MS Roberts, a 54-year-old African American woman with cirrhosis due tohepatitis B, presented with a new right upper quadrant abdominal pain. Acomputed tomographic scan showed a poorly defined mass with indistinct borderslocated near the portal vein. The likely diagnosis was hepatocellularcarcinoma, probably unresectable due to proximity to the portal vein. Agastroenterologist tells her that the mass may represent cancer but that aliver biopsy is needed to establish the diagnosis. She agrees to the biopsy.Her physician reads the biopsy report while Ms Roberts is sitting in a clinicexamining room.
What is the best way for the physican to handle the disclosure of this newswith Ms Roberts? In this article, we review the empiric research that canguide physicians in communicating bad news.
METHODS
We performed a MEDLINE search using the index terms communicationand bad news and screened the resulting 633 citations for relevance.We targeted citations based on empiric research. Additional literature waselicited from two excellent reviews.1,2
How do patients and physicians experience the delivery of badnews?
A useful definition of bad news is that it “results in a cognitive,behavioral, or emotional deficit in the person receiving the news thatpersists for some time after the news is received.”1 Thus,the determination of what news is bad constitutes a subjective judgment in themind of the receiver, so when physicians assume they are delivering bad news,they may influence patients' responses.
Patients report a variety of emotional reactions to hearing bad news. In astudy of patients who were diagnosed as having cancer, the most frequentresponses were shock (54%), fright (46%), acceptance (40%), sadness (24%), and“not worried” (15%).3 In addition, patient confusioncan be an important contributor to distress commonly seen after a bad newsdiscussion. The biggest source of patient misunderstanding is technicallanguage. For example, in a study of 100 women with a diagnosis of breastcancer, there was substantial misunderstanding of prognostic and survivalinformation, with 73% not understanding the term “median” survivalwhen it was used by their physician.4 Furthermore, they did notagree on the numeric equivalent of a “good” chance ofsurvival.
Physicians are inaccurate at detecting patient distress during bad newsencounters, and this may worsen patients' experiences. In an intensivequalitative study of five oncologists, only one of the five was able toreliably assess patient distress resulting from bad news. In other words,physicians' ability to accurately assess anxiety or depression related to abad news consultation was no better than chance.5 These findingscontrasted with the physicians' self-assessment of their own performance: theyrated their performance favorably and were highly satisfied withit.6
Many physicians experience intense emotions of their own when theycommunicate bad news to a patient. Ptacek and Eberhardt proposed a model ofthe stress associated with bad news that relates the physician's experience tothat of the patient (Figure).1 This model describes the physician'santicipatory stress before delivering bad news and suggests that physicians'stress peaks during the clinical encounter, whereas the patients' stress peakssome time afterward. This stress model can help physicians anticipate thechallenges involved in communicating bad news, and some aspects of it havebeen empirically verified. In a large survey of oncologists, 20% reportedanxiety and strong emotions when they had to tell a patient that her conditionwould lead to death.7 In a more detailed study of 73 physicians, 31(42%) indicated that, while the stress often peaks during the encounter, thestress from a bad news encounter can last for hours to 3 or more daysafterwards.8
How competent are physicians at communicating bad news?
When asked to rate overall physician performance, patients are generallypositive, but they also report that their needs and perferences are not alwaysmet during bad news discussions. Among 148 patients with breast cancer ormelanoma, about 60% reported that their physicians' communication about badnews was “excellent” or “good,” but 22% also reportedthat their physicians seemed nervous or uncomfortable.3 Gapsbetween patients' needs and physician performance are particularly apparentwhen patients are asked whether physicians discussed the implications of thebad news. In a study of cancer disclosure experiences, only 14% of patientsthought that diagnostic disclosure is the most important aspect of a bad newsdiscussion; many patients thought that prognosis (52% of patients) andtreatment (18% of patients) were more important. In the patients with breastcancer or melanoma, 57% wanted to discuss life expectancy, although only 27%of physicians actually did.3 Most of these patients (63%) wanted todiscuss the effects of cancer on other aspects of life, yet only 35% reportedhaving these discussions. In another study, patients reported rarely receivingprognostic information.9
Qualitative studies characterize how physician competence in delivering badnews can fall short. A qualitative study of 79 patients with chronic andterminal illnesses, along with 68 family members and health care workers,found that two important factors made for good communication of bad news. Thefirst was a willingness to talk about dying, and the second was disclosing badnews sensitively.10 Poor delivery of bad news stemmed from beingtoo blunt, discussing bad news at a time and place not appropriate for aserious conversation, and conveying the sense that there was no hope. Also,patients discussed the need for physicians to maintain a balance betweensensitivity and honesty in discussing prognosis.11
Studies that examined the quality of physician competence in discussingdo-not-resuscitate orders and prognosis, issues that may follow bad news,characterize other shortcomings. In an audiotape study of physiciansdiscussing do-not-resuscitate orders with hospitalized patients, physiciansspent 75% of the time talking and missed opportunities to allow patients todiscuss their personal values and goals.12 In a study ofphysicians' communication of prognosis, physicians reported that even ifpatients with cancer requested survival estimates, they would provide a frankestimate only 37% of the time and would provide no estimate, a consciousoverestimate, or a conscious underestimate most of the time (63%).9Taken together, these studies suggest that physician competence atcommunicating bad news is suboptimal.
