Jco 2004 02 166
Jco 2004 02 166
T H E A R T O F O N C O L O G Y:
JOURNAL OF CLINICAL ONCOLOGY When the Tumor Is Not the Target
From the Department of Medical On- available through physicians, the health
cology, Princess Margaret Hospital/Uni- HERE’S THE CASE
versity Health Network, University of
care team, media, and other sources, can
Toronto, Toronto, Ontario, Canada; A 74-year-old war veteran presents with make for a complex decision-making pro-
Medical Psychology Research Unit, bowel obstruction, is found to have multiple cess for patients.
Royal Prince Alfred Hospital, and De-
partment of Cancer Medicine, Univer-
liver and omental metastases at the time of Patients increasingly define themselves
sity of Sydney, Sydney, Australia. surgery, and undergoes palliative resection as consumers of health care. Many cancer
Submitted February 25, 2004; accepted of a primary colorectal cancer. The patient, patients today want detailed information
February 25, 2004. asymptomatic following his postoperative about their cancer diagnosis, prognosis, and
N.L. is supported in part through an recovery, is referred to a medical oncologist, treatment options, and many wish to be ac-
American Society of Clinical Oncology who, recommending palliative systemic tive participants in medical decision mak-
Career Development Award.
chemotherapy, initiates fluorouracil-based ing. Support for involving patients in mak-
Authors’ disclosures of potential con-
flicts of interest are found at the end of
treatment. After 4 months, the patient is ing decisions about their care is mounting
this article. admitted once for febrile neutropenia, and with an increasing weight of evidence in sev-
Address reprint requests to Natasha has experienced fatigue, diarrhea, and eral areas of medicine, demonstrating that
Leighl, MD, Princess Margaret Hospital/ mouth sores. His doctor deems these toxic- this can contribute to better quality deci-
University Health Network, 5-222, 610
University Ave, Toronto, Ontario, Can-
ities minor, while the patient feels that the sions and can improve health outcomes.
ada M5G 2M9; e-mail: Natasha.Leighl@ treatment has compromised his quality of Among cancer patients, those offered
uhn.on.ca. life. His oncologist tells him that his scans choices in their treatment show better psy-
© 2004 by American Society of Clinical are unchanged. The patient wonders chologic adjustment and health-related
Oncology
whether this treatment has been worth- quality of life, while those feeling that they
DOI: 10.1200/JCO.2004.02.166 while, and whether he had other options at have had little control over their disease
the start of treatment. and treatment have a poorer psychosocial
Treatment decision making in ad- outcome.1,2 Moreover, patients who per-
vanced cancer is complex for many reasons. ceive that their physicians are making an
The goals of treatment are often palliative: to effort to facilitate their involvement in de-
improve disease-related symptoms and life cision making tend to be more involved in
quality for a limited period of time, and in that process and have greater satisfaction
some cases, to prolong average survival by and physician loyalty.3,4
weeks or months. The limited tumor selec-
tivity of many agents used in systemic cancer
PATIENT PREFERENCES FOR
therapy, however, means that substantial INFORMATION AND INVOLVEMENT
treatment-related toxicity is common,
though developments in targeted therapy Patients have variable preferences for infor-
may eventually change this. When the goal mation and involvement in their cancer
of treatment is not cure, treatment decisions care, with predictive variables including age,
are further complicated by patient and sex, education, and performance status. A
caregiver denial, anxiety, physical distress, survey of 1,012 Canadian women with early
and emotional distress. In addition, po- breast cancer revealed that most of these
tential misunderstandings of information, women wanted detailed information about
1759
their disease, that 22% wanted to select their own treatment, study of 244 Australian cancer patients, less than 20% cor-
that 44% wanted to select treatment collaboratively with rectly estimated the chance of treatment achieving cure,
their doctor, and that 34% wanted their doctor to select the prolonging life, or palliating their symptoms. While patient
treatment.5 A small study of 48 patients with colon cancer denial contributed to misunderstanding in this study, pa-
reported that while these patients have similar information tient rating of the clarity of information received was also
needs compared with patients with breast cancer, they have predictive of their understanding.10 What patients under-
strikingly lower involvement preferences.6 Indeed, 78% stand, or misunderstand, about their prognosis can lead to
wanted to play a passive role in decision making and 80% polarized decisions about management. A study of 916
perceived that they did so. Similarly, a study of 57 men with hospitalized patients with advanced non–small-cell lung or
prostate cancer found that the men wanted to be well in- colon cancer found that while doctors were reasonably ac-
formed, though 58% preferred that their doctor make the curate in their prediction of patient life expectancy, 82% of
final treatment decision.7 Estimates of the proportion of patients overestimated their life expectancy, and 59% were
cancer patients who achieve their desired involvement in significantly more optimistic about their life expectancy
their treatment decision making range from 34% to 42%.4,5 than their physician.11 The perception patients had of their
prognosis had a profound effect on their subsequent treat-
PATIENT INFORMATION RESOURCES ment decisions. Overly optimistic patients were nearly three
times more likely to pursue aggressive therapy over sup-
To be active participants in their care, patients must have an portive care, but their survival was no better than those who
accurate understanding of information about their disease pursued supportive care alone. Physicians are also contrib-
and treatment options. This requires both the provision of uting to the increasingly aggressive management of cancer
accurate information and its successful comprehension. patients at the end of life, with as many as 18% of patients
There are documented problems in both areas. Studies receiving chemotherapy within 14 days of death.12 Clearly,
examining the adequacy of information provided during physicians are obliged to determine what their patients wish
medical consultations have found that many patients are to know, to provide information on prognosis, including
not equipped to make informed decisions.8 Gattellari et al life expectancy, the evidence-based impact of therapy on
examined initial oncology consultations from a subset of survival and quality of life, and to review all treatment
118 patients with advanced cancer, and found that only options, including supportive care alone, in the setting of
58% were told about their life expectancy.8 While the goals advanced cancer.
