, treatment and prognosis and how much the patient wanted, whether the doctor was reassuring/hopeful and willing to discuss patient feelings, when the patient was able to ask questions and whether the patient had a say in the treatment plan. These items cover nearly all of the issues raised in guidelines relating to breaking bad news [1, 37, 9, 11, 14]. Most items were adapted from ones used in a previous study of communication in cancer patients [8]. Additional items were developed to cover areas found to be important from focus groups with melanoma patients and a literature review of expert opinion. Each item comprised a stem and a short list of response categories (Tables 1, 2 and 3).


View this table:
[in this window]
[in a new window]
 
Table 1. Relationship between communication factors related to diagnosis and patients’ satisfaction, anxiety and depression levels
 

View this table:
[in this window]
[in a new window]
 
Table 2. Relationship between communication factors related to prognosis and patients’ satisfaction, anxiety and depression levels
 

View this table:
[in this window]
[in a new window]
 
Table 3. Relationship between communication factors related to treatment and patients’ satisfaction, anxiety and depression levels
 
Patient satisfaction
Satisfaction with communication about diagnosis, prognosis and treatment was measured using three items with the structure: "How would you rate the way ‘the diagnosis of cancer, or the implications of the diagnosis, or treatment for your cancer’ was discussed with you?". The response options were excellent, good, satisfactory, inadequate and poor. The satisfaction measures were developed for this current study as existing measures did not assess the particular issues relevant to breaking bad news: specifically diagnosis, prognosis and treatment discussions.

Psychological morbidity
The Hospital Anxiety and Depression Scale (HADS) [21] was used to measure psychological morbidity. It has two subscales, comprising seven items each to assess depression and anxiety separately.

Analysis and treatment of data
Response categories that were conceptually similar, such as ‘strongly agree’ and ‘agree’, were amalgamated to provide sufficient cases in each cell for analysis. For the satisfaction measures, the response options were collapsed into two groups: high (‘excellent’ and ‘good’) and mid to low satisfaction (‘satisfactory’, ‘inadequate’ and ‘poor’). Wilcoxon sign tests were used to investigate differences between levels of satisfaction for matters related to diagnosis, implications of diagnosis and treatment. Chi-square analysis was used to examine differences in satisfaction levels in relation to different communication experiences. To assess the relationship between communication experiences and levels of psychological morbidity over time, a series of multivariate analyses of variances (MANOVAs) was employed. The between-subjects factors were the communication factors. The within-subjects factors were psychological morbidity (either anxiety or depression) at the three time periods. The analyses were performed separately for anxiety and depression and for each of the items assessing communication factors. Only the significant associations are reported in the text. The significance level was set at P = 0.05.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Description of sample
The majority of the sample was male (60%). Mean age was 58 years and ranged between 19 and 88 years. The majority (57%) had some secondary school education, 15% had completed secondary school, 8% had attended technical college and 15% had a university degree (with 5% missing data). Nearly three-quarters of the sample was married or in a de facto relationship, 8% were single, 8% widowed and 8% separated or divorced (with 3% missing data). The most common site for the melanoma was on the limbs (46%), followed by the trunk (44%) and 14% were located on the head or neck. Less than 1% were occult. Eighty-six per cent of patients presented with localised disease and 10% with lymph node involvement (with 4% missing data). The mean Breslow thickness was 3.14 mm (SD 1.84 mm) with a range of 1.5–12.0 mm.

The majority of patients were very satisfied with communication, with 64%, 57% and 79% of patients describing the communication as ‘excellent’ or ‘good’ for diagnosis, prognosis and treatment discussions, respectively. However, patients expressed greater satisfaction with discussions about diagnosis compared with discussions about prognosis (z = 3.39; P = 0.0007) and patients were more satisfied with discussions of treatment plans than discussions of diagnosis (z = 3.60; P = 0.0003) or prognosis (z = 5.65; P = 0.0001).

