336
Influence of Physician Communication
on Newly Diagnosed Breast Patients'
Psychologic Adjustment and
Decision- making
Cleora S. Roberts, Ph.D., Charles E. Cox, M.D., Douglas S. Reinfgen, M.D.,
Walter F. Baile, M.D., and Michael Giberfini, Ph.D.
Background. Physician-patient communication is of
critical importance when a breast cancer diagnosis is
made, because the emotionally overwhelmed patient
must be educated about her disease and available treat-
ments so she can participate in decisions about her care.
A research study addressed the hypothesis that patients
whose surgeons used psychotherapeutic techniques dur-
ing the cancer diagnostic interview would have better
psychologic adjustment to their cancer.
Methods. One hundred women surveyed 6 months af-
ter surgery completed the Cancer Diagnostic Interview
Scale (CDIS) and the SCL-90-R,a measure of psychologic
well being.
Results. Factor analysis of the CDIS revealed that the
physician's caring attitude was perceived by the women
as most important, with information-giving as a much
weaker component. Multiple regression analysis sup-
ported the hypothesis that psychologic adjustment was
predicted by physician behavior during the cancer diag-
nostic interview. Other significant predictors of adjust-
ment were a history of psychiatric problems and premor-
bid life stressors.
Conclusions. Provision of information needed for de-
cision-making appears to be valued largely within the
context of a caring physician-patient relationship. Spe-
cific surgeons' behaviors believed to facilitate patient ad-
justment include expressing empathy, allowing sufficient
time for patients to absorb the cancer diagnosis, provid-
ing information, and engaging the patient in treatment de-
cision-making. Cancer 1994; 74:336-41.
Presented at the National Conference on Breast Cancer, Boston,
Massachusetts, August 26-28, 1993.
From the H. Lee Moffitt Cancer Center and Research Institute,
the University of South Florida, Tampa, Florida.
Supported in part by a Research and Creative Scholarship
Award from the University of South Florida.
Address for correspondence: Cleora Roberts, Ph.D., H. Lee
Moffitt Cancer Center and Research Institute, 12902 Magnolia Drive,
Tampa, FL 33612-9497.
Key words: breast cancer, psychologic adjustment, phys-
ician-patient communications, treatment decision-mak-
ing, patient education.
Physician-patient communication is of critical impor-
tance when a breast cancer diagnosis is made, because
the patient must deal simultaneously with a potentially
life-threatening diagnosis and the loss or altered ap-
pearance of a valued body part. In addition, the surgeon
must educate an emotionally overwhelmed patient
about her disease and available treatments so she can
participate in making decisions about her care.
In a previous study of psychosocial adjustment to
breast cancer, many patients reported that their
surgeon's caring attitude during the cancer diagnostic
interview was a n important part of their adjustment.
These observations led to the present study with the hy-
pothesis that the patient whose surgeon used psycho-
therapeutic techniques, such as listening, expressing ac-
curate empathy, and providing information and reas-
surance, would experience less psychologic distress in
her subsequent adjustment to the diagnosis and treat-
ment procedures.
Review of the Literature
The effects of physician communications on patient
outcomes have been studied extensively. Most studies
have focused on physician behaviors such as informa-
tion-giving, partnership-building, and interpersonal
skills and how these behaviors related to patient out-
comes of satisfaction with medical care, compliance,
and recall of information.',* Buller and Buller' tested a
social interaction explanatory model and concluded
that physician use of an "affiliative" communication
style (defined as friendly encouraging, open honest,
empathic and listening attentively) was related most
Influence of Physician CommunicationlRoberts et al.
strongly to patient satisfaction and positive evaluation
of medical care. Hall, Roter, and Katz' proposed a
theory of provider-patient reciprocation based on a dis-
tinction among task behaviors (information-giving,
question-asking, and technical competence) and socioe-
motional behaviors (partnership-building, social con-
versation, and interpersonal competence). They hy-
pothesized that a physician's task performance, such as
that of information-giving, is a reflection of caring and
positive regard and therefore has a more positive effect
on patient satisfaction and compliance than does mere
adeptness in socioemotional tasks alone.
