Annals of Internal Medicine 2002 (Oct 15); 137 (8): 660–664 ~ FULL TEXT
David M. Eisenberg, MD; Roger B. Davis, ScD; Susan L. Ettner, PhD;
Scott Appel, MS; Sonja Wilkey; Maria Van Rompay; Ronald C. Kessler, PhD
Department of Obstetrics and Gynecology UHNSO,
Oregon Health and Science University,
3181 SW Sam Jackson Park Road,
Portland, OR 97239, USA.
Increasing use of complementary and alternative medical (CAM) therapies by patients, health care providers, and institutions has made it imperative that physicians consider their ethical obligations when recommending, tolerating, or proscribing these therapies. The authors present a risk-benefit framework that can be applied to determine the appropriateness of using CAM therapies in various clinical scenarios. The major relevant issues are the severity and acuteness of illness; the curability of the illness by conventional forms of treatment; the degree of invasiveness, associated toxicities, and side effects of the conventional treatment; the availability and quality of evidence of utility and safety of the desired CAM treatment; the level of understanding of risks and benefits of the CAM treatment combined with the patient's knowing and voluntary acceptance of those risks; and the patient's persistence of intention to use CAM therapies. Even in the absence of scientific evidence for CAM therapies, by considering these relevant issues, providers can formulate a plan that is clinically sound, ethically appropriate, and targeted to the unique circumstances of individual patients. Physicians are encouraged to remain engaged in problem-solving with their patients and to attempt to elucidate and clarify the patient's core values and beliefs when counseling about CAM therapies.
From the FULL TEXT Article:
Because the use of complementary and alternative medical
(CAM) therapies is becoming increasingly widespread
in the United States [1, 2], physicians with traditional
training now confront a complex array of ethical
dilemmas. The conventional physician in the United States
is relatively unfamiliar with CAM therapies, a lack of consensus
exists among orthodox health care providers about
CAM use within conventional medicine, and scientific evidence
for CAM therapies is rapidly shifting. Such an environment
creates a potential for serious physician–patient
conflict, as well as patient harm, when CAM and conventional
medical therapies intersect. [3–5] Some CAM therapies
in and of themselves can be hazardous, but harm can
also occur indirectly when patients choose less effective
CAM treatments instead of conventional methods that
have demonstrated efficacy. In individual situations in
which evidence supporting CAM therapies is good, more
subtle forms of harm can manifest when a provider, either
through bias against such therapies or lack of knowledge,
presents conventional treatment as the only available option.
In such situations, a patient could be deprived of
potentially useful CAM treatments, may be unable to actively
participate in decision making, and may be left without
medical supervision and monitoring while using CAM
Physician–patient conflict may also occur when the
patient considers the use of CAM treatments that are not
supported by evidence-based medicine. Although some
CAM therapies have been subjected to randomized, controlled
trials, evidence for many types of CAM therapies is
quite thin or simply does not exist. Some of these therapies
also have a strong spiritual component that relates directly
to a person’s belief system, and some patients choose such
treatments despite a lack of evidence of benefit—or worse,
despite clear evidence that such treatments are ineffective.