Communicating the diagnosis
Satisfaction. Table 1 presents results for the relationship between diagnosis communication factors and patients’ satisfaction and psychological morbidity. Patients who felt that their doctor prepared them for the diagnosis reported comparatively higher levels of satisfaction ({chi}2 = 26.49; df = 2; P = 0.00001). Patients who reported that the length of time between first seeking help and confirmation of cancer was ‘about right’ were significantly more satisfied ({chi}2 = 13.40; df = 2; P = 0.0012). Using the word ‘cancer’ was associated with higher levels of satisfaction ({chi}2 = 11.65; df = 1; P = 0.0006). Having the family present for diagnosis was linked to higher satisfaction ({chi}2 = 4.13; df = 1; P = 0.04), as was having the people the patient wanted present, including being alone with the doctor if that was the patients’ preference ({chi}2 = 10.47; df = 1; P = 0.0012). Highest satisfaction was linked to the patient reporting they had been told ‘everything’ ({chi}2 = 35.70; df = 2; P = 0.00001) and agreeing the information had been presented clearly ({chi}2 = 53.02; df = 3; P = 0.00001). Satisfaction was also linked to the clinician individually matching the amount of information given with the amount wanted by the patient ({chi}2 = 18.62; df = 1; P = 0.00002). Similar results were obtained for the provision of written information ({chi}2 = 3.77; df = 1; P = 0.05) and individually matching the provision of written information to patient needs ({chi}2 = 12.96; df = 1; P = 0.00032). Patients agreeing with the statements that the doctor was willing to discuss patient feelings ({chi}2 = 20.69; df = 1; P = 0.00001) and that the doctor was reassuring ({chi}2 = 13.10; df = 1; P = 0.00029) reported higher satisfaction than patients who did not agree.

Anxiety. Patients who were prepared for the diagnosis experienced lower levels of anxiety (F = 7.52; df = 2, 106; P = 0.001). Lower anxiety was also observed amongst patients who had the people present who they wanted present to hear the diagnosis, including being alone with the doctor (F = 8.89; df = 1, 107; P = 0.004). Patients who perceived that the amount of information given to them at diagnosis was ‘everything’ and who agreed that it was clearly presented reported lower levels of anxiety than those who did not (F = 5.77, df = 2, 103, P = 0.004; F = 3.48, df = 3, 105, P = 0.018, respectively). Being given as much information about the diagnosis as desired, be that everything or as little as possible, was also linked to lower anxiety (F = 4.36; df = 1, 105; P = 0.039). Similarly, individually matching the provision of written information to patient needs was linked to lower anxiety (F = 17.11; df = 1, 102; P = 0.0001). Attending to the patient’s emotional reactions, specifically the clinician being reassuring and willing to discuss patient feelings, was significantly related to lower anxiety (F = 5.96; df = 1, 104; P = 0.016; F = 15.72; df = 1, 102; P = 0.0001, respectively).

Depression. Individually matching the provision of written information to patient needs was linked to lower depression (F = 5.83; df = 1, 102; P = 0.018). While the main effect for the communication factor, using the word ‘cancer’ was not significantly related to depression (F = 0.02; df = 1, 103; P = 0.891); there was a significant interaction effect between this factor and time (F = 3.07; df = 2, 206; P = 0.048). Simple contrasts revealed that compared with others, patients who reported that the word ‘cancer’ was used expressed lower levels of depression at baseline; however, there was no difference between the two groups at later time periods.

Prognosis and implications of the diagnosis
Satisfaction. Table 2 presents the results for the relationship between prognosis communication factors and patients’ satisfaction and psychological morbidity. Patients who reported that their doctor had talked about the severity of the cancer expressed higher levels of satisfaction ({chi}2 = 36.50; df = 1; P = 0.00001). Similar relationships were found between patient satisfaction and the doctor talking about the patients’ life expectancy ({chi}2 = 12.28; df = 1; P = 0.00046) and the doctor talking about how the cancer would affect other aspects of life ({chi}2 = 6.36; df = 1; P = 0.012). Higher satisfaction was also associated with individually matching the topics discussed to the needs of patients: specifically, satisfying the need to discuss life expectancy ({chi}2 = 10.43; df = 1; P = 0.0012) and how cancer would affect other aspects of life ({chi}2 = 9.49; df = 1; P = 0.0021).

Anxiety. Lower levels of anxiety were observed amongst patients who reported having their need for life expectancy discussions satisfied (F = 5.29; df = 1, 106; P = 0.023) and being able to discuss their questions the same day as the diagnosis (F = 3.03; df = 2, 104; P = 0.05).

