1. Patient Education and Counseling 60 (2006) 10–15
www.elsevier.com/locate/pateducou
Ability of primary care physician’s to break bad news:
A performance based assessment of
an educational intervention
Gilad E. Amiel a,c, Lea Ungar b,c, Mordechai Alperin b,c,
Zvi Baharier b,c, Robert Cohen d, Shmuel Reis b,c,*
a
Bnai-Zion Medical Center, Department of Urology, Haifa, Israel
Clalit Health Services, Haifa District, Department of Family Medicine, Haifa, Israel
c
The Technion-Isreal Institute of Technology, Ruth and Bruce Rappaport Faculty of Medicine,
P.O. Box 9649, Bat-Galim, Haifa 31096, Israel
d
Center for Medical Education, Hebrew University, Faculty of Medicine, Jerusalem, Israel
b
Received 14 August 2004; received in revised form 15 April 2005; accepted 23 April 2005
Abstract
Objective: We have previously described a breaking bad news (BBN) training program for primary care physicians [Ungar L, Alperin M,
Amiel GE, Beharier Z, Reis S. Breaking bad news: structured training for family medicine residents. Patient Educ Couns 2002;48:63–68].
In this paper, we present the assessment of an educational intervention aimed at improving this important skill.
Methods: The assessment tool was an eight station objective structured clinical examination (OSCE) utilizing standardized patients (SPs).
Intervention and control groups of 17 general practitioners (GP) each were evaluated before and after an educational intervention, or a Balint
group (control).
Results: Intervention group GPs significantly increased their average grade on the post-test as compared to the pre-test (58.5, S.D. 12.7 versus
68.4, S.D. 9.2), effect size 0.94. Improvement in the control group was minimal (pre-test 57, S.D. 10.4 versus 58.1, S.D. 9.5 for the post-test),
effect size 0.23. Reliability of the OSCE was a = 0.81.
Conclusion: The performance assessment used in this study proved to be a reliable and valid tool to assess the ability of physicians to break
bad news. It provided evidence of the effectiveness of the intervention.
Practice implications: BBN training can and should be evaluated by valid and reliable measures. SPs can serve as reliable evaluators of BBN
training.
# 2005 Elsevier Ireland Ltd. All rights reserved.
Keywords: Breaking bad news; OSCE; Continuing medical education; General practitioners
1. Introduction
There was a time when it was an acceptable practice to
break bad news to a patient who was suffering from a
terminal illness by mail, often without even seeing the
patient [1]. Fortunately, the medical profession has made
tremendous strides in dealing with this area of practice [2,3].
Consensus guidelines on how to break bad news to patients
* Corresponding author. Tel.: +972 4 8295402; fax: +972 4 8295249.
E-mail address: reis@netvision.net.il (S. Reis).
as outlined by Rosenbaum et al. [4] Buckman [5] and by
Baile et al. [6] represent some of the many attempts to
establish basic principles for breaking bad news (BBN).
A number of studies have shown that physicians
experience difficulty when required to deliver bad news
[7]. Lack of skills and the reluctance to deal with the
patient’s feelings have been reported as the main causes for
physicians’ avoidance of this task [8,9]. To overcome these
problems, courses for breaking bad news have been
implemented [10]. Of crucial importance is the effectiveness
and outcome of such interventions, i.e. do they improve the
0738-3991/$ – see front matter # 2005 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.pec.2005.04.013

2. G.E. Amiel et al. / Patient Education and Counseling 60 (2006) 10–15
ability of participating physicians to breaking bad news, and
to what degree do participants retain these skills. Assessment of the impact of such courses on competence is rare [2].
We were unable to find studies that reported on the
development of a reliable performance based assessment of
the ability of physicians to deliver bad news to patients. The
purpose of this study was to: (1) evaluate the reliability and
validity of a competence based assessment, utilizing
simulated patients as evaluators, to assess primary care
physicians’ ability to deliver bad news; (2) evaluate the
effectiveness of a training program in breaking bad news
offered to a group of general practitioners (GPs) as part of a
continuing medical education (CME) program.