How should physicians communicate bad news?
Most American patients want to have straightforward, honest discussionswith their physicians.13 They also want their physicians to besensitive in these conversations, and they value hope.11 Some ofthese patients, however, want basic rather than extensiveinformation.14 In addition, patients and physicians identify avariety of barriers to discussing bad news, and individuals differ on theirrelative importance.15 Thus, an approach to communicating bad newsthat encourages physicians to respond to the needs of individual patients maybe more successful than a standardized script, although no comparativeevaluations of bad news protocols have been reported.
Several recommendations on communicating bad news were endorsed by amultidisciplinary panel of experts and also rated as “essential”or “desirable” by more than 70% of 100 patients withcancer.2 Many of these are found in published protocols forcommunicating bad news,16-18 and they are summarized in theTable.
No one way is best to discuss different aspects of prognosis becausepatients differ in how they want to hear the news. For instance, among womentreated for breast cancer, there was no consensus on whether they preferred apositively framed message (eg, 43% preferred discussing “chance ofcure”) or negatively framed message (eg, 33% preferred discussing“chance of relapse”).4 Physicians need to inquirewhether their communication is satisfying the patient's needs and be ready toreframe information.
Table 1.
A step-by-step protocol for communicating badnews*
| Step | Description |
|---|---|
| Prepare for the encounter | If possible, have advance discussion with patient about who will be present |
| Find a location with adequate privacy | |
| Arrange adequate time for discussion | |
| Review the clinical information | |
| Assess the patient's understanding | Introduce everyone present |
| Assess the patient's understanding of the situation | |
| Find out how much patient wants to know | |
| Discuss the news | Provide information honestly and in simple language |
| Tailor amount of medical details and technical language to patient wishes | |
| Respond to the patient's emotions | Encourage patients to express their emotions |
| Acknowledge the patient's emotions and empathize | |
| Tolerate silence | |
| Offer to discuss implications of the news, including | Prognosis |
| Treatment options | |
| Effect on quality of life | |
| Assistance talking to others | |
| Identify support services | |
| Summarize the discussion | Restate important points |
| Ask if there are any other questions | |
| Arrange a follow-up time for patient and family questions and concerns | |
| Document the discussion in the medical record |
From Girgis and Sanson-Fisher,2 Baile et al,16Buckman and Baile,17 and von Gunten et al.18
Recall aids can assist physicians in giving bad news. Audiotapes of thepatient-physician consultation have been shown to improve recall of importantinformation and to reduce anxiety in some patients.19 Theseaudiotapes are typically listened to four to six times after the visit, oftenby family members or friends who were not present. Similarly, writtensummaries have also been shown to improve recall of important information, butpatients tend to prefer audiotapes.19
Does competence in delivering bad news make a difference topatients?
