S P E C I A L T H E M E A R T I C L E

Teaching Medical Students and Residents Skills for

Delivering Bad News: A Review of Strategies
Marcy E. Rosenbaum, PhD, Kristi J. Ferguson, PhD, and Jeffrey G. Lobas, MD
ABSTRACT
Although delivering bad news is something that occurs
daily in most medical practices, the majority of clinicians
have not received formal training in this essential and
important communication task. A variety of models are
currently being used in medical education to teach skills
for delivering bad news. The goals of this article are (1) to
describe these available models, including their advan-
tages and disadvantages and evaluations of their effective-
ness; and (2) to serve as a guide to medical educators who
are initiating or rening curriculum for medical students
and residents. Based on a review of the literature and the
authors own experiences, they conclude that curricular
efforts to teach these skills should include multiple ses-
sions and opportunities for demonstration, reection, dis-
cussion, practice, and feedback.
Acad Med. 2004;79:107117.
D
elivering bad news, a task that occurs in any
medical practice, can be daunting for the clini-
cian. Although it is most often thought of as
communicating about life-threatening illness, the
imminence of death, or communicating about the death of a
loved one to a family member, Bor et al.
1
provide a useful and
more inclusive denition of bad news: . . . situations where
there is either a feeling of no hope, a threat to a persons
mental or physical well-being, a risk of upsetting an estab-
lished lifestyle, or where a message is given which conveys to
an individual fewer choices in his or her life. Given this
denition, delivering bad news is something that occurs daily
for most practicing clinicians.
How bad news is delivered can have a signicant impact
on patients perspective of illness, their long-term relation-
ships with clinicians, and both patient and provider satisfac-
tion.
25
Several authors have reported that patients had
signicantly more distressing feelings toward clinicians they
felt delivered the news in an inappropriate manner.
25
Practicing physicians and residents have been shown to
lack both condence and skill in performing this basic
clinical task.
69
A number of factors can contribute to this
discomfort, such as feeling responsible for patients misfor-
tune, perceptions of failure, unresolved feelings about death
and dying, concerns about patients responses to the news,
and clinicians concerns about their own emotional responses
to the circumstances.
7
Another contribution to low condence and discomfort in
this task is that the majority of practicing physicians have
reported having received no formal training in effectively
communicating bad news.
6,10,11
Thus, until recently, most
practitioners learned to give patients bad news through trial
and error and perhaps by observing role models during their
training. Because negative role models for giving bad news
are common,
6
relying on experience and role-modeling may
result in communication patterns that do not meet patient
needs rather than in effective approaches to this task. There-
fore, teaching the skills for delivering bad news increases the
likelihood that physicians will learn how to deliver bad news
effectively.
Much has been written about the skills necessary for
effective delivery of difcult news, including extensive re-
views of the literature and creation of consensus guidelines
for this practice.
2,3,7,12,13
In the literature specically focus-
ing on educational interventions, several useful content
models have been developed and implemented in both un-
Dr. Rosenbaum is assistant professor of family medicine and Dr. Lobas is
professor of pediatrics, Roy J. and Lucille A. Carver College of Medicine; Dr.
Ferguson is associate professor of community and behavioral health, College
of Public Health. All are at the University of Iowa, Iowa City.
Correspondence should be addressed to Dr. Rosenbaum, 1204 MEB, Uni-
versity of Iowa College of Medicine, Iowa City, IA 52245; e-mail:
marcy-rosenbaum@uiowa.edu.
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 107
dergraduate and graduate settings. For example, the SPIKES
model (setting, perception, invitation, knowledge, empathy,
summary and strategy) developed by Buckman
7
for delivering
difcult news is used in many medical schools (see Table 1).
In this article, we review published reports (based on
Medline searches) of strategies that have been used to teach
effective delivery of bad news to medical students and resi-
dents.* We describe available models and offer our opinions
based on our experiences and on our review of the broader
medical education literature on the advantages and disad-
vantages of each strategy (see Table 2). We also discuss
ndings from evaluations of these models. This article pro-
vides a guide to medical educators who are initiating or
rening curriculum for medical students and residents to
learn this essential and important communication task.
Based on our review of the literature, we conclude that,
optimally, any curriculum should include a model for effec-
tive delivery of bad news (e.g., SPIKES), and opportunities
for learners to discuss relevant issues, and practice and
receive feedback on their skills. Potential strategies for pro-
viding education in bad-news delivery include lectures,
small-group discussions, role-playing with peers and stan-
dardized patients (SPs), and teaching in the context of
patient care.
STRATEGIES FOR TEACHING SKILLS FOR DELIVERING
BAD NEWS
Didactic Approaches
In a comprehensive review published in 1997, Billings and
Block found that lectures were the most widely used strategy
for teaching end-of-life content in the medical curriculum,
under which bad-news delivery is often addressed.
14
In a lecture on delivering bad news described in one
report, residents learned death notication skills as part of an
advanced cardiac life support course.
15
This lecture focused
on methods of notication, understanding grief and after-
notication issues. We found no other published reports of a
sole reliance on lectures to teach learners about delivering
difcult news. Several studies have discussed using interac-
tive lecture formats to convey basic information and as a
catalyst for discussion and skills practice in subsequent small-
group sessions.
1619
In one example, two faculty provided an
interactive lecture on delivering difcult news.
16
They in-
volved the audience and role-played both poor and effective
encounters, using elements of the model described by Buck-
man.