of recommended treatment were reviewed in most consul-
tations, the impact of treatment on quality of life was dis- DECISION AIDS FOR PATIENT MEDICAL
cussed with only 36% of patients. The concept of uncer- DECISION MAKING
tainty of benefit was introduced in 60% of consultations,
but supportive care was discussed as a treatment option in There are several potential methods to promote patient
only 44% of consultations. Existing patient information understanding and to facilitate decision making. These in-
materials, which can supplement information discussed in clude patient information aids, techniques to reduce patient
the consultation, are often not in an acceptable, compre- anxiety, communication training for patients and/or physi-
hensible, useful format for patients. Further, information cians, and decision aids (DAs). DAs are defined as “interven-
available through the Internet is often inaccurate, mislead- tions designed to help people make specific and difficult
ing, or too complex for the average patient. However, there choices among options by providing information on the op-
are many initiatives to improve this, such as the American tions and outcomes relevant to the person’s health status.”13
Society of Clinical Oncology’s People Living With Cancer More than 50 studies of DAs for cancer patients have
Web site (http://www.peoplelivingwithcancer.org), and the been published, with most targeting patients considering
National Cancer Institute’s Cancer Information service primary or adjuvant treatment of early stage breast or pros-
(http://cancer.gov/cancerinfo/). tate cancer. However, DAs have been developed to address
Only those patients who understand their prognosis issues of cancer screening, genetic testing, cancer preven-
can make informed choices about their care, or participate tion, and, less commonly, the management of advanced
in treatment decisions in accordance with their values. But cancer. Many of these are developmental studies, but more
even when patients receive the required information, a sig- groups are testing DAs through randomized trial designs.
nificant proportion misunderstands this information. The Whelan et al14 conducted a randomized trial in patients
reasons for this are multifactorial, relating to physician and with node-negative breast cancer considering chemother-
patient communication techniques, information overload, apy, comparing the effect of the standard medical consulta-
as well as patient anxiety and possible denial. Studies have tion versus the consultation plus use of a validated decision
shown that as many as a third of cancer patients misunder- board (DB). Patients randomly assigned to use of the DB
stand information they have received.9,10 For example, in a demonstrated a significant improvement in knowledge
about disease and treatment, and greater satisfaction with to help patients decide between supportive care (or radical
decision making up to 12 months after the initial consulta- radiotherapy in locally advanced disease), with or without
tion; however they exhibited no statistically significant dif- chemotherapy.18,19 The aids describe the different treat-
ference in patient anxiety, consultation duration, or treat- ment options and the adverse effects of each treatment
ment choice between the two arms, nor in physician choice, including an assessment of the effect of each treatment
satisfaction with decision making. Patients randomly as- choice on the patient’s physical and social functioning. This
signed to use the DB were more active participants in their information is followed by a structured interview to help the
medical decision making than those randomly assigned to patient work through trade-off exercises to clarify the patient’s
the standard consultation. Ravdin et al15 also presented the values for the outcomes of median survival and 1-year or
results of a randomized trial in patients with early-stage 3-year survival with either treatment option. Both have been
breast cancer, using the computer program “Adjuvant!” demonstrated to be feasible and useful.