Depression. Patients reporting that the doctor talked about their life expectancy were more likely to express lower levels of depression than other patients (F = 5.41; df = 1, 107; P = 0.022). Whether the doctor talked about the severity of the cancer was not significantly related to depression (F = 3.07; df = 1, 107; P = 0.082) over time; however, there was a significant interaction effect between this factor and time (F = 3.86; df = 2, 214; P = 0.023). Simple contrasts revealed that compared with others, patients who reported that the severity of the cancer had been discussed, expressed lower levels of depression at baseline; however, there was no difference between the two groups at later time periods.

Treatment
Satisfaction. Table 3 presents the results for the relationship between treatment communication factors and patients’ satisfaction and psychological morbidity. Patients reported higher satisfaction with treatment discussions if they perceived that the amount of information that they had been given by their doctor was ‘everything’ ({chi}2 = 13.05; df = 1; P = 0.0003) or if they had been given as much information as they had wanted whether that be ‘everything’ or less than that ({chi}2 = 15.16; df = 1; P = 0.0001). Higher satisfaction was also associated with agreeing that the information about treatment had been presented clearly ({chi}2 = 9.81; df = 1; P = 0.0017) and the patient believing that they had been presented with all available treatment options ({chi}2 = 39.50; df = 1; P = 0.00001).

Anxiety. None of the communication factors measured relating to treatment were significantly related to anxiety.

Depression. Patients who perceived that they had had a major say in the treatment plan reported lower levels of depression than those who did not (F = 6.27; df = 1, 107; P = 0.014).


    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Satisfaction with communication
In general, the patients in this study were satisfied with the communication they had with their clinicians. Satisfaction was highest for discussions of treatment, lower for discussions of diagnosis and lowest for discussions of prognosis. The comparatively lower levels of satisfaction with discussions of prognosis is not surprising as speculating about how the future will unfold, however informed, is inherently uncertain. This is particularly true for melanoma, which is notoriously unpredictable in an individual patient. While it is important that the doctor explores whether the patient wishes to discuss prognosis, as most patients say they do, few rate it as one of the most important communication issues [6]. High satisfaction with treatment discussions may reflect the fact that these discussions involve planning a constructive path forward, providing hope for the patient and returning the doctor to familiar territory.

Communication strategies that optimise patients’ satisfaction and psychological health
Different communication strategies were found to be important in this study, depending upon which outcome criteria were used, highlighting the necessity of considering a range of patient outcomes. Previous work has shown that most of the communication features that have been recommended by experts were also endorsed by patients as important [2]. Many of these patient-preferred communication features were linked with higher satisfaction levels, indicating that patient preferences are a good guide to what influences patients’ satisfaction. This lends support to the trend for more patient-centred consultations where the clinician elicits and appropriately responds to the patient’s preferences, wishes and concerns. Frequently, the communication strategies that were associated with higher satisfaction were also linked with lower psychological morbidity, suggesting that patients can recognise practices that will facilitate their adjustment.

A few communication procedures that are recommended by many experts were not linked to psychological morbidity or satisfaction in this study. They were that a familiar doctor should tell the diagnosis face-to-face, not over the telephone, in a private room. However, given that these recommendations are endorsed by patients as important features of communication in a consultation [1, 2, 3, 7], they should be maintained as part of best practice. The familiarity of the doctor who gave the diagnosis was not actually measured in this study. In the majority of cases, the person giving the diagnosis was a general practitioner who was familiar to the patient, after which the patient was referred to a specialist for further treatment. There were insufficient cases of another health professional, such as a nurse, being present during the diagnosis disclosure to test whether this influenced patient outcome.