2. Method
2.1. Course framework and teaching modalities
In 1991, a mandatory course for second year family
medicine trainees on how to break bad news was introduced
into our residency training program. Since 1996, this course
has also been offered as a CME course for practicing GPs.
The guiding textbook for this course has been ’How to break
bad news’ by Buckman [5].
A group of certified family physicians and a social worker
identified common and important situations dealing with bad
11
news in primary care that served as the basis for developing
the teaching program. Based on this list, a blueprint of 14
relevant encounters for teaching and discussion was constructed. Each of the fourteen 90-min small group sessions
included four elements: (1) a theoretical component, dealing
with methods of managing stress and crisis intervention; (2)
clarifying personal attitudes and coping with providers’
emotions when breaking bad news; (3) communication skills;
(4) practicing communication by interviewing simulated
patients. A detailed description of the course has been
published elsewhere [11].
2.2. The examination
A performance based assessment tool, an objective
structured clinical examination (OSCE) was developed to
evaluate primary care physicians’ ability to deliver bad news
to patients. The OSCE format was chosen, since it provided
an opportunity to simulate multiple doctor–patient encounters in a standardized setting. This method has been shown to
be reliable and valid, and has been widely used to assess the
performance of medical students, residents and practicing
physicians [12]. Eight 15-min stations representing breaking
bad news scenarios commonly encountered by primary care
physicians were developed (Table 1).
Based on the course curriculum, a list of the skills required
for providing bad news and coping with patients’ feelings was
Table 1
Topic, description and communication challenge of the eight OSCE stations in breaking bad news
Station No.
Topic
Description
Communication challenge
1
Anger due to missed diagnosis
Coping with a patient’s anger
2
Reactive depression
3
Perceiving that death is imminent
4
Fear of illness and disability
5
Difficulty in accepting the role
of a patient
6
Coping with uncertainty before
definite diagnosis
Breaking unexpected bad news
to a patient
A 24-year-old student who was treated with
NSAID after complaints of pain in right
knee later diagnosed to suffer from a tumor
in the right Tibia
Sleep disorders and loss of appetite in a
young mother of a 4-month-old baby that
was diagnosed as Down syndrome
Home-visit to a 65-years-old pharmacist
suffering from end-stage cancer of lung
who wishes to discuss with the doctor
his coming death
A 35 years old woman who was recently
diagnosed as suffering from multiple
sclerosis and didn’t receive information
concerning the disease. Appears tense
and anxious
A 40-year-old with risk factors for heart
disease who denies repeated measurements
of high blood pressure
A 50-year-old woman comes to the office
after palpating a lump in her breast
A 60-year-old patient who was sent to a
gastroscopy after recurrent upper abdominal
discomfort. You thought it is a peptic ulcer,
but the biopsy results show cancer of stomach
A 25-year-old woman who comes to the office
to receive test results which show that she is
suffering from Hodgkin’s disease
7
8
Breaking bad news to a patient
who does not wish to know
her condition
Treatment of reactive depression
after the birth of a disabled child
Coping with a conversation on
an approaching death
Coping with a patient with anxiety
after learning about a serious illness
Coping with denial and getting
compliance
Preparing toward probable bad
news in an uncertain situation
Breaking bad news to an alert and
intelligent patient who wishes to
receive every bit of information
Coping with a patient who does
not wish to know any details about
her disease and getting compliance
from the patient

3. 12
G.E. Amiel et al. / Patient Education and Counseling 60 (2006) 10–15
outlined by the course directors. A group of six senior family
physicians that served as tutors in the Department of Family
Medicinewere asked to list common and/or critical encounters
where they were required to deliver bad news to patients in their
clinic. These encounters served as the examination blue print,
and eight scenarios were developed. The course tutors
evaluated the cases to assure that they would require the
utilization of the skills taught in the course if they were to be
dealt with successfully.
An OSCE consisting of eight ‘long cases’ was
hypothesized to be sufficient to achieve acceptable
reliability.