Physician-patient communications in the oncology
setting have been addressed by S ~ h a i n ,G~ a n ~ and
,~
Hermann,5among others. Blanchard, Labrecque, Ruck-
deschel, and Blanchard6 explored cancer patients' per-
ceptions of physician behaviors and objective measures
of physician behavior as predictors of patient satisfac-
tion. The data confirmed their hypothesis that patient
perceptions were better predictors of satisfaction.
A number of studies have addressed the provision
of information, particularly undesirable information, to
cancer patients as the most salient issue in physician-
patient communication.7~sIn a review of this literature,
SiminofP concluded that nearly all studies point to se-
rious gaps in patient recall and understanding of the
information they are given. One possible explanation of
poor patient recall is physician provision of routinized
information rather than communications tailored to the
individual patient, as was reported in a study of com-
munication patterns of breast surgeons.8
Information and Decision-making
An important goal of patient education is to enable in-
formed decision-making and treatment choices. Two
studies have reported that openness in communication
encouraged patient de~ision-rnaking.~,~ On the other
hand, Cassileth et al.7 and Blanchard et a1.I' found that
although nearly all patients (90% and 92%, respec-
tively) wanted full information about their cancer, far
fewer (25% and 69%, respectively) wanted to be in-
volved in treatment decisions.
Concordance in Cancer Patient-Physician
Communication
Schain3 has proposed that ideal physician communica-
tion with breast cancer patients should be tailored to
the patients' needs or coping styles to reduce patient
distress. That is to say that patients differ in their ability
to integrate threatening information and that concor-
dance between disclosure of information and the pa-
337
tient's ability to deal with this information would be the
ideal.
In summary, patient satisfaction with physician
communication has been linked to interpersonal com-
petence and information-giving, with suggestions made
that physician communication tailored to patient cop-
ing styles may have optimal results. The purpose of this
study was to go a step beyond patient satisfaction to
explore the influence of physician communication on
breast cancer patients' psychologic adjustment.
Methods
Subjects
A sample of 25 newly diagnosed breast cancer patients
was drawn consecutively from the patient files of each
of two of the authors (C.E.C. and D.S.R.) who are breast
surgeons at the H. Lee Moffitt Cancer Center and Re-
search Institute. To sample patients treated by many
different surgeons, an additional 50 patients were re-
cruited via announcements at professional meetings
and in the newspaper. All of the 100 patients were
women.
The patients ranged in age from 29 to 82 years
(mean, 55.0; SD = 11.5). Sixty-four percent had a mas-
tectomy, and 36% elected to have a lumpectomy fol-
lowed by radiation treatments. Seventy-two percent of
the subjects were married, and 48% were employed
outside of the home. The average time-lapse after sur-
gery was 5.7 months at time of data collection.
Instruments
The first step was to design a questionnaire, the Cancer
Diagnostic Interview Scale (CDIS), to assess the pa-
tient's perceptions of her surgeon's behavior during the
interview in which she was informed of her cancer di-
agnosis and the treatment options available to her. Pa-
tient reports are considered to be valid based on the au-
thors' clinical observations that most cancer patients
can recall exactly when, where, and how they were told
they had cancer, although recall of technical and diag-
nostic and treatment information may be incomplete.
Patients' accounts strongly resemble the phenomenon
of "flashbulb memories", which are described by
Brown and Kulik"' as memories of events that are unex-
pected and biologically significant.