Physician competence in delivering bad news influences patient adjustmentto illness, anxiety, depression, hope, and decision making. In a study of 100patients with breast cancer surveyed 6 months after surgery, adjustment toillness correlated with physician behavior during the cancer diagnosticinterview and with the patient's history of psychiatric problems and premorbidlife stressors.20 Interestingly, the study's findings indicate thatthe physician's caring attitude was more important than the informationprovided during the clinical encounter. In another study, patients whoperceived that the provision of information was handled poorly during aninitial cancer consultation were twice as likely to be depressed or anxiousthan patients who were satisfied.21 Patients who have concerns thathave not been addressed are also more likely to be depressed.22
Bad news discussions also influence patient hope. In a descriptive study,56 patients recently diagnosed as having cancer reported that physicianscontributed to their hope in a variety of ways and that giving information ina sensitive way increased hope.23 However, because of concern aboutdamaging hope, both patients and physicians may collude to avoid talking aboutdifficult information.24 Similarly, in a study of patients withadvanced AIDS, physicians reported that fear of destroying a patient's hope isone of the most common and important barriers to discussing end-of-lifecare.15
The link between the communication of bad news and patients' subsequenttreatment decisions is not entirely clear. However, in a study of patientswith cancer who were seriously ill, those who unrealistically overestimatedtheir survival were more likely to choose life-prolonging therapy and to diein the hospital after attempted cardiopulmonary resuscitation or mechanicalventilation.25 This study emphasizes the effect of inaccuratepatient understanding and suggests that improved communication about bad newsmay influence patients' choices about life-sustaining treatments.
How do cultural differences influence communication of bad news?
Patients of different ethnic backgrounds vary in their preferences abouthow to hear about bad news such as a cancer diagnosis. In a study involvingEuropean, African, Mexican, and Korean Americans, Blackhall and colleaguesdemonstrated a wide variation in patients' willingness to discuss a diagnosisof metastatic cancer openly.26 Many of these families address theissue indirectly by focusing on practical logistics.27 Patientsfrom cultures different from those of their physicians may have worseexperiences with the delivery of bad news. In one study, nonwhite patients whohad advanced AIDS rated the quality of patient-physician communication aboutend-of-life care lower than white patients with advanced AIDS.28 Itmay be particularly important for physicians to openly address cross-culturaldifferences in patients' preferences about the delivery of badnews.
Table 2.
| Exploring cultural beliefs in discussing badnews* |
|---|
|
Modified from Kleinman et al29
In some cultures, even articulating bad news may be associated with adverseconsequences. In a qualitative study of Navajos, Carrese and Rhodes describehow the Navajo concept of hozho (“harmony”) influencescommunication; patients and providers should think and speak in a positive wayand avoid thinking or speaking in a negative way, which could constitute adangerous violation of values.29 This view may be more widespreadthan many realize. In a study of patients with advanced AIDS, Curtis andassociates showed that African Americans with AIDS were more likely than whitepatients with AIDS to believe that discussing death could bring deathcloser.15
These findings indicate that physicians must be alert for situations inwhich their cultural beliefs and values may differ from those of theirpatients. In situations where cultural beliefs may differ widely, thequestions in the Box can be used to develop a commonunderstanding.30
The physician began by asking Ms Roberts how she was doing and whatthoughts she had about the needle biopsy. Ms Roberts appeared nervous, so thephysician addressed her anxiety by asking what she was most worried about. Thepatient expressed concern that a diagnosis of cancer would seriously affecther sister and husband. The physician then asked Ms Roberts if she was readyto go on, and after hearing “yes,” told her that the needle biopsyconfirmed the diagnosis of cancer of the liver. Ms Roberts was upset andtearful, but after a short time of quiet emotional support, the physician wenton to discuss the next steps: a consultation with an oncologist, treatment ofpain, treatment of anxiety, help talking with Ms Roberts' husband and sister,and consultation with a social worker about applying for disability.
CONCLUSION
Communicating bad news is a fundamental physician skill. Physicians shouldbe aware that their own sense of what constitutes a good encounter may differfrom that of many patients, especially when cultural backgrounds differ. Theseconversations, when handled well, can help patients feel informed and hopefuland physicians feel affirmed in their commitment to care for patients.
Figure 1.
Stress experienced by physician and patient in the discussion of bad news(from Ptacek and Eberhardt1).
Figure 2.
Most patients want honest informative discussions about their prognosis
Competing interests: None declared
- Patients report a wide variety of reactions to bad news
- Physicians are unable to accurately detect patient distress in bad newsencounters
- Many physicians experience intense emotions when they communicate badnews
- Patients desire a balance of sensitivity and honesty when receiving badnews
- Physicians should be alert to cultural preferences that limitdisclosure