7
Trigger videotapes, showing dramatized bad-news en-
counters, can also be used in this process.
19
After each
demonstration, the audience was asked to identify effective
and noneffective behaviors on the part of the clinician, based
on the patients communication needs. Then steps in effec-
tive delivery of bad news (see Table 1) were presented in
detail while referring to both case examples.
In an alternate approach, an audience member was asked
to volunteer to give bad news to a SP. For a student audience,
a scenario that required little medical knowledge was pro-
vided. For a more advanced audience (residents and practic-
ing clinicians), volunteers were asked to identify a typical
situation. As this example shows, a spontaneous demonstra-
tion has the advantage of being perceived as more genuine.
20
In addition, learners in the audience can more easily imagine
themselves in the volunteers position and ponder what they
would do in a similar circumstance. The disadvantage is that
some common ineffective or effective behaviors will be left
out if they are not scripted.
Several education programs have used speaker panels to
present information about delivering bad news.
10,17,21
In one
example, parents of children in whom cancer had been
diagnosed described their responses and needs in relation to
* We limited this review to models of bad news education described in
published articles indexed in Medline. Attention should be drawn to the
availability of descriptions of other models for bad news education in the
End of Life/Palliative Education Resource Center (EPERC) Online Data-
base, managed by the Medical College of Wisconsin. This database (www.
eperc.org) provides peer-reviewed descriptions of curricula focused on
delivering bad news and other end-of-life communication skills, and is a
useful resource for persons interested in developing or enhancing education
in these areas.
Table 1
The SPIKES Protocol for Delivering Bad News to Patients*
Step Description of Task
Setting Establish patient rapport by creating an appropriate setting
that provides for privacy, patient comfort, uninterrupted
time, setting at eye level, and inviting signicant other(s)
(if desired).
Perception Elicit the patients perception of his or her problem.
Invitation Obtain the patients invitation to disclose the details of the
medical condition.
Knowledge Provide knowledge and information to the patient. Give
information in small chunks, check for understanding,
and frequently avoid medical jargon.
Empathize Empathize and explore emotions expressed by the patient.
Summary and
strategy
Provide a summary of what you said and negotiate a
strategy for treatment or follow-up.
*From Baile WF, Kudelka AP, Beale EA, et al. Communication skills training in oncology:
description and preliminary outcomes of workshops on breaking bad news and managing
patient reactions to illness. Cancer. 1999;86:88797. Baile et al.s protocol was adapted from
Buckman R. How to Break Bad News: A Guide For Healthcare Professionals. Baltimore: Johns
Hopkins University Press, 1992.
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 108
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
bad news, and they elded questions from the audience.
17
In
another example, a panel of clinicians discussed their ap-
proaches to delivering bad news and described the challenges
they had faced.
10
The main advantage of lectures is that they take minimal
time and faculty resources to deliver content to a wide
audience. However, they allow for only limited assessment of
learner needs, limited discussion of issues, and no chance for
practice and renement of the skills discussed.
Small-Group Discussions
Reported interventions using small-group discussion sessions
for teaching delivery of difcult news have included trigger
tapes, demonstrations, case descriptions, or presession read-
ings to generate discussion (see Table 3).
2024
These tools are
used in a manner similar to didactic approaches but include
opportunity for learners to discuss the issues raised. For
example, during a one-hour case conference in internal
medicine, a student or faculty member was invited to give
bad news to a SP in front of the group as a catalyst for
discussion during the session.
20
In another intervention,
during two-hour sessions with groups of 1618 second-year
medical students, group members discussed their perceptions
of bad-news tasks and challenges, watched two videotapes on
delivering bad news, and then interacted with a handicapped
child and his or her parents, or a patient with cancer. This
approach is particularly innovative in including actual pa-
tients as part of the small group discussion.
22
Although small-group discussions give learners an oppor-
tunity to discuss their concerns more deeply and explore
their reactions, these discussions can require more faculty
time than do lectures to reach the same number of learners,
and there is no opportunity for skills practice and feedback.
Small Groups with Peer Role-Playing
Some small-group interventions include giving learners an
opportunity to practice and receive feedback on their skills
through peer role-play exercises following discussion of basic
bad-news delivery issues.
2,10,17,25,26
Some interventions have
Table 2
Advantages and Disadvantages of Strategies for Teaching Medical Students and Residents Skills for Delivering Bad News
Strategy Advantages Disadvantages
Didactic approaches Presents core concepts to large numbers of learners efciently Little opportunity for discussion
Minimal faculty time and resources No opportunity for practice and feedback
Learners are anonymous
Opportunity for efcient use of skills demonstration and use of
speaker panels can be done efciently
Small-group discussion Opportunity to discuss issues, skills, and concerns No opportunity for practice and feedback
Faculty time intensive
Small-group, peer role-play Opportunity to discuss issues, skills and concerns Variable ability of learners to portray patients
Skills practice with feedback Faculty time intensive
Insight into patient perspective
Small-group standardized patient role-play Multiple scenarios show range of approaches and patient
responses
Peer performance anxiety
Standardized patients and faculty time intensive
Skills practice with feedback from faculty, peers, and
standardized patients
Less realistic than one-to-one standardized
patient encounter
More realistic than peer role-play
One-to-one standardized patient encounters Skills practice with feedback from standardized patients or
faculty
More realistic than group encounters
No group discussion
No exposure to different approaches and patient
responses
Faculty or standardized patient intensive
Teachable moments in clinical settings Actual context of patient care Clinical time restraints
Observation, demonstration, and feedback Patient privacy
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 109
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
Table 3
Summary of Literature on Strategies for Teaching Medical Students and Residents Skills for Delivering Bad News to Patients*
Format Authors Level of Learner Strategies Assessment Measures/Results
Lecture Pollack 1999
15
PGY One-hour lecture on death notication Design: Random assignment to lecture or nonlecture
group and all participated in SP death-notication
encounter
Measures: SP global rating
Results: Lecture group did signicantly better than
nonlecture group
Small-group
discussion
Edinger et al. 1999
20
MS3 Demonstration SP role-play Not available
Romm 2002
21
Obstetrics/
Gynecology
PGY
Panel of parents
Group discussion
Learner satisfaction
Knox et al. 1989
22
MS2 Trigger videos
Discussion with family of disabled child
Demonstration cancer diagnosis role-play
Learner satisfaction three months and 18 months after the
seminar
McNeilly et al.