Simulated patients (SPs) or chronically ill patients with
previous experience in role-playing were trained to present
the scenarios. Each SP received 8 h of training, which
included dealing with the communication challenges and the
rating scales to be used for assessing the performance of the
physicians. Trainers discussed with SPs their personal
attitudes concerning breaking bad news, the objectives of the
course and acceptable standards in breaking bad news
according to the literature. A single simulated patient was
assigned to each station, and participated in both the pre- and
post-tests.
SPs evaluated the candidates utilizing global ratings. Two
5-point Likert scale questionnaires were developed for each
station. The first was a 7-item communication scale to assess
principles and techniques in breaking bad news, common to
all stations (Table 2). Based on the known guidelines for
breaking bad news [4–6], items for evaluating basic
communication skills were selected. The second questionnaire was a 3-to-4-item questionnaire, tailored for each
scenario. Items chosen from the above guidelines were
aimed at evaluating the specific communication problem/s
the physician was required to deal with in each station.
For example, one of the stations dealt with a 35 years old
woman, who was discharged from a neurological ward, with
a diagnosis of multiple sclerosis. The communication
challenges the doctor had to cope with in this station were:
1. To acknowledge the patient’s anxiety.
2. To inform her about the natural (slowly progressive)
history of the disease and treatment options.
3. To assure her of his/hers support during the illness.
According to these challenges, the SP had to complete
three evaluation items (on a 1–5 Likert scale):
1. To what extent did the doctor tried to find out what you
know about the disease and its prognosis?
2. To what extent did the doctor explains the natural
development of the disease?
3. To what extent did the doctor check your understanding
of his/her explanation?
Candidates were assessed on a total of 10–11 items using
a 5-point rating scale in each of the eight stations.
Table 2
Breaking bad news OSCE 1–5 global rating communication scale No. 1 (common to all stations)
(1) To what extent did the doctor use appropriate verbal techniques in order to convey comfort and trust, encouraging you to cooperate during the
interaction? (i.e. used open-ended questions, used lay language, did not talk too fast or gave long speeches)
very much 5 4 3 2 1 not at all
(2) To what extent did the doctor express non-verbal empathy toward your situation? (i.e. maintained attentive pose and eye contact; body language that
conveyed warmth, sympathy and encouragement; touched you if it was appropriate)
very much 5 4 3 2 1 not at all
(3) To what extent did the doctor assess the presence of family and other resources that might help you cope with the situation? (i.e. wife, parents,
children, friends; how good is the relationship and how much can you rely on it; what kind of support can you expect to receive: emotional or
financial, if great expenses are under way due to the situation)
very much 5 4 3 2 1 not at all
(4) To what extent did the doctor manage to express personal commitment to you and your problem, and his full dedication while helping you
throughout your struggle? (i.e. scheduled a close follow-up appointment to further discuss the situation; offered to refer or inquire with experts
in the field; gave you the feeling that he/she cares what will happen to you in the near future)
very much 5 4 3 2 1 not at all
(5) To what extent did the doctor manage to give you as a patient a sense of hope without denying the truth or describing it unrealistically?
(i.e. gave you a feeling that your problem is treatable; discussed options in a positive way; maintained a balance between explaining benefits
of a treatment and side-effects)
very much 5 4 3 2 1 not at all
(6) To what extent did the doctor address your feelings? (i.e. asked a specific question about how you feel; touched upon specific concerns
you have raised; expressed support when emotions have arisen)
very much 5 4 3 2 1 not at all
(7) Overall, to what extent were you satisfied with the doctor? Would you wish to continue to be his patient in the future
very much 5 4 3 2 1 not at all

4. G.E. Amiel et al. / Patient Education and Counseling 60 (2006) 10–15
13
Effectiveness of the intervention was determined by
comparing pre- and post-course OSCE scores of the study
and control groups.