Previous research indicated that three classes of
physician behavior tend to improve patient adjustment
to cancer diagnosis: (1) providing information about the
~ a n c e r(2)
; ~ using interpersonal skills such as empathy,
listening, and conveying positive regard; and (3) instill-
ing hope."*'2 Therefore, the original scale contained
338 CANCER Supplement July I , 1994, Volume 74,No. 1

eight Likert items for each of these three factors. Pre- Table 1. CDIS Eigenvalues and Factor Loadings for
testing of the CDIS on 94 patients with different types Phase I1 Data for 136 Breast Cancer Patients*
of cancer, however, resulted in a two-factor solution Factor loading
with no support of instillation of hope as a separate fac-
tor. Eighteen of the original items were retained. A Factor I , Factor 2,
Abbreviated item Caring Information Factor 3
panel of experts, specifically a surgeon, a psychologist,
and a psychiatrist, all of whom had extensive clinical Doctor understood my fears 0.74 0.59
experience with breast cancer patients, rated the scale I felt hopeful 0.55 0.67
items to determine if they could be categorized reliably Doctor encouraged me to
express my feelings 0.58 0.58
as information-giving or interpersonal skills. Eleven Believed in my doctor 0.60 0.73
items were judged by the panel to reflect interpersonal Felt I was in good hands 0.60 0.77
skills, and seven were categorized as information-giv- Doctor warm and caring 0.74
ing. Doctor cared about me 0.77
Principal components analysis using varimax pro- Doctor remained detached7 0.63
cedures was repeated for a sample of 136 breast cancer Wish doctor took time to
patients, including the 100 in the present sample, who answer questionst 0.57
Doctor abrupt with newst 0.66
completed the CDIS approximately 6 months after sur-
Doctor impatient with my
gery. A three-factor solution was obtained, with the first questionst 0.69
factor explaining 43% of the variance and the second I was given information 0.63
and third factors each contributing considerably less ex- Doctor discussed treatments 0.72
planation (7.6% and 6.3%, respectively.) We then ex- Doctor explained procedures 0.75
amined the structure matrix to ascertain which items I understood information 0.71
correlated most strongly with factor 1 (Table 1). They Doctor uncomfortable when I
was emotionalt 0.59 0.65
were as follows: ”I felt my doctor cared about me as a
I needed more time to ask
person,” ”my doctor understood my fears and con- questionst 0.65 0.60
cerns,” and “my doctor is a warm and caring person.” Wish doctor had been more
This factor was labeled “caring”. Of the 11 items judged hopefult 0.70
to indicate a physician’s interpersonal skills, 10 had Eigenvalues 7.79 1.37 1.14
loadings of 0.55 or more on Factor 1. Three items rated CDIS: ductal carcinoma in situ.
by the judges as information-ping also were loading Loadings less than 0.55 are omitted
t Reverse scored item.
on Factor 1 with coefficients of 0.57 or more.
The two items most clearly defining Factor 2 were
”my doctor discussed different treatments available for
my type of cancer” and ”my doctor explained the need sures of each of these variables were made as discussed
for tests and procedures.” These items clearly reflect in- below.
formation-giving and were judged unanimously as Premorbid psychologic adjustment. Premorbid
such by the raters as well. Interpretation of the rather psychologic adjustment was measured in two ways.
weak Factor 3 is unclear. Cronbach’s alpha, a reliability First, to assess premorbid life stressors, each patient was
coefficient, for the CDIS was 0.92 on both administra- asked about changes in five areas of her life during the
tions, indicating that the scale has excellent internal year before her cancer diagnosis. These five areas were
consistency. (1)health, (2) work situation, (3) home and family situ-
Psychologic status. Psychologic status was as- ation, (4) personal and social life, and (5) financial or
sessed with the SCL-90-R, a standardized instrument legal status. Each patient who reported a life change
that evaluates the psychologic symptomatic distress of was asked to rate subjectively, on a scale of 1 to 10, the
medical patients.I3 It is a 90-item self-report symptom amount of stress created by the change. A subject’s
inventory assessing distress in nine dimensions. All score could range from 0 (no life changes) to 50 (change
items are used to compute the Global Severity Index, in all five areas with maximum subjective distress re-
which represents the best single indicator of distress. ported for each change).
Factors other than the physician’s handling of the A second measure of premorbid psychologic ad-
diagnostic interview are known to influence patient ad- justment was obtained by inquiring about prior treat-
justment to breast cancer. Previous research has dem- ment for emotional problems and the use of psycho-
onstrated that premorbid psychologic a d j ~ s t m e n t , ’ ~ , ’ ~tropic medications. A psychiatrist (W.F.B.)and psychol-
physical health ~ t a t u s , ’ ~and
- ’ ~ social also ogist (M.G.) reviewed patient answers and assigned a
are predictors of the cancer patient’s well being. Mea- score of 1(no psychiatric history), 2 (history of mild psy-
Influence of Physician Communication/Roberts et al.
chiatric problems), or 3 (significant psychiatric history).