2001
23
MS3 Presentation of Buckman model
Trigger videos
Application of model to videos
Design: Pre/post knowledge and attitude
Measures: Students asked to name six steps in Buckman
model and if they had a plan for breaking bad news
Results: Signicant improvement in bad news knowledge
and attitude after the seminar
Angelos et al.
1999
24
Surgery PGY Small-group discussion
Limited role-play with prepared cases
Video review
Design: Pre/post condence questionnaire and learner
satisfaction
Measure: Self-assessed condence in explaining bad
news
Results: No signicant difference before and after the
seminar
Small-group
peer role-
play
Vetto et al. 1999
10
MS12 Self-study readings
Clinician panel
Group discussion
Written case-based peer role-play
Design: Comparison of objective structured clinical
examination scores from intervention and
nonintervention groups
Measures: SP-rated knowledge and humanistic skills and
faculty-rated humanistic skills
Results: Intervention group did signicantly better on
humanistic skills with no signicant difference on
knowledge
Morgan et al.
1996
17
Pediatrics PGY1 Didactic
Panel of parents
Learner satisfaction
Peer role-play with resident-generated
cases
Magnani et al.
2002
25
MS2 Clinical incidents
Trigger videos
Learner satisfaction
Written case based peer role-play, three
cases
Reection questions
Ungar et al. 2002
26
PGY2 14 sessions, 90 minutes each Learner satisfaction
Group discussion
Written case-based peer role-play
Video review
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 110
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
Table 3 (Continued)
Format Authors Level of Learner Strategies Assessment Measures/Results
Small-group Rosenbaum et al. MS3 Lecture/demonstration Design: Pre/post questionnaire
SP role-play 2002
16
SP role-play with others watching, ve
cases
Measures: Self-assessed comfort in delivering bad news in
different situations
Results: Signicant increase in comfort after the session
Garg et al. 1997
18
MS3 Video critique Design: Pre/post questionnaire
Small-group exercises
Peer role-play
SP role-play
(Four cases chosen out of 12 possible)
Measures: Self-assessment of whether students had a plan
for approaching delivering bad news and if felt competent
to do so
Results: Signicant increases posttest on both measures
Cushing et al. 1995
27
MS45 Discussion Design: Pre/post questionnaire
Video critique
Peer role-play
SP role-play
Measures: Students were asked to give level of condence
in specic bad-news situations and list as many things a
clinician can do to help recipients when giving bad news
Results: Signicant increases in condence level and longer,
more comprehensive list of steps in effective bad news
Van Winkle et al. 1998
28
MS4 Discussion Learner satisfaction
SP role-play with others watching, three
cases
Tolle et al. 1989
29
PGY1 SP role-play with learners consulting as
team
Learner satisfaction
Group feedback
Kahn et al. 2001
30
MS3 SP role-play with four learners taking
turns on same case
Group feedback
Discussion
Design: Pre/post self-efcacy questionnaire
Measure: Self-assessment of I am comfortable giving bad
news to patients
Results: Signicant increase in self-efcacy
Fortin et al. 2002
31
MS2 Mini-lecture Learner satisfaction
SP role-play with others watching
Feedback
Serwint et al. 2002
32
PGY2 All day seminar Learner satisfaction, use of techniques
Video triggers Self-assessed 18 months after the seminar
SP role-play with others watching, two cases
Lectures
One-on-one
SP
encounters
Coletti et al. 2001
33
MS3 Reading packet
SP with feedback
Design: Comparison between SP encounter and non-SP
encounter groups on end-of-rotation clinical practice
examination bad-news SP station
Measures: SP ratings on numbered item evaluation form
addressing content and communication skills
Results: SP encounter group did signicantly better on the
examination
Greenberg et al. 1999
34
Pediatrics
PGY23,
fellows
SP encounter with feedback
SP encounter without feedback
Design: Comparison across two SP encounters, before and
after feedback.
Measures: I. SP ratings on content checklist, Gibb trust
scale, and National Board of Medical Examiners Patient
Perception Questionnaire. II. Resident self-assessed
comfort level pre/post
Results: Signicant improvement in content categories of
communication and follow-up. Signicant differences in
counseling skills. No signicant differences in Patient
Perception Questionnaire
continued on next page
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 111
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
used preprepared cases in which one learner portrayed the
patient and the other acted as the clinician delivering the
news.