Table 3
Results of pre- and post-test scores for study and control groups
2.3. Study and control group
Pre-test
58.5 (S.D. 12.7) 57 (S.D. 10.4)
Post-test
68.4 (S.D. 9.2) 58.1 (S.D. 9.5)
Statistical significance p < 0.01
–
Thirty-four GPs participating in a CME program were
invited to participate in the study. Seventeen physicians who
selected the ‘breaking bad news’ course served as the study
group, and 17 physicians who selected a ‘Balint group’
course served as the control group [13]. The two groups were
matched for age, sex and years in practice as GPs. Mean age
of participants was 43.7years (S.D. 6.74) and 46.4years
(S.D. 5.49), respectively (t = À1.228; p = 0.208; z = 1.623;
p = .106), 10 female and 7 male doctors constituted the BBN
group, while 11 and 6, respectively, constituted the Balint
group (x2 = 0.125; p = 0.50), mean years of practice as GPs
was 17.8 years and 19.4 years, respectively (t = À0.769;
p = 0.432; z = À0.828; p = 0.413).
In the Balint course, GPs held group discussions about
patients and situations in their clinic that triggered exceptional
emotional reactions. A clinical psychologist and a senior
family physician conducted these discussions. GPs tried to
understand their personal feelings of transference and counter
transference and to get an insight as to their influence on
doctor–patient interaction. Participants in both groups confirmed that they had not undergone previous training for
content of the modules offered. Both the study and control
groups took the OSCE as a pre-test before starting
their respective courses. They took the same post-test examination, i.e. with the same case scenarios and same SP in each
station, at the completion of their respective courses. SPs had
no information concerning the course the participants
attended, they were blind to treatment versus control group.
The specific scenarios in the examination were intentionally
not dealt with in the breaking bad news course.
2.4. Analysis
Mean scores and standard deviations were calculated for
each participant, station and the rating scales used in the
examination. Reliability for the overall examination was
calculated by the internal consistency statistic Chronbach
alpha. Mean scores and standard deviations were calculated
for both experimental and control groups. Independent
sample t-tests was used to determine if there were significant
differences between the study and control groups on entry
into the program ( p < 0.05), and effect size was calculated
to determine the impact of the interventions given to both
study and control groups [14].
3. Results
All 34 physicians took both the pre- and post-tests
(Table 3). Overall mean score for the pre-test was 57.3, S.D.
Study group
(BBN) (n = 17)
Control group
Statistical
(Balint) (n = 17) significance
–
p < 0.01
Range: 20–100 (100: high performance). Overall result in the pre-test and
post-test OSCE in breaking bad news for general practitioners. The study
group took a breaking bad news course and the control group a ‘Balint’
course.
11.3 (range: 20–100). No significant difference on the pretest was found between the scores of the GPs from the study
and control groups (58.5, S.D. 12.7 versus 57, S.D. 10.4,
respectively; range: 20–100). Overall reliability of the pretest was high for a 2 h OSCE (a = 0.81).
The GPs in the study group significantly increased their
average grade on the post-test as compared to the pre-test
(58.5, S.D. 12.7 versus 68.4, S.D. 9.2; range: 20–100), effect
size 0.94, whereas the improvement in the performance of
the control group was minimal (pre-test 57, S.D. 10.4 versus
58.1, S.D. 9.5, for the post-test; range: 20–100), effect size
0.23. Overall reliability of the post-test was a = 0.78.
4. Discussion and conclusion
4.1. Discussion
At no time is effective communication more important
and challenging than when a physician is required to deliver
bad news or tragic information to patients and their families.
Receiving a medical diagnosis may be overwhelming
regardless of the care the physician takes in communicating
the news. Jonsen et al. have stated, ‘‘the truth may be brutal,
but the telling of it should not be’’. ([15]).
Little is known to date about actual physician performance in providing bad news and the emotional support they
may or may not provide the patient. Many courses and
guidelines aimed at improving the ability of physicians to
present bad news have been described [2–6,16–25]. Most of
these models and guidelines focused on the technique of
breaking bad news, neglecting the accompanying emotions
and personal attitudes of the involved physicians.
The instruction module offered to physicians in the study
group focused on providing the knowledge, skills and
attitudes that would assist them with breaking bad news in
diverse situations. The cases presented during the course
represented a wide variety of possibilities that doctors might
encounter during their day-to-day practice. Some of the
cases were designed to deal not only with delivering the bad
news but also on coping with its emotional consequences.
