“Burnout”, the end result of repeated exposure to stress, is characterised by apathy, depression, health problems and the eventual inability to work.1 Cancer specialists are at particular risk of burnout because of the nature of their work, with 28%–56% reportedly suffering at least one episode of burnout in their career.1,2 Communicating with patients has been identified as a significant source of stress for these doctors,2,3 particularly when it involves breaking bad news4,5 and if they feel inadequately trained for the task.2 With cancer specialists conducting an average of 35 bad-news discussions per month,5 the accumulated effect of these events can have an impact on the doctor’s health and the quality of care provided to patients.6
Baile and colleagues5 surveyed 167 cancer specialists attending the 1999 Annual Meeting of the American Society of Clinical Oncology. The most difficult communication tasks reported were those dealing with a lack of curative treatment options for a patient, euthanasia, cancer recurrence, treatment failure, and do-not-resuscitate (DNR) orders. Doctors reported feeling unable to deal with emotional reactions to bad news5 and having a fear of eroding all hope in the patient.7
Practice-related stressors place further strain on doctors and their interactions with patients. In a study of Australian general practitioners, time constraints and phone interruptions during patient consultations were rated among the top three most frequent stressors.8 Cancer specialists have reported that time constraints are a major barrier to providing adequate information to patients9 and to involving them in treatment decisions.10
Our survey was part of a larger study on stress and burnout among members of the Clinical Oncological Society of Australia (COSA).11 In May 2007, members of COSA (n = 1322) were informed about an upcoming survey on stress and burnout. Of these, 1157 agreed to receive further information. After receiving the invitation, 165 declined further contact and nine reported that they no longer worked in cancer care. Of the 740 respondents to the larger survey, 151 identified themselves as oncologists or palliative care specialists (the remainder were nurses, psychologists, social workers and members of other non-medical specialties). The 151 cancer specialists were asked to complete an Internet-based “consultation stress questionnaire”.
The consultation stress questionnaire was a modified version of that used by Baile et al.5 A shorter questionnaire was piloted among eight COSA members. The final questionnaire contained 11 items measuring the difficulty experienced when conducting various consultation tasks and eight items measuring stress experienced in relation to various practice factors. Responses were reported on a four-point Likert scale (“not at all”, “somewhat”, “quite a bit” or “a lot”). Participants also provided demographic information (age, sex, education level) and practice information (cancer specialty, number of years in current position, geographical region of work).
Participants received the survey invitation either via email or by post (if requested). Invitations provided a web address for accessing the questionnaire and personal log-in details. Three reminders were sent to non-responders 2, 3 and 6 weeks after the initial invitation. For participants who received the postal invitation, the second reminder also included a printed copy of the survey.
Responses were cross-tabulated for each item across the four response categories. Subsequently, differences in age (median, 49 years) and number of years in current position (median, 17 years) were examined by dividing the sample into two groups using median splits and using χ2 tests. To ensure adequate numbers in each cell for between-group analyses, the response categories “somewhat”, “quite a bit” and “a lot” were subsumed into one category (“stress”) and compared with “not at all” (“no stress”).
The questionnaire and protocol were approved by the University of Sydney and University of Newcastle human research ethics committees.
Of the 151 eligible participants identified, 134 completed the questionnaire (89% response rate). Of the 17 non-respondents, 16 were surgical oncologists and one was a radiation oncologist. Based on figures from the Royal Australasian College of Physicians and the Royal Australasian College of Surgeons, our sample represented about 20% of practising oncologists and 4% of palliative care specialists in Australia.
Demographic characteristics of the sample were as follows: 70% male; 91% working in metropolitan areas; 63% medical oncologists, 20% radiation oncologists, 11% surgical oncologists and 6% palliative care specialists; mean age, 49.1 years (SD, 10.1 years; range, 31–76 years); and mean length of experience in current position, 17.6 years (SD, 10.6 years; range, 1–40 years).
Self-reported levels of difficulty in conducting particular consultation tasks are summarised in Box 1. Discussing high-cost drugs with patients who doctors believe cannot afford them caused the greatest difficulty, with 78% of doctors reporting at least some difficulty. Other difficult discussions were about treatment failure (67%), DNR orders (63%), transition from curative to supportive care (62%), and cancer recurrence (62%).
Self-reported levels of stress experienced during different practice situations are presented in Box 2. Almost all doctors reported being at least somewhat stressed when there was a long line of patients waiting to see them (97%) and when they had inadequate patient information (93%). Other sources of stress included interruptions during consultations (86%), seeing patients when doctors were unwell themselves (85%), and managing office staff (79%).
More male doctors than female doctors had difficulty responding to patients’ emotions (65% v 46%; χ2 = 4.1; P < 0.05). More female than male doctors had difficulty discussing complementary therapies (59% v 37%; χ2 = 5.6; P < 0.05) and found the following situations stressful: having inadequate patient information (100% v 90%; χ2 = 3.9; P < 0.05), interruptions (97% v 82%; χ2 = 5.5; P < 0.05), seeing patients when they were unwell themselves (97% v 81%; χ2 = 6.1; P < 0.01), and disagreeing with colleagues in multidisciplinary meetings (76% v 59%; χ2 = 3.5; P < 0.05).
Younger doctors (aged < 49 years) had greater difficulty discussing complementary therapies than older doctors (57% v 28%; χ2 = 11.5; P < 0.01) and were more stressed by consultation interruptions (93% v 80%; χ2 = 4.7; P < 0.05).