25,26
Preprepared cases can be especially appropriate for
learners who have little actual experience to draw on. In one
example, second-year medical students were introduced to
issues about delivering bad news through clinicians describ-
ing their experiences and then the students critiqued trigger
videotapes. Students then role-played detailed, written bad-
news encounters, and answered a series of questions.
25,26
Some interventions for higher-level learners use learner-
generated cases. These cases can be elicited either before or
during the actual session. For example, one intervention
based a seminar on cases written by the institutions own
residents.
17
Using cases generated during sessions, group
members identied a clinical experience they would like to
reenact (often one they felt did not go as well as they would
have liked).
11,26
This approach allows participants to address
concerns they feel they need to work on, making it especially
relevant for them.
In one approach to learner-generated cases, the learner
provides medical information, patient circumstance, patient
reaction, and clinicians approach to the encounter. Then, a
group member portrays the patient, and the clinicianlearner
delivers the news in a different way than he or she did in the
actual encounter, applying some of the concepts already
discussed in the group. The group provides feedback about
ineffective and effective behaviors demonstrated in the en-
counter and alternative ways to approach the situation.
Alternatively, the clinician whose case is being role-played
takes on the role of the patient, and another group member
takes on the role of the clinician. In this conguration, the
person who generated the case gains insight into both a
different way to approach the case and what the patient may
have been experiencing in this encounter. In both situations,
Table 3 (Continued)
Format Authors Level of Learner Strategies Assessment Measures/Results
Goldschmidt et al.
1987
35
MS4, Family
medicine
PGY1
SP encounter with feedback
SP encounter
Learner satisfaction
Rosenbaum et al.
1996
36
PGY1 SP encounter with faculty feedback Design: Pre/post questionnaire
Measure: Self-assessed rating of condence in giving bad
news
Results: Signicantly less condence after one encounter
Jewett et al. 1982
37
Pediatrics
PGY23
One SP encounter with feedback at
beginning of rotation
Design: Posttest control group design, comparison of
performance before training and after
One SP encounter with feedback at end
of rotation
Measures: SP rating of critical information giving, clarity
of information, and interpersonal skills
Results: Signicant improvement in critical information,
general improvement in clarity, and improvement in
interpersonal skills, especially listening and partnering
with patients
Roth et al. 2002
38
Medicine PGY1 SP encounter with feedback from SP and
observing faculty
Learner satisfaction
Pan et al. 2002
39
MS3 Small-group session using role-play,
reection, and discussion.
Learner satisfaction
SP encounter with feedback from faculty
Clinical
teaching
Muir et al. 1999
40
MS4 Didactic
Rounds
Learner satisfaction
Bedside modeling
Videotaped SP
*MS, medical student; PGY, postgraduate year (resident); SP, standardized patient.
We did not report measure for studies that relied on learner satisfaction. All authors reported high ratings of interventions by learners. Assessment methods are explained only to the extent they were
clearly explained in the original article.
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 112
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
if groups of three are formed then one person can observe and
provide feedback.
In summary, role-playing allows learners to practice their
skills, receive feedback, and gain insight into the patients
perspective; it also generates discussion. Peer role-playing is
less demanding of resources and organizational needs than
role-playing with SPs. Disadvantages to role-playing are
variation in learners abilities to portray patients in a realistic
manner, familiarity among peers, and more faculty time
required than less interactive sessions.
Small Groups Using Standardized Patients
The majority of published educational interventions that
focus on delivering bad news used SPs who can be trained to
portray patient responses to bad news in a realistic and
standardized manner.
16,18,2732
These interventions included
portrayal of multiple scenarios, giving most (if not all)
participants a chance to practice delivering difcult news.
Using multiple scenarios in bad-news role-playing sessions
can provide insight into the common and contrasting patient
responses and skills needed in different situations and also
allows for exposure to different learners approaches to the
task.
16
For example, two reported interventions with medical
students used a combination of preprepared cases and stu-
dent-generated cases in role-plays with SPs during small-
group sessions (four to ten students each).
18,27
Each student
role-played with the SP, and then the group proceeded with
feedback and discussion. Some interventions have used
closed-circuit television with small groups, allowing learners
to watch as individual group members deliver bad news to a
SP in a separate room.
16,28
Closed-circuit observation sys-
tems can provide a more realistic context than can perform-
ing directly in front of the group and the setting allows the
observers to comment as the encounter proceeds. In one
example, each student in a small group of students delivered
difcult news to a different SP while being observed by others
over closed circuit television.
16
In another example, four to
six students observed over closed circuit television as three
other students each took a turn delivering difcult news in a
variety of situations. Each scenario was followed by feedback
and discussion with a faculty facilitator, the students and the
SP. In this conguration, not all participants may practice
delivering the news but they are exposed to multiple ap-
proaches and varying scenarios.
Use of SPs in role-play situations gives learners an oppor-
tunity to practice their skills with skilled and nonfamiliar
patients and receive feedback from peers, SPs, and faculty.
This role-playing, however, requires intense use of both
faculty and SP time and audiovisual support resources if
closed-circuit television is used. In addition, having to per-
form in front of ones peers can be intimidating for some and
creates a less realistic situation than a one-to-one encounter
with SPs.
One-to-One Learning with Standardized Patients
Several schools have reported using one-to-one encounters
between learners and SPs as their primary approach to
teaching delivery of difcult news.