Interrater concordance was 100%.
Health status. Health status was judged by phys-
ician ratings (C.E.C. and D.S.R.) of responses to a ques-
tionnaire regarding the patient’s recovery from surgery,
side effects from radiation and chemotherapy, level of
physical activity, and health problems unrelated to the
cancer diagnosis. Interrater concordance was 0.84 using
the Kappa statistic, which corrects for chance
agreement. The health status score could range from 0
(totally asymptomatic) to 12 (indicative of moderate
symptoms or impairment in all areas).
Social support. Social support was measured with
a self-report instrument, the Social Support Question-
naire, which was developed specifically for breast can-
cer patients2’ This Likert scale contains 24 items re-
garding emotional support from spouse, family mem-
bers, and friends.
Data collection techniques. Patients who re-
sponded to a letter from their surgeon or to a newspaper
announcement of the study were interviewed by tele-
phone by the principal investigator (C.S.R.)or research
assistants. The study was explained in detail, and in-
formed consent was obtained from all participants.
Data regarding premorbid life stressors, psychiatric his-
tory, and health status were obtained during the tele-
phone interview, which took place approximately 6
months after the patient’s surgery. The interviewers
used probes and follow-up questions. In addition, each
patient was encouraged to provide anecdotal data about.
her cancer diagnostic interview. These data provided
additional insights to supplement the quantitative mea-
sures obtained from the CDIS. The remaining question-
naires were mailed to patients to complete and return.
Results
Overall ratings of physician behavior during the cancer
diagnostic interview were high, with a mean of 76.8
(S.D. = 13.9) of a possible score of 90 and a range of 30-
90.
Multiple regression analysis was used to examine
the effects and the magnitude of those effects of the in-
dependent or predictor variables (cancer diagnostic in-
terview, social support, health status, premorbid life
stressors, and premorbid psychiatric history) on the de-
pendent variable of psychologic adjustment, the Global
Severity Index. Three of the five independent variables
entered the regression equation (Table 2).
Discussion
The results of this study support the hypothesis that a
surgeon’s use of basic psychotherapeutic techniques
339
Table 2. Regression Analysis of
Psychological Adjustment
Variable Beta R RZ
~
Psychiatric history 0.306 0.358 0.128
Cancer diagnostic interview -0.233 0.459 0.210
Premorbid life stressors 0.207 0.499 0.249
can have a significant positive influence on a patient‘s
psychologic well being. The findings underscore the po-
tential impact of the physician-patient relationship in
improved quality of life for cancer patients. Although
physician-patient communication previously has been
linked to patient satisfactionz1and specifically to cancer
patient satisfaction,” the present findings go a step fur-
ther by demonstrating that patients’ psychologic symp-
toms, such as depression and anxiety, also can be in-
fluenced by these interactions.
A history of treatment for emotional problems and
recent life stressors place the breast cancer patient at
risk for psychologic problems. The lack of correlation
between social support and psychologic adjustment in-
dicates that support from spouse, family, and friends
may be less influential than perceived support from the
physician during the early stages of diagnosis and treat-
ment. Health-care givers should not assume neither
that the presence of a supportive family will foretell a
good adjustment nor, conversely, that the absence of
this social network will forebode a poor adjustment.
Similarly, physical health was unrelated to psychologic
symptoms in this population, a finding also reported by
Taylor et al.23
Data derived from patient comments and the social
science literature indicate that the surgeon should do
the following during the cancer diagnostic interview:
1. Carefully select an appropriate place and time for
the interview. An ideal situation is a prearranged
return office visit, with the suggestion that the pa-
tient bring a spouse or other significant person. In
some situations, the cancer diagnosis may be given
after completion of the biopsy. If this is the more
feasible plan, the patient should not be approached
until she has regained full consciousness.
Surgeons of nearly all of the women in the study
followed this principle. One patient, however, was
given the diagnosis by telephone, and another (a
nurse) was told she had cancer by a pathologist col-
league while she was on duty. In yet another case,
the husband was given the news and left with the
burden of informing his wife. Several patients re-
called being given the diagnosis while still groggy
from anesthesia.