Compared with more experienced colleagues, doctors who had been less than 17 years in their current position were more likely to report difficulty in discussing high-cost drugs (85% v 69%; χ2 = 4.4; P < 0.05) or complementary therapies (55% v 29%; χ2 = 9.0; P < 0.01), and in seeing patients when they themselves had problems at home (67% v 44%; χ2 = 6.7; P < 0.01). They also reported being more stressed by consultation interruptions (93% v 78%; χ2 = 6.4; P < 0.05) and disagreements in multidisciplinary meetings (72% v 54%; χ2 = 4.6; P < 0.05).
Our study revealed that most Australian cancer specialists have particular difficulty discussing high-cost drugs with patients they believe cannot afford them. The second most difficult topic was treatment failure. Practice-related factors, affecting interactions with patients, also caused stress in over 60% of our respondents.
Although anti-cancer drugs are the most rapidly growing group of drugs within the Pharmaceutical Benefits Scheme (PBS),12 a large number of effective drugs are listed only for particular indications. Our findings echo those of a survey by Thomson and colleagues of 184 Australian medical oncologists, which revealed that only a third of oncologists would discuss a new drug with a patient if the drug was not PBS-subsidised.13 However, doctors have an ethical, and most likely legal,14 duty to inform patients about any clinically relevant treatment options and side effects.13 These findings suggest that cancer specialists would benefit from guidelines on how to discuss unfunded high-cost drugs. We support the view of Thomson and colleagues that such discussions could follow the initial presentation of standard options,13 and add that doctors should outline differences between treatment options in costs versus the likelihood of benefits.
The fact that the doctors in our survey reported the least difficulty with disclosing cancer diagnosis or being honest about a patient’s prognosis is an improvement on the findings of a decade ago, suggesting that training in these topics has reduced discomfort.5,15 However, given that two-thirds reported difficulty with topics surrounding treatment failure, doctors may need targeted communication skills training on moving the emphasis to symptom control, exploring realistic goals and other forms of “hope”, and providing psychosocial support.16 Similarly, discussions about DNR orders could be an opportunity to identify and reaffirm the patient’s values about hoping for a peaceful death.17 We recommend that cancer specialists read the recently released guidelines on how to best discuss these and other end-of-life issues.18
Discussing complementary therapies caused particular difficulty for female doctors, younger doctors and less experienced doctors in our survey. This contrasts with a study that showed female and younger doctors had a more favourable attitude towards complementary therapies.19 Our results may reflect a lack of confidence in such discussions due to gaps in knowledge.20 Currently, complementary medicine is not acknowledged in the Australian medical curriculum.21
While some stressors cannot easily be ameliorated in busy cancer clinics, the incidence of others can be reduced by small organisational changes. For example, long appointment lists that do not allow sufficient time for each patient to be seen should be curtailed. Is it reasonable that newly referred cancer patients be seen within a 15-minute consultation? It may be helpful to inform patients before their appointment of how much time they have been allocated.22 Communication aids that allow cancer patients to identify questions to ask their doctor before their appointment may also reduce consultation times.23 Prohibiting phone and staff interruptions, which reportedly consume 4% of oncology consultations,24 is a relatively simple strategy.
Cancer care often involves meetings of a multidisciplinary team, making maintenance of complete patient records and sharing of information difficult. Having electronically accessible records can minimise problems of inadequate patient information25 and help reduce phone and staff interruptions.
However, the attitudes and habits of staff at all levels can be barriers to changing practices. Education of, and consultation with, medical, management and administration staff about specific stressors, together with a policy of promoting a stress-free environment championed by senior staff, will aid in the successful adoption of changes.26
Although ameliorating the difficulties experienced by cancer specialists in dealing with patients is not simple, our findings highlight areas to target in evidence-based training. Small organisational changes, such as reducing phone and staff interruptions during consultations and informing patients how much time is allocated for their consultation, may have immediate effects on reducing doctors’ stress levels.
1 Level of difficulty reported by cancer specialists for various consultation tasks
2 Level of stress reported by cancer specialists for various practice-related situations
Abstract
Objective: To determine what aspects of communicating and consulting with cancer patients are viewed as difficult and stressful by cancer specialists in Australia.
Design, participants and setting: Anonymous, cross-sectional, Internet-based survey completed by 134 cancer specialists between June and August 2007. Participants, who were all members of the Clinical Oncological Society of Australia, included oncologists and palliative care specialists.
Main outcome measures: Degree of difficulty perceived for various consultation tasks; level of stress reported during various practice-related situations.
Results: Doctors had the most difficulty discussing high-cost drugs with patients they knew could not afford them, followed by topics relating to treatment failure. They had the least difficulty telling patients they had cancer or being honest about prognosis. The most stressful practice situations included having incomplete patient information to conduct the consultation and having a long line of patients waiting for a consultation. At least 62% of respondents reported experiencing some degree of stress in all the practice situations presented. There were differences in difficulty and stress experienced as a function of the doctor’s sex, age and clinical experience.
Conclusions: Targeted, evidence-based guidelines and communication courses are required to better equip cancer specialists for providing non-directive advice about unsubsidised high-cost drugs and for offering different forms of hope in the context of treatment failure. Implementing small organisational changes — such as reducing interruptions during consultations and informing patients of the duration of their allocated consultation — may also help reduce stressful practice situations.