3339
This approach has
most often been used with learners who have already had
patient care experiences. In some approaches, the SP was the
main teacher during this intervention. In one example, in
two SP encounters students learned about delivering bad
news during surgery and obstetrics/gynecology rotations.
33
After reviewing written materials on techniques for deliver-
ing difcult news, each student delivered difcult news (rec-
tal cancer diagnosis or pregnancy loss) to the SP, who
afterward provided feedback to the learner on strengths and
suggestions for improvement. Using this approach minimizes
demand on faculty time but requires more intensive use of SP
time for training and teaching sessions.
Other examples reported using faculty observers and feed-
back in one-to-one simulated sessions. One approach used
two SP encounters of providing a cancer diagnosis to train
residents and fourth-year students during a family practice
rotation.
35
After the second encounter with a SP, faculty
members reviewed the videotape with learners and provided
feedback on skills improvement. One advantage of this
approach is that it provides an opportunity for faculty to
observe learners actually delivering news to a patient, albeit
a standardized one, which is often difcult for faculty during
clinical rotations.
One-to-one SP encounters eliminate the discomfort that
can accompany role-playing in front of groups and can
provide a more clinically realistic encounter. In addition, the
time commitment for both learners and faculty is minimized.
Disadvantages of this approach are that learners do not have
an opportunity to benet from observing multiple approaches
and multiple patient responses to bad news.
Teachable Moments in Delivering Bad News
Although rarely described in the literature, faculty have
ample opportunities to teach and reinforce skills for deliver-
ing bad news in the direct context of clinical care. These
teachable moments can be identied and used in inpatient
ward round and outpatient stafng settings.
6,8,19,4042
Before
a bad-news encounter, faculty members can discuss concerns
and possible approaches to bad-news delivery. They can ask
the learner(s) about their experiences and concerns regard-
ing delivering bad news, and thus assess their learning needs
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 113
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
and levels of comfort with the task. This also encourages
approaching the encounter with a set plan in mind for
delivering the news.
18
In addition, learners can ask questions
and be queried regarding their perceptions of the patients
reactions and needs, as well as the effectiveness of ap-
proaches. Providing self-study resources such as articles and
videotapes can also be useful for enhancing learning and skill
development in these venues.
18
Role modeling and demonstration are important ways
trainees can learn and rene their bad-news delivery skills. In
a curriculum at our institution, we told third-year medical
students that if they found themselves in a position of giving
bad news that they were uncomfortable with at their level of
training, the students could ask a supervisor to deliver the
news while they observed.
16
One of the challenges to bedside
teaching of this skill is to maintain patients privacy during
this emotionally charged encounter while still being able to
teach learners these important skills. Setting aside time
outside of rounds with one or two learners is advised, in order
to deal with the interaction sensitively and effectively while
maximizing the experience for the learners. The patient
should be informed of the reason for the learners(s) pres-
ence in this type of encounter. If bad news must be delivered
during rounds we suggest that the encounter be saved until
the end of rounds to avoid time pressure and also allow the
attending physician to limit the number of learners that
participate. In addition to giving enough time to the patient,
this also can allow time immediately after the encounter to
process with the learner(s).
Observing learners giving difcult news can also allow
faculty to provide feedback to improve skills. However, roles
of the learner(s) and the attending physician need to be
clearly dened before a bad-news encounter. For example,
with a residentphysician who has his or her own patient
pool, the attending can offer to accompany the resident as a
resource if he or she cannot answer patient questions. The
resident can inform the patient of the attendings role as an
observer. The challenge in observing learners with patients is
for the attending physician to resist the temptation to dom-
inate the encounter and have the patient focus on the
attending physician. However, with observation, feedback
on actual performance can be even more effective than
giving feedback regarding SP encounters.
Teaching about bad-news delivery in the context of actual
patient care can open the door to identify ways for learners to
improve and acknowledge the emotional challenges that ac-
company being the bearer of bad news. In addition, faculty can
relay both negative and positive outcomes from different ap-
proaches they have tried. Finally this approach allows for ap-
plication of skills in the context in which they will be used.
Potential disadvantages include time constraints for teaching in
the clinical setting and concerns about patients privacy.
EVALUATION OF TEACHING STRATEGIES
The number of published reports evaluating different strate-
gies for teaching skills for delivering bad news are limited and
the majority limit their assessment of the impact of educa-
tional interventions to learner satisfaction and condence
rather than assessing change in learners actual behavior (see
Table 3). However, the following evidence points to the
advantage of using some strategies over others.
We found only one report specically that evaluated a
lecture for teaching bad-news delivery.
15
In a prospective,
randomized study, 18 of 36 residents received a lecture on
death notication. Members in the lecture group performed
signicantly better in a death-notication encounter with a
standardized survivor. Although this report indicates a lec-
ture can have some immediate impact on learners skills for
delivering bad news, the literature on teaching communica-
tion skills in general and delivering bad news in particular
argues against sole reliance on lectures to teach these behav-
ioral skills.
13,19,4143
Learners must have an opportunity to
practice the skills before they can internalize them. In com-
bination with other methods, however, didactic presentation
of the principles of delivering bad news can provide impor-
tant baseline information for discussion and practice. When
using lectures, we encourage teachers to employ more inter-
active techniques, such as incorporating demonstrations,
role-plays, panels, and audience feedback, as a way to engage
the audience and help with retention of information.
The majority of reports on small-group activities to teach
skills for delivering bad news relied on learner self-assess-
ments of condence before and after the intervention, and
learner evaluation of the usefulness of the educational activ-
ity (see Table 3).