Diagnosis
The results suggest that several recommendations made in the literature [1, 7, 11] may be important considerations in disclosing the diagnosis in a way which will optimise patient satisfaction and psychological well-being. These strategies are intended to let patients know what is happening, to give them a sense of control and to reassure them. The doctor should prepare the patient for the possibility of a cancer diagnosis. If an appointment has been made to disclose the diagnosis, the patient should be asked whether they would like to bring a relative. The word ‘cancer’ should be used, as it appears to be associated with less initial depression. The doctor should clearly present the patient with as much information as they want, including written information and give them the opportunity to ask questions. If the general practitioner has insufficient knowledge or resources, the specialist should perform this role. Discussing the patient’s feelings and appropriately responding to them with realistic reassurance also appears to be important in lowering psychological morbidity. Perceived delay in receiving the diagnosis was linked to lower satisfaction but not psychological morbidity. Indeed, it may be desirable to accept short delays and wait for favorable circumstances, such as a consultation with relatives present, to deliver a cancer diagnosis rather than tell immediately in poor circumstances [7].

Prognosis and implications of the diagnosis
Communicating the seriousness of the patient’s cancer, their life expectancy and how cancer might affect their lifestyle were all linked with less depression. It may be that because people fear the unknown, having an open discussion about prognosis alleviates depressive feelings associated with an uncertain future. In any case, this finding provides some reassurance to those clinicians anxious not to cause harm by disclosing ‘too much’. However, it is also possible that less depressed individuals are also more likely to be given more information about prognosis by doctors who judge them better able to cope with this information.

Individualising disclosure about life expectancy according to the patient’s preferences was associated with lower anxiety but not with lower depression. The delicate negotiation required to ensure tailored information provision while maintaining hope is likely to be most difficult in the area of prognosis because of the difficulties inherent in forecasting the future and because it can lead to considerations of mortality.

Treatment
As in earlier studies with breast cancer patients [18], melanoma patients who felt that they had a major say in their treatment plan reported less depression. Satisfaction about treatment discussions was associated with clearly presented and tailored information about treatment, including discussion of treatment options. Presenting all of the treatment options related primarily to those patients with lymph node involvement, where real options do exist. However, earlier work by Fallowfield et al. [18] with breast cancer patients has shown that it is not having real treatment options that promotes psychological well-being, but rather having a doctor who fosters choice and involves the patient in determining the treatment plan.

Strengths and limitations of this study
This study is one of the first to explore naturalistically in a large sample whether doctors’ communication practices produce detectable patient outcomes. The study targeted specific communication practices recommended in consensus guidelines. Short- and long-term outcomes were explored. In order to reduce variability, only one patient sample (melanoma patients) was included. This allowed inclusion of both males and females, and provided an important addition to the data provided by many other communication studies based only or primarily on breast cancer samples. Interestingly, in this melanoma sample, many of the findings reported in earlier studies were replicated [8, 16, 18].

However, there are some important limitations to this study. First, while the list of communication practices studied in this paper was extensive, it was not exhaustive. Secondly, a patient’s experiences and recall of the consultation may not accurately reflect what actually occurred. However, there is evidence to suggest that emotionally charged events are like flashbulb memories and are recalled more accurately than other events [20]. In another study, our group compared patient report with actual doctor behaviour and found a high correlation between the two [22]. Third, the imputation of causation is always an issue in the analysis of longitudinal matrices of association. For instance, it is plausible that patients who were initially anxious were less likely to perceive their doctor as reassuring than patients who were not initially anxious. Or other factors, such as physical health, may influence both patient recall and reported satisfaction. However, there were some objective features of the consultation, unlikely to be affected by poor recall or anxiety, such as who was present to hear the diagnosis and the provision of written information, that were related to patient outcome measures.

Future directions
Future work could extend these results by investigating the association between psychological morbidity and the more subjective communication features by tape-recording the consultation and independently rating the occurrence and effectiveness of these communication features. Alternatively, communication strategies linked to patient outcomes in this research could be included in a training program for clinicians and evaluated in a randomised controlled trial. Given the expense of transcribing and coding tape-recorded consultations and conducting randomised controlled trials, this study was an important first step to generate hypotheses.

Conclusions
Breaking bad news is recognised as a difficult and stressful task. The present study has provided preliminary evidence that implementing recommended communication strategies may optimise patients’ satisfaction with communication and their psychological well-being, both in the short and the longer term.


    Acknowledgements
 
This research was funded by the University of Sydney Cancer Research Fund. Dr Penelope Schofield is a National Health and Medical Research Council Post-doctoral Research Fellow.