Documentation of the effectiveness of courses devoted to
learning how to break bad news, or the assessment of
physician’s competence in breaking bad news are scarce. In

5. 14
G.E. Amiel et al. / Patient Education and Counseling 60 (2006) 10–15
the few attempts that were found in the literature, the
decision whether a physician is capable of adequately
breaking bad news to a patient was usually based on one (or
two at the most) interactions evaluated by communication
experts [4,19–22]. The OSCE used in this study proved to be
a reliable and valid tool to assess the ability of physicians in
breaking bad news. Validity of the OSCE has been achieved
through: (1) the questionnaire (content validity), which was
developed by two experts in patient–doctor communication
including BBN; (2) the OSCE scenarios, which were
developed by three expert family physicians, based on their
clinical experience. The large effect size (close to one
standard deviation) for the study group provides evidence of
the effectiveness of the intervention in this group of
physicians.
Global ratings of performance on OSCE stations have
been shown to have promising psychometric properties [25].
Subjective global ratings have been shown to yield more
reliable and valid information on the performance of
practicing physicians than the objective detailed checklists
of the traditional OSCE, which are characterized by a lack of
flexibility in rewarding different approaches to problem
solving [26]. The utilization of Likert scale type global
ratings as opposed to detailed checklists as the assessment
tool in our stations enabled the SPs to assess the quality of
the doctor–patient interaction, and not merely the technical
aspects of breaking bad news.
It is well documented in the literature that SPs can be
appropriate evaluators of communication skills [3,16]. Our
findings also demonstrate the feasibility of utilizing welltrained SP as the evaluator of the physicians’ breaking bad
news skills, and not necessarily a communication expert.
SPs have also provided feedback to examinees after the
encounter, which may prove to be an effective educational
methodology [4].
As demonstrated in our study, the experience of the GPs
does not always reflect competence. It is questionable,
therefore, whether physicians who have practiced for
many years, but have not obtained necessary communication skills, possess the required proficiency to adequately
provide bad news to their patients. The present study
demonstrated the effectiveness of the breaking bad news
course developed for veteran GPs. Furthermore, by
comparing the study and control groups, we demonstrated
that a course focusing on the specific skills required to
deliver bad news is significantly superior to a more diffuse
experience of discussing communication issues and
personal experiences as in the Balint group. It would
appear that the consciousness raising experience and
opportunity to discuss personal experience with colleagues, such as is practiced in the Balint group, does in of
itself not improve doctors’ competence in delivering bad
news, and dealing with patients’ feelings.
Our study demonstrated that the OSCE can be utilized as
a reliable and valid tool to assess physicians’ competence in
BBN. Moreover, it can evaluate the quality of BBN training,
as reflected by the participants’ performance at the end of the
course. However, it will be necessary to further examine if
our measure can serve as an effective screening tool for the
identification of family physicians who are in need of
enhancement of their communication skills. A major
drawback of this assessment tool, as pointed out by the
examinees, is the artificial situation in which the physician is
required to deliver bad news to eight consecutive patients.
Examinees reported that although they were obviously
aware that it was a simulation, the OSCE was extremely
demanding emotionally. Undoubtedly, in real life, family
physicians do not experience such intensive encounters
successively. Other drawbacks of the present study include
the small size of the population, and the fact that participants
chose the intervention instead of being selected randomly.
Although most of them did not know what a Balint group
was, their choice may express their personal perception that
they do not need the breaking bad news course. However, in
dealing with emotions, it was speculated that the Balint
group might also contribute to breaking bad news skills as
much as a specific course.
Furthermore, we realize that this is not a direct reflection
of performance in actual practice, and the predictive validity
is yet to be determined. Therefore, this study is not complete
without evaluation of these acquired skills as applied to real
patients. The extent of long-term retention of these skills and
the timing of reinforcement are unexplored fields that
provide future challenges in the research of breaking bad
news.
4.2. Conclusions
The OSCE utilizing SPs as evaluators can be utilized as a
reliable and valid tool to assess the communication skill of
breaking bad news to patients.
4.3. Practice implications
BBN training can and should be evaluated by valid and
reliable measures.
SPs can serve as reliable evaluators of BBN training.
References
[1] Forrester J. Postal diagnosis: breaking the bad news in the 17th
century. Br Med J 1995;331:1694–6.