1618,21,24-27,30
Almost all interventions
using these assessments reported signicant changes in
learner self-condence and high ratings of the usefulness of
the training. One study found no signicant differences in
condence in delivering bad news after a small-group discus-
sion session with residents.
24
The authors postulated that this
was due to lack of opportunity for feedback and practice of
these skills within the session. The majority of small-group
studies indicate that learners desire more training opportu-
nities and the opportunity to practice with SPs.
16,22,25,27
Some studies have evaluated the impact of small-group
activities on learners knowledge and attitudes. For example,
in a pre/post study, learners were able to describe the six steps
in Buckmans model for delivering bad news and were more
likely to have a plan for giving bad news following a small-
group intervention in which learners applied the model to
trigger videos.
23
In another pre/post comparison study, learn-
ers could provide longer, more comprehensive lists of steps in
effective bad-news delivery after small-group, SP sessions
than they could before the intervention. We found one study
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 114
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
evaluating the impact of small-group training on learners
behaviors.
10
Objective structured clinical examination scores
of students who had participated in small-group training and
those who had not were compared and the comparison
demonstrated signicantly better humanistic behavior scores
(e.g., communication and empathy skills) among those stu-
dents who had participated in the training sessions.
One-to-one learning with SPs allows for simultaneous
assessment of actual bad-news delivery behaviors during and
after educational interventions. In most of these interven-
tions faculty and/or SPs use checklists to identify learners
strengths and weaknesses. In two reports, residents partici-
pated in encounters with SPs and were provided once with
feedback.
35,36
In addition to demonstrating skills in certain
areas of bad-news delivery (learners concern for patients,
honesty, and appropriate follow-up plans), faculty observers
of these encounters identied areas for improvement, such as
providing too much data and scientic information during
the encounter.
35
Studies suggest the limitation of only pro-
viding one opportunity for learners to practice their skills and
receive feedback without having the chance to practice
applying behaviors recommended in the feedback. For exam-
ple, one study
36
found that residents ratings of their own
abilities were actually lower after a one-time encounter with
feedback than they were before the encounter. In contrast,
reports by others
33,34
demonstrate improvement of learner
skills when compared across two simulated encounters. For
example, one of these studies
33
found signicant differences
in content and communication skills between students who
had received training and feedback through previous simu-
lated encounters for delivering bad news and those who had
not received training and feedback.
We found no published reports that systematically evalu-
ated the effectiveness of learning about delivering bad news
in the less formal settings of inpatient wards and outpatient
clinics. A few studies have reported student and resident
experiences in bad-news delivery in the context of patient
care. Two recent reports
42,44
found that many students and
residents received little guidance from or opportunity to
debrief with faculty around these bad-news encounters. They
also reported that students and residents desired this guid-
ance and found discussion, observation, and feedback bene-
cial when provided in these contexts.
CONCLUSIONS
There are a variety of approaches available for teaching skills
in bad-news delivery. All of the interventions we describe
here have been rated highly by learners and have demon-
strated impact on learner self-condence and, in some cases,
learner knowledge and behaviors.
Adult learning is best facilitated through instruction that
is interactive and learner-centered, draws on previous expe-
rience and knowledge, is relevant to the learners practice,
allows the learner to apply what is being learned in a timely
manner, and includes the opportunity for feedback and
reection. Based on these adult learning principles and
ndings in the education literature on delivering bad news,
we conclude that the most effective interventions present
basic steps to effectively delivering bad news, and provide
opportunities for learners to discuss concerns, practice, and
receive feedback on their skills.
Our recommendation of best practices in teaching skills
for delivering bad news echoes recommendations made by
others.
6,19
In addition, evaluations that include observation
of actual behaviors point to the benet of learners having
more than one opportunity to practice and receive feedback
so that they can try out new behaviors they may not have
demonstrated in their rst encounter. It is striking that in
evaluation of many of the interventions, learners indicated a
desire for more training and opportunities for practice. In
addition, researchers need to examine the impact of educa-
tional interventions on learners actual behaviors and learn-
ers long-term retention of these skills. Measures have been
developed specically for assessing skills for delivering bad
news that could be used for this purpose.
9
This review demonstrates there are many models for
teaching skills for delivering bad news; ones choice will
depend on resources available in terms of faculty, SPs, and
curricular time. In addition, deciding when to provide this
training to learners will depend on available resources, but
the training is likely to be most effective if it is provided early
and often. As suggested by Kurtz et al.,
43
following a helical
model where communication skills are reiterated and rein-
forced throughout medical training is essential to maximum
skill development. Thus, prior to or early in their direct
patient care experiences, medical students can benet from
training by having an opportunity to practice giving bad
news in a safe, simulated environment before having to
deliver bad news with actual patients and families. Early
training also gives students a framework in which to critically
evaluate role models they may observe giving bad news on
the wards and in the clinics. As students and residents have
increased patient-care responsibilities, new and more com-
plex aspects of bad-news delivery can arise. It has also been
argued that students communication skills tend to degrade
over the four years of medical school if the skills are not
reinforced.
43
Faculty who have the skill to recognize and
capture these teachable moments in the context of clinical
rotations can help learners discuss these issues and hone their
skills. At the resident level, formal instruction and practice
in delivering bad news guarantees that all residents, regard-
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 115
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
less of where they went to medical school, are equipped to
adequately perform this task.