(http://www.adjuvantonline.com) to assist patients and For metastatic prostate cancer, one DA provided a
physicians with adjuvant systemic therapy decisions. Patients letter that patients took home, explaining two potential
randomly assigned to use the computer program were signifi- hormonal treatment options—surgical castration versus
cantly more satisfied with the decision-making process and therapy with a luteinizing hormone-releasing hormone.20
had a better understanding of adjuvant treatment. Patients More than 90% of patients and their wives indicated satis-
randomly assigned to use the program were more likely to faction with their treatment decision, revealing that they
select either a chemotherapy-containing regimen (39% v 36%; would select the same treatment again. Chadwick et al21
P ⫽ .04) or no systemic adjuvant therapy (15% v 11%; P ⫽ .04) reported a case series of 51 patients with advanced prostate
than those not randomly assigned to use the program. cancer who were offered a structured interview to assist
These encouraging results with decision-support tools them with decision making about medical or surgical orchi-
in early-stage cancer patients do not necessarily predict the dectomy, reviewing the procedures, their benefits, and ad-
success of similar tools in the setting of advanced cancer. verse effects. These patients identified treatment conve-
For patients with incurable cancer, the prognosis and treat- nience and the doctor’s recommendation as the major
ment goals differ substantially from the goals of adjuvant considerations contributing to their treatment decision.
therapy. The trade-off between benefits and toxicities of Lastly, a DA has been developed for patients with met-
palliative therapy is more complex. Advanced-stage cancer astatic colorectal cancer to facilitate decision making for the
patients often have greater need for emotional support and systemic management of their disease and to improve pa-
symptom control, in addition to their information and tient understanding about their disease and treatment op-
decision-involvement needs. Five studies have been pub- tions.22 Evidence from randomized trials and individual
lished describing the development of DAs for patients with patient meta-analyses describing the potential benefits and
metastatic cancer, and one for locally advanced cancer.16-21 toxicities of different standard treatment options, including
However, none one of these has been evaluated through a supportive care alone, was incorporated and illustrated us-
randomized trial. ing graphic formats, with a values clarification exercise. The
Elit et al16 developed and tested a DB to elicit patient aid, in the format of a take-home booklet and audiotape,
preferences for therapy in advanced epithelial ovarian can- was well received and proved to be feasible, highly accept-
cer. This DB describes two cancer chemotherapy options able to patients, valid, and reliable. Review of the DA using
for patients with suboptimally debulked ovarian cancer, a pre-/posttest design in an initial pilot study of 27 patients
with their potential adverse effects and possible outcomes. significantly improved knowledge (P ⫽ .013), and did not
Although one of the treatment options in this DB is no seem to increase anxiety. A randomized trial of the DA in
longer widely used, the use of the board made it possible to this patient population is ongoing, to evaluate its effective-
provide prognostic information to 98% of patients, which ness in the clinical setting with respect to patient knowl-
was previously uncommon, and the DB was shown to be a edge, decisional conflict, decision satisfaction, treatment
reliable, valid method of sharing information about ad- decisions made, and patient well-being.
vanced ovarian cancer with patients.
DAs have also been developed for patients with lung CONCLUSION
cancer. Fiset et al,17 developed a DA for patients with met-
astatic non–small-cell lung cancer—a workbook and an Treatment decisions in advanced cancer are difficult. The
audiotape for patients to take home. Using a pre-/posttest uncertain benefits of systemic anticancer treatment must be
design, the aid was demonstrated to improve patient weighed against likely toxicity, in a situation where the goal
knowledge of options and outcomes, and reduce decisional of treatment is not cure. While many decision supports exist
conflict. Most physicians reviewing the aid found it accept- for early-stage cancer patients, there are few for advanced
able. For patients with locally advanced and metastatic cancer patients, who arguably have a greater need for deci-
non–small-cell lung cancer, Brundage et al developed DAs sion support.
www.jco.org 1761
The potential benefits of DAs include enhanced patient of treatment decisions at the end of life with patient values,
understanding, reduced decisional conflict, enhanced har- and potentially improving patient well being.
monization between patient values and treatment deci-
sions, and greater patient involvement and satisfaction with THE CASE REVISITED
decision making. Two systematic reviews have been re-
ported regarding 18 randomized trials that studied the ef- The patient participated in a developmental study of a DA
fects of DAs on improving patient decision making and for advanced colorectal cancer patients. He learned, among
patient outcomes in patients with early-stage cancer.13,23 other things, that supportive care alone was a valid option
These meta-analyses demonstrate that, compared with con- for the management of advanced cancer, as is expectant
trols, most of the DAs studied produced higher patient management in the case of the asymptomatic patient. He
knowledge scores, lower decisional-conflict scores, and also learned more about the potential survival benefit from
more active patient participation in decision making. No chemotherapy, and the concept of progression-free sur-
differences in anxiety or satisfaction with decisions were vival. On reflection, he felt that it had been worthwhile
seen.13 The effectiveness of these promising decision sup- trying the chemotherapy, but, given the toxicities that he
port tools for advanced-stage cancer patients and their role had incurred with chemotherapy, he approached his oncol-
in current oncology practice should be better defined ogist at his next visit about stopping treatment and observ-
through additional randomized trials. DAs and supports to ing the course of his disease with supportive care alone.
facilitate decision making for patients grappling with the
■ ■ ■
diagnosis of advanced cancer can further our goal of im-
proving the quality of decision making and overall care in Authors’ Disclosures of Potential
advanced-stage cancer patients, through enhancing in- Conflicts of Interest
formed consent, decision satisfaction, better harmonizing The authors indicated no potential conflicts of interest.