[2] Fallowfield L, Jenkins V. Communicating sad, bad, and difficult news
in medicine. Lancet 2004;363:312–9.
[3] Rosenbaum ME, Kreiter C. Teaching delivery of bad news using
experiential sessions with standardized patients. Teach Learn Med
2002;14:144–9.
[4] Rosenbaum ME, Ferguson KJ, Lobas JG. Teaching medical students
and residents skills for delivering bad news: a review of strategies.
Acad Med 2004;79:107–17.
[5] Buckman R. How to break bad news—a guide for health care
professionals. Baltimore, MD: The Johns Hopkins University Press,
1992.

6. G.E. Amiel et al. / Patient Education and Counseling 60 (2006) 10–15
[6] Baile WF, Buckman R, Lenzi R, Glober G, Beale EA, Kudelka AP.
SPIKES—A six-step protocol for delivering bad news: application to
the patient with cancer. Oncologist 2000;5:302–11.
[7] Sykes N. Medical students’ fears about breaking bad news. Lancet
1989;2:564.
[8] Dosanjh S, Barnes J, Bhandari M. Barriers to breaking bad news
among medical and surgical residents. Med Educ 2001;35:
197–205.
[9] Cantwell BM, Ramirez AJ. Doctor–patient communication: a study of
junior house officers. Med Educ 1997;3:17–21.
[10] Ptacek JT, Ptacek JJ, Ellison NM. ‘‘I’m sorry to tell you. . .’’ physicians’ reports of breaking bad news. J Behav Med 2001;24:
205–17.
[11] Ungar L, Alperin M, Amiel GE, Beharier Z, Reis S. Breaking bad
news: structured training for family medicine residents. Patient Educ
Couns 2002;48:63–8.
[12] Sharp PC, Pearce KA, Konen JC, Knudson MP. Using standardized
patient instructors to teach health promotion interviewing skills. Fam
Med 1996;28:103–6.
[13] Brock CD, Stock RD. A survey of Balint group activities in U.S.
family practice residency programs. Fam Med 1990;22:33–7.
[14] Lipsey MW. Designed sensitivity. London, UK: Sage Publications,
1990. p. 31–32.
[15] Jonsen A, Siegler M, Winslade W. Clinical ethics, 3rd ed., New York,
NY: McGraw-Hill, 1992. p. 53.
[16] Ladyshewsky R, Gotjamanos E. Communication skill development in
health professional education: the use of standardized patients in
combination with a peer assessment strategy. J Allied Health
1997;26:177–86.
15
[17] Cushing AM, Jones A. Evaluation of a breaking bad news course for
medical students. Med Educ 1995;29:430–5.
[18] Betson CL, Fielding R, Wong G, Chung SF, Nestel DF. Evaluation of
two videotape instruction programmes on how to break bad news for
Cantonese speaking medical students in Hong Kong. J Audiovisual
Media Med 1997;20:172–7.
[19] Balie 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;86:887–97.
[20] Ptacek JT, Eberhardt TL. Breaking bad news: a review of the literature.
J Am Med Assoc 1996;276:496–502.
[21] 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;80:792–800.
[22] Ambuel B, Mazzone MF. Breaking bad news and discussing death.
Prim Care 2001;28:249–67.
[23] Greenberg LW, Ochsenschlager D, O’Donnell R, Mastruserio J, Cohen
GJ. Communicating bad news: a pediatric department’s evaluation of a
simulated intervention. Pediatrics 1999;103:1210–6.
[24] Viadya VU, Greenberg LW, Patel KM, Strauss LH, Pollack MM.
Teaching physicians how to break bad news: a 1-day workshop using
standardized parents. Arch Pediatr Adolesc Med 1999;153: 419–22.
[25] Cohen R, Rothman AI, Poldre P, Ross J. Validity and generalizability
of global ratings in an objective structured clinical examination. Acad
Med 1991;66:545–8.
[26] Norman GR, Davis D, Lamb S, Hannah E, Caulford P, Kaigas T.
Competency assessment of primary care physicians as part of a peer
review program. J Am Med Assoc 1993;270:1046–51.