340 CANCER Supplement JuZy 1,1994, Volume 74, No. 1
2. Verbalize empathy to the patient and convey
awareness that the cancer diagnosis may a devas-
tating emotional impact. A simple statement such
as ”I know this news is very upsetting” gives the
patient permission to express her emotions without
feeling ashamed of being upset and tearful. Direct
eye contact and touching the patient also convey a
caring attitude.
3. Allow the patient sufficient time to absorb the news
and react emotionally before proceeding with med-
ical explanations or information. David SpiegelZ4
comments that in similar situations he cautions
medical students, ”Don’t just do something. . .
stand there.” Although there is wide variation
among individuals, most patients will regain a fair
degree of emotional control within several minutes.
The surgeon who sits down during the diagnostic
interview communicates acceptance, positive re-
gard for the patient, and willingness to devote time
to the important task at hand.
4. Provide information about the cancer and treat-
ment options. From both a medical and legal stand-
point, the advantages of patient education and in-
formed consent are clear. From a psychosocial per-
spective, patient education is viewed as a
moderator of psychologic d i s t r e ~ sEmpiric
.~ studies
have demonstrated that patient education reduces
treatment-related anxiety and depres~ion.’~
Based on her work with cancer patients, Hermann’
concluded that the skill with which physicians
communicate information to patients and families
is the most powerful tool at their disposal in helping
people cope with illness. CasselZ6conceptualized
information as a therapeutic tool that can
strengthen the relationship between physicians and
patients.
Communicating information at the point of diagno-
sis should be viewed as an interactive process, with
the surgeon listening carefully to the patient to un-
derstand her concerns and fears regarding breast
cancer and its treatments. A patient whose mother
died of breast cancer will interpret facts and statis-
tics very differently from a patient whose mother is
a 30-year survivor of the disease. Likewise, the loss
of a breast is viewed by some women as more trau-
matic than the potential loss of life.
A brief assessment of the individual patient’s ability
to integrate threatening information may facilitate
the decision as to how much information to give
during the diagnostic interview. Social scientists
have described two patient response styles: (1)
monitors, who want to know everything about their
situation, and (2) blunters, who want to be pro-
tected from frightening or painful information.
These and other issues are covered in depth in
Schain’s insightful article on physician-patient
communication about breast ~ a n c e r . ~
Although providing information is endorsed for its
psychotherapeutic value, it should be recalled that
the factor analysis of the CDIS showed that patients
perceived the importance of information-giving as
secondary to their surgeon’s caring attitude. Further
research on the relative importance of each of these
factors is indicated.
5. Engage the patient in treatment decision-making.
The benefits of increased patient participation in
health care decisions have been demonstrated and
certainly hold true for breast cancer patients, most
of whom can be offered breast preservation or mas-
tectomy with or without reconstructive surgery.
Physicians are encouraged to adopt the techniques
of cognitive behavior therapists who recognize the
psychotherapeutic value of restoring a sense of con-
trol to patients in crisis.27Cognitive therapists em-
phasize the collaborative nature of treatment and
an active role for the patient. Schain3observed that
active participation in decision-making and assum-
ing responsibility for one’s own treatments have
been suggested as possible contributors to the con-
trol of one’s disease and recovery.
A small number of patients in this study expressed
anger with physicians who attempted to rush or bypass
the decision-making process by scheduling the surgery
before discussing the diagnosis and treatment options
with them. A few patients, however, preferred situa-
tions in which the surgeon made decisions for them.
Variations in patient preferences underscore the impor-
tance of listening and attending to the individual pa-
tient.
Although it might be beneficial to involve mental
health professionals at the time of a breast cancer diag-
nosis, this is not always feasible, because both surgeon
and patient are eager to begin treatment. The surgeon
has the knowledge and skills needed by the patient,
who views the physician as the key actor in whose
hands her fate will rest. It may be wiser to train the
surgeon or other members of the cancer treatment team
in basic principles of psychotherapy rather than refer
the patient elsewhere for help in dealing with the psy-
chosocial sequelae of a cancer diagnosis.
References
1. Buller MK, Buller DB. Physician’s communication style a n d pa-
tient satisfaction. J Health SOCBehav 1987; 28:376-88.