9. MacKillop WJ, Stewart WE, Ginsberg AD, application of a bedside decision instrument.
REFERENCES et al: Cancer patients’ perceptions of their dis- Gynecol Oncol 62:329-335, 1996
ease and its treatment. Br J Cancer 58:355-358, 17. Fiset V, O’Connor AM, Evans W, et al:
1. Fallowfield LJ, Hall A, Maguire P, et al: 1988 Development and evaluation of a decision aid for
Psychological effects of being offered choice of 10. Gattellari M, Butow PN, Tattersall MH, et patients with stage IV non-small cell lung cancer.
surgery for breast cancer. BMJ 309:448, 1994 al: Misunderstanding in cancer patients: Why Health Expect 3:125-136, 2000
2. Street RL, Voigt B: Patient participation in shoot the messenger? Ann Oncol 10:39-46, 18. Brundage MD, Feldman-Steward D,
deciding breast cancer treatment and subse- 1999 Cosby R, et al: Cancer patients’ attitudes toward
quent quality of life. Med Decis Making 17:298- 11. Weeks JC, Cook EF, O’Day SJ, et al: treatment options for advanced non-small cell
306, 1997 Relationship between cancer patients’ predic- lung cancer: Implications for patient education
3. Kaplan SH, Greenfield S, Gandek B, et al: tions of prognosis and their treatment prefer- and decision support. Patient Educ Couns 45:
Characteristics of physicians with participatory ences. JAMA 279:1709-1714, 1998 149-157, 2001
decision-making styles. Ann Intern Med 124: 12. O’Connor AM, Rostom A, Fiset V, et al: 19. Brundage MD, Feldman-Stewart D, Cosby
497-504, 1996 Decision aids for patients facing health treat-
R, et al: Phase I study of a decision aid for
4. Gattellari M, Butow PN, Tattersall MHN: ment or screening decisions: Systematic review.
patients with locally advanced non-small-cell
Sharing cancer decisions. Soc Sci Med 52:1865- BMJ 319:731-734, 1999
lung cancer. J Clin Oncol 19:1326-1335, 2001
1878, 2001 13. Earle CC, Neville BA, Landrum MB, et al:
20. Cassileth BR, Soloway MS, Vogelzang NJ,
Trends in the aggressiveness of cancer care near
5. Degner LF, Kristjanson LJ, Bowman D, et et al: Patients’ choice of treatment in stage D
the end of life. Proc Am Soc Clin Oncol 22:522,
al: Information needs and decisional preferences prostate cancer. Urology 33:S57-S62, 1989
2003 (abstr 2100)
in women with breast cancer. JAMA 277:1485- (suppl 5)
14. Whelan TJ, Sawka C, Levine M, et al:
1492, 1997 21. Chadwick DJ, Gillatt DA, Gingell JC: Med-
Helping patients make informed choices: A ran-
6. Beaver K, Bogg J, Luker KA: Decision- ical or surgical orchidectomy: The patients’
domized trial of a decision aid for adjuvant che-
making role preferences and information needs: choice. BMJ 302:572, 1991
motherapy in lymph node-negative breast can-
A comparison of colorectal and breast cancer. cer. J Natl Cancer Inst 95:581-587, 2003 22. Leighl NB, Butow PN, Clarke SJ, et al:
Health Expect 2:266-276, 1999 15. Ravdin PM, Siminoff L, Hewlett J, et al: When the goal of treatment is not cure: Devel-
7. Davison BJ, Degner LF, Morgan TR: Infor- Evaluation of impact of communication tool gen- opment of a decision aid for patients with met-
mation and decision-making preferences of men erated by the computer program, Adjuvant!, on astatic colorectal cancer—Presented at Medical
with prostate cancer. Oncol Nurs Forum 22: patients with early breast cancer and their doc- Oncology Group of Australia Annual Meeting,
1401-1408, 1995 tors. Proc Am Soc Clin Oncol 20:31a, 2001 (abstr Sunshine Coast, Queensland, Australia, August
8. Gattellari M, Voigt K, Butow PN, et al: 119) 10-12, 2001
When the treatment goal is not cure: Are cancer 16. Elit LM, Levine MN, Gafni A, et al: Pa- 23. Molenaar S, Sprangers MA, Postma-
patients equipped to make informed decisions? tients’ preferences for therapy in advanced epi- Schmit FC, et al: Feasibility and effects of deci-
J Clin Oncol 20:503-513, 2002 thelial ovarian cancer: Development, testing, and sion aids. Med Decis Making 20:112-127, 2000