    Footnotes
 
+ Correspondence to: Dr P. E. Schofield, Department of Haematology & Medical Oncology, Peter MacCallum Cancer Institute, Locked bag 1, A’Beckett Street, Victoria 8006, Australia. Tel: +61-3-9656-3560; Fax: +61-3-9656-1250; E-mail: penelope.schofield{at}petermac.org Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Girgis A, Sanson-Fisher RW. Breaking bad news: consensus guidelines for medical practitioners. J Clin Oncol 1995; 13: 2449–2456.[Abstract]

2. Schofield PE, Beeney LJ, Thompson JF et al. Hearing the bad news: the Australian melanoma patient’s perspective. Ann Oncol 2001; 12: 365–371.[Abstract]

3. Peteet JR, Abrams HE, Ross DM, Stearns NM. Presenting a diagnosis of cancer: patient’s views. J Fam Pract 1991; 32: 577–581.[ISI][Medline]

4. Sardell AN, Trierweiler SJ. Disclosing the cancer diagnosis. Procedures that influence patient hopefulness. Cancer 1993; 72: 3355–3365.[ISI][Medline]

5. Maguire P. Breaking bad news. Eur J Surg Ca 1998; 24: 188–199.

6. Baile WF, Buckman R, Lenzi R et al. SPIKES—A six-step protocol for delivering bad news: application to the patient with cancer. Oncologist 2000; 5: 302–311.[Abstract/Free Full Text]

7. Ptacek JT, Eberhardt TL. The patient–physician relationship. Breaking bad news: a review of the literature. JAMA 1996; 276: 496–492.[Abstract]

8. Butow PN, Kazemi JN, Beeney LJ et al. When the diagnosis is cancer. Patient communication experiences and preferences. Cancer 1996; 77: 2630–2637.

9. Ellis PM, Tattersall MHN. How should doctors communicate the diagnosis of cancer to patients? Ann Med 1999; 31: 336–341.[ISI][Medline]

10. Lind SE, DelVecchio Good M-J, Seidel S et al. Telling the diagnosis of cancer. J Clin Oncol 1989; 7: 583–589.[Abstract]

11. Fallowfield L. Giving sad and bad news. Lancet 1993; 341: 476–478.[CrossRef][ISI][Medline]

12. Ptacek JT, Ptacek JJ, Ellison NM. "I’m sorry to tell you . . . " Physicians reports of breaking bad news. J Behav Med 2001; 24: 205–217.[CrossRef]

13. Ptacek JT, Ellison NM. Health care providers’ perspectives on breaking bad news to patients. Crit Care Nurs Q 2000; 23: 51–59.[Medline]

14. Ong LML, de Haes JCJM, Hoos AM, Lammes FB. Doctor–patient communication: a review of the literature. Soc Sci Med 1995; 40: 903–918.[CrossRef][ISI][Medline]

15. Ptacek JT, Ptacek JJ. Patients’ perceptions of receiving bad news about cancer. J Clin Oncol 2001; 19: 4160–4164.[Abstract/Free Full Text]

16. Roberts CS, Cox CE, Reintgen DS et al. Influence of physician communication on newly diagnosed breast patients’ psychologic adjustment and decision-making. Cancer 1994; 74: 336–341.[ISI][Medline]

17. Rutter DR, Iconomou G, Quine L. Doctor–patient communication and outcome in cancer patients: an intervention. Psych Health 1996; 12: 57–71.

18. Fallowfield LJ, Hall A, Maguire GP, Baum M. Psychological outcomes of different treatment policies in women with early breast cancer outside a clinical trial. BMJ 1990; 301: 575–580.[ISI][Medline]

19. Simes RJ, Tattersall MHN, Coates A et al. Randomised comparison of procedures for obtaining informed consent in clinical trials of treatment for cancer. BMJ 1986; 293: 1065–1068.[ISI][Medline]

20. Brown R, Kulik J. Flashbulb memories. In Neisser V (ed.): Memory Observed. San Francisco: Freeman 1982; 23–40.

21. Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand 1983; 67: 361–370.[ISI][Medline]

22. Gattelari M, Butow P, Tattersall MHN. Sharing decisions in cancer care. Soc Sci Med 2001; 51: 1865–1878.[CrossRef]