In recommending more faculty-intensive educational in-
terventions to teach skills for delivering bad news (e.g., small
groups, role-plays with feedback, clinical teaching), it follows
that faculty also need and deserve training in providing this
instruction. Several successful models have been imple-
mented to train practicing physicians, some for the rst time,
to deliver bad news.
11,45,46
Thus, training will help improve
physicians own interactions with patients, as well as their
ability to teach others in formal settings and to identify
learning opportunities in the context of patient care.
Learning to deliver bad news effectively is an important
part of providing good medical care, maintaining productive
relationships with patients, and enhancing patient and phy-
sician satisfaction. Through educational interventions, the
bad-news encounter can be made less distressing for both
clinician and patient. To incorporate effective behaviors for
delivering bad news into practice, we encourage medical
education programs to commit the necessary resources to
provide a comprehensive approach to teaching this task, one
that includes multiple sessions and opportunities for demon-
stration, reection, discussion, practice, and feedback.
We are grateful to Dr. Jerold Woodhead for supporting our implementation
and renements of the curriculum for bad-news delivery and assisting in
thinking through the issues involved in this review. We are also grateful to
Dr. John F. Wilson for inspiration and assistance in development and
implementation the curriculum for bad-news delivery. In addition, we want
to thank Dr. Paul Casella for his editorial assistance.
REFERENCES
1. Bor R, Miller R, Goldman E, Scher I. The meaning of bad news in HIV
disease. Couns Psych Q. 1993;6:6980.
2. Girgis A, Sanson-Fisher RW. Breaking bad news 1: current best advice
for clinicians. Behav Med. 1998;24:539.
3. Ptacek JT, Eberhardt TL. Breaking bad news: a review of the literature.
JAMA. 1996;276:496502.
4. Falloweld LJ. Giving sad and bad news. Lancet. 1993;341:4768.
5. Maguire P. Breaking bad news. Eur J Surg Oncol. 1998;24:18891.
6. Falloweld LJ. Things to consider when teaching doctors how to deliver
good, bad and sad news. Med Teach. 1996;18:2730.
7. Buckman R. How to Break Bad News: A Guide For Healthcare Profes-
sionals. Baltimore: Johns Hopkins University Press, 1992.
8. Dosanjh S, Barnes J, Bhandari M. Barriers to breaking bad news among
medical and surgical residents. Med Educ. 2001;35:197205.
9. Eggly S, Afonso N, Rojas G, Baker M, Cardozo L, Roberson RS. An
assessment of residents competence in the delivery of bad news to
patients. Acad Med. 1997;72:3979.
10. Vetto JT, Elder NC, Tofer WL, Fields SA. Teaching medical students
to give bad news: does formal instruction help? J Cancer Educ. 1999;
14:137.
11. Baile WF, Kudelka AP, Beale EA, et al. Communication skills training
in oncology: description and preliminary outcomes of workshops on
breaking bad news and managing patient reactions to illness. Cancer.
1999;86:88797.
12. Ellis PM, Tattersall MHN. How should doctors communicate the
diagnosis of cancer to patients? Ann Med. 1999;31:33641.
13. Rabow MW, McPhee SJ. Beyond breaking bad news: how to help
patients who suffer. West J Med. 1999;171:2603.
14. Billings JA, Block S. Palliative care in undergraduate medical educa-
tion. Status report and future directions. JAMA 1997;278:7338
15. Pollack M. Educating new resident physicians in death notication.
Acad Med. 1999;74:721.
16. Rosenbaum ME, Kreiter C. Teaching delivery of bad news using expe-
riential sessions with standardized patients. Teach Learn Med. 2002;14:
1449.
17. Morgan ER, Winter RJ. Teaching communication skills: an essential
part of residency training. Arch Pediatr Adolesc Med. 1996;150:638
42.
18. Garg A, Buckman R, Kason Y. Teaching medical students how to break
bad news. Can Med Assoc J. 1997;156:115964.
19. Harden RM. Twelve tips on teaching and learning how to break bad
news. Med Teach. 1996;18:2758.
20. Edinger W, Robertson J, Skeel J, Schoonmaker J. Using standardized
patients to teach clinical ethics. Med Educ Online. 1999;4:6.
21. Romm J. Breaking bad news in obstetrics and gynecology: educational
conference for resident physicians. Arch Women Ment Health. 2002;
5:1779.
22. Knox JD, Thomson GM. Breaking bad news: medical undergraduate
communication skills teaching and learning. Med Educ. 1989;23:258
61.
23. McNeilly DP, Wengel SP. The ER seminar: teaching psychothera-
peutic techniques to medical students. Acad Psych. 2001;25:193200.
24. Angelos P, DaRosa DA, Derossis AM, Kim B. Medical ethics curricu-
lum for surgical residents: results of a pilot project. Surgery. 1999;126:
7017.
25. Magnani JW, Minor MA, Aldrich JM. Care at the end of life: a novel
curriculum module implemented by medical students. Acad Med. 2002;
77:2928.
26. Ungar L, Alperin M, Amiel GE, Beharier Z, Ries S. Breaking bad news:
structured training for family medicine residents. Patient Educ Couns.
2002;48:638.
27. Cushing AM, Jones A. Evaluation of a breaking bad news course for
medical students. Med Educ. 1995;29:4305.
28. Van Winkle NW, Rosenbaum M, Redwood S. Improving fourth-year
students complex interviewing skills. Acad Med. 1998;73:590.