2. Hall ]A, Roter DL, Katz NR. Meta-analysis of correlates of pro-
Influence of Physician Communication/Roberts et al.
vider behavior in medical encounters. Medi Care 1988; 26:657-
72.
3. Schain WS. Physician-patient communication about breast can-
cer: a challenge for the 1990s. Surg Clin North Am 1990; 70:917-
36.
4. Ganz PA. Patient education as a moderator of psychosocial dis-
tress. ] Psychosocial Oncol 1988; 6:181-97.
5. Hermann JF. Psychosocial support: interventions for physicians.
Semin Oncol 1985; 12:466-71.
6. Blanchard CG, Labrecque MS, Ruckdeschel JC, Blanchard EB.
Information and decision-making preferences of hospitalized
adult cancer patients. Soc Sci Med 1988; 27:1139-45.
7. Cassileth BR, Zupkis RV, Sutton-Smith MS, March V. Informed
consent: why are its goals imperfectly realized? N Engl ] Med
1980; 302:866-900.
8. Taylor KM. Telling bad news: physicians and the disclosure of
undesirable information. Sociol Health Illness 1988; 10:109-32.
9. Siminoff LA. Cancer patient and physician communication:
progress and continuing problems. Ann Behnu Med 1989; 11:
108-12.
10. Brown R, Kulik J. Flashbulb memories. In: Neisser U, editor.
Memory observed. San Francisco: Freeman and Co, 1982:23-40.
11. Blum RH, Blum DS. Psychosocial care of the cancer patient:
guidelines for the physician. ] Psychosoc Oncol 1988; 6:119-35.
12. Weisman A. Coping with cancer. New York: McGraw-Hill, 1979.
13. Derogatis LR. The SCL-90-R administration, scoring and proce-
dures manual-11. Towson, MD: Clinical Psychometric Re-
search, 1983.
14. Plumb MM, Holland JC. Comparative studies of psychological
function in patients with advanced cancer, 11: interviewer-rated
current and past psychological symptoms. Psychosom Med 1981;
43:243-54.
15. Lansky SB, List MA, Herrman CA, Ets-Hukin G , DasGupta TK,
341
Wilbanks GD, et al. Absence of major depressive disorder in fe-
male cancer patients. j Clin Oncol 1985; 3:1553-60.
16. Craig TJ, Comstock GW, Geiser PB. The quality of survival in
breast cancer: a case control comparison. Cancer 1974; 33:1451-
7.
17. Aaronson, NK. Quality of life: what is it? how should it be rnea-
sured? Oncology 1988; 269-74.
18. Wortman C. Social support and the cancer patient. Cancer 1984;
53:2339-60.
19. Funch DP, Mettlin C. The role of support in relation to recovery
from breast surgery. SOCSci Med 1982; 16:91-8.
20. Northouse LL. Social support in patients’ and husbands’ adjust-
ment to breast cancer. Nurs Res 1988; 37:91-5.
21. Ware JE, Davies A, Stewart AL. The measurement and meaning
of patient satisfaction. Health Med Care Sen, Rev 1978; 1:l-15.
22. Blanchard CG, Ruckedeschel JC, Fletcher BA, Blanchard EB. The
impact of oncolopts’ behaviors on patient satisfaction with
morning rounds. Cancer 1986; 58:387-93.
23. Taylor SE, Lichtman RR, Wood JV, Bluming AZ, Dosik GM, Lei-
bowitz RL. Illness-related and treatment-related factors in psy-
chological adjustment to breast cancer. Cancer 1985; 55:2506-
...I
13.
24. Speigel D. Does living better mean living longer? Effects of group
support on metastic breast cancer patients. Presented at the an-
nual meeting of the National Association of Oncology Social
Workers; 1991 Apr 29; Monterey, California.
25. Mumford E, Schlesinger HF, Glass GV. The effects of psycho-
logical intervention on recovery from surgery and heart attacks:
an analysis of theliterature.Am]PublicHealth 1982; 72:141-51.
26. Cassel E. Talking with patients. Vol 2. Cambridge: MIT Press,
1985.
27. Scott J, Williams MG, Beck AT. Cognitive Therapy in Clinical
Practice: An Illustrative Casebook. New York: Routledge, 1989.