29. Tolle SW, Cooney TG, Hickam DH. A program to teach humanistic
skills for notifying survivors of a patients death. Acad Med. 1989;506.
30. Kahn MJ, Sherer K, Alper AB, et al. Using standardized patients to
teach end-of-life skills to clinical clerks. J Cancer Educ. 2001;16:1635.
31. Fortin AH, Haeseler FD, Angoff N, et al. Teaching pre-clinical medical
students an integrated approach to medical interviewing: half-day work-
shops using actors. J Gen Intern Med. 2002;17:7048.
32. Serwint JR, Rutherford LE, Hutton N, et al. I learned that no death is
routine: description of a death and bereavement seminar for pediatrics
residents. Acad Med. 2002;77:27884.
33. Coletti L, Gruppen L, Barclay M, Stern D. Teaching students to break
bad news. Am J Surg. 2001;182:203.
34. Greenberg LW, Ochsenschlager D, ODonnell R, Mastruserio J, Cohen
GJ. Communicating bad news: a pediatric departments evaluation of a
simulated intervention. Pediatrics. 1999;103:12107.
35. Goldschmidt RH, Hess PA. Telling patients the diagnosis is cancer: a
teaching module. Fam Med. 1987;19:3024.
36. Rosenbaum ME, Wilson HJ, Sloan DA. A clinical instruction module
for delivering bad news. Acad Med. 1996;71:529.
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 116
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D
37. Jewett LS, Greenberg LW, Champion LAA, et al. The teaching of crisis
counseling skills to pediatric residents: a one-year study. Pediatrics.
1982;70:90711.
38. Roth CS, Watson KV, Harris IB. A Communication Assessment and
Skill-Building Exercise (CASE) for rst-year residents. Acad Med.
2002;77:7467.
39. Pan CX, Soriano RP, Fischberg DJ. Palliative care module within a
required geriatrics clerkship: taking advantage of existing partnerships.
Acad Med. 2002;77:9367.
40. Muir JC, Krammer LM, von Gunten CF. Training physicians in pallia-
tive care. Generations. 1999;23:915.
41. Makoul G. Medical student and resident perspectives on delivering bad
news. Acad Med. 1998;73:S35S37.
42. Wear D. Face-to-face with it: medical students narratives about their
end-of-life education. Acad Med. 2002;77:2717.
43. Kurtz S, Silverman J, Draper J. Teaching and Learning Communica-
tion Skills in Medicine. Oxford, England: Radcliffe Medical Press,
1998.
44. Orlander JD, Fincke BG, Hermanns D, Johnson GA. Medicine resi-
dents rst clearly remembered experiences giving bad news. J Gen
Intern Med. 2002;17:82540.
45. Miller SJ, Hope T, Talbot DC. The development of a structured rating
schedule (the BAS) to assess skills in breaking bad news. Br J Cancer.
1999;80:792800.
46. Faulkner A, Argent J, Jones A, OKeeffe C. Improving the skills of
doctors giving distressing information. Med Educ. 1995;29:3037.
Teaching and Learning Moments
NOT JUST A REGULAR CUSTOMER
As a third-year resident at University of Illinois at Chicago, Mary was a regular customer. She was a frail white
woman in her early 50s with grace in her features, but suffered from cirrhosis due to alcoholism. Mary required
repeated admissions for encephalopathy and variceal bleedings.
In one of her many admissions to the intensive care unit, I had to evaluate her. Seeing her repeatedly as an intern
and a resident, it seemed routine work. Her husband of more than 30 years recited her familiar medical history to me
and I thought she was in her usual encephalopathic state. As I proceeded to examine her, a small grin appeared on
her face. As I was doing a routine head and neck examination, I heard her whisper weakly, You have a nice smell
doctor. I didnt know what to say and was stunned by the comment. In that moment, I realized I had lost the
humanistic part of my patient during the years of seeing her repeatedly.
In our routine work and dealings with patients with chronic diseases, physicians tend to develop a not again
approach to these patients and ignore the fact that these chronic cases appreciate life and its subtletiesjust like
everyone else. A fresh smell of a person had made such an impact amongst the aroma of alcohol, Betadine, and other
chemicals in the hospital. Marys surprising comment changed my attitude towards patients with chronic diseases. I
learned to separate the disease and the human being in my patients more easily. This also reminded me of the quote:
Patients are evaluating you while you are evaluating the patients.
1
Over the next few months, Mary continued to
be admitted and I continued to treat her until she died of massive variceal bleeding. Not only did Mary change my
perspective of managing patients with chronic illnesses, but her evaluation of me gave her comfort and a smile. I am
proud that I was able to give her that moment of comfort. I continue to try to bring comfort and smiles to all of my
patients, especially my regular customers.
NAUMAN TARIF, MD
Dr. Tarif is assistant professor of medicine and consultant nephrologist, Department of Medicine, College of Medicine,
King Saud University, Riyadh, Saudi Arabia.
REFERENCE
1. Orient JM, Sapira JD. Sapiras Art and Science of Bedside Diagnosis. 1st ed. New York: Lippincott Williams & Wilkins, 1998:9.
AC A D E M I C ME D I C I N E , VO L . 7 9 , NO . 2 / F E B R U A R Y 2 0 0 4 117
S K I L L S F O R DE L I V E R I N G BA D NE W S , C O N T I N U E D