The challenge for physicians in such situations is to provide
ethical, medically responsible counseling that respects
and acknowledges the patient’s values. Physicians should
not violate their own values in the process of responding to
a patient’s needs or abandon the practice of evidence-based
medicine in order to provide support for their patients’
beliefs. A long-standing, carefully nurtured partnership between
physician and patient that has grown over time may
be strained or completely destroyed if common ground in
such situations cannot be found. Choosing to withdraw
from care of a patient because of a fundamental conflict is
a decision that requires intense self-examination by the
physician and consultation with trusted colleagues; it is a
decision that is never made lightly.  Physician–patient
conflict arising from issues related to use of CAM therapies
is common, but the need to withdraw from care should be
rare if physicians are willing to engage creatively with patients
when problems seem unsolvable. [7, 8] However,
even physicians who are willing to consider CAM treatments
in some circumstances find it difficult to know how
to responsibly and ethically advise patients in this unfamiliar
realm of medicine. They feel uninformed about CAM
therapies and ill-equipped to judge the quality of CAM
We present two prototypical cases that illustrate the
challenges already discussed as well as other ethical challenges
that will arise as use of CAM therapies in the United
States becomes even more common. The use of case-based
reasoning, or casuistry, in questions of CAM therapy use
allows a fuller appreciation of the way in which circumstances
play an intrinsic role in moral judgments. [9, 10]
Specific details of each case are particularly relevant in
guiding ethical decision making (Table). This information
forms a risk– benefit framework in which varying degrees
of illness, efficacy, safety, and patient choice help guide a
conventional physician to recommend, tolerate, or proscribe
the use of CAM therapies. Application of this familiar
type of risk– benefit analysis to questions of CAM use
can help providers develop a treatment plan that responsibly fulfills their ethical obligations while recognizing and
allowing consideration of the unique circumstances of each
Ms. P., a 52–year-old woman, sees a naturopath for
her primary care. The naturopath refers Ms. P. to a gynecologist
for evaluation of an abnormal Papanicolaou smear
performed at an annual examination in the naturopath’s
office. Cervical biopsies reveal adenocarcinoma in situ, a
premalignant condition that, if left untreated, can progress
to invasive adenocarcinoma. The gynecologist recommends
a cervical conization or hysterectomy to ensure that all
abnormal tissue is identified and removed. Ms. P. is a Reiki
practitioner, trained to perform a type of energy healing
that is said to involve moving energy through the body in
order to balance it. She tells the physician that she plans to
pursue meditation, colonics, and yoga and to work with
her Reiki master rather than have surgery. She further states
that she does not want any “surgical invasion of her bodily
integrity.” The gynecologist feels unable to change Ms. P.’s
mind, is concerned about her prognosis, and is unsure
about remaining in the physician–patient relationship.
Analysis of this case begins with consideration of the
factors presented in the Table. Ms. P. has a precancerous
condition that, if left untreated, has a high likelihood of
progressing to invasive cancer within several years. The
conventional treatment offered is admittedly invasive but
at this point should be completely curative. Ms. P. is
highly motivated to seek CAM therapies, but although
there are some data on the beneficial effect of her chosen
treatments for some conditions [11, 12], no data support
successful treatment of a precancerous condition. In addition,
these treatments are generally used to increase overall
well-being rather than to cure a specific physiologic abnormality.
Ms. P. is legally competent; the law presumes persons
are competent unless a judge has ruled them incompetent.
However, a physician faced with a seeming inexplicable
decision by a patient must explore the patient’s capacity for
decision making. Capacity requires at least an ability to
understand relevant information, to appreciate one’s medical
choices and their possible consequences, to communicate
a choice, and to engage in rational deliberation about
one’s own values in relation to the physician’s recommendations
about treatment options.  Review of Ms. P.’s
records showed that she had no psychiatric history, but
that she had been sexually abused as a child and, as an
adult, had received counseling regarding this history. She
had been practicing for the previous 4 years as a Reiki
therapist, trained to move energy throughout the body in a
way that is said to provide balance and healing. Ms. P.
exhibited no verbal or behavioral signs of mental illness.
Once the physician was satisfied that the patient’s
decision-making capacity was adequate, the physician had
to ensure that Ms. P. was fully informed by carefully explaining
the potential risks involved and attempting to persuade
her to change her mind. The physician approached
the patient by saying, “As far as I am able to tell, you don’t
have cancer now, but the condition you have often does
progress to cancer, and the more time that passes before
you have these cells removed, the more likely it is that they
eventually will turn into cancer. I’m concerned that if you
delay treatment by trying some of these other methods,
none of which have been shown to be effective in treating
your condition, you may end up with cancer, which will be
much more difficult, if not impossible, to cure.” The patient
should show understanding of the risks, a willingness
and ability to communicate about them, and a capacity to
Ms. P. listened carefully to everything the physician
said, and, as the physician listened equally intently, it became
clear to him that Ms. P.’s choice reflected her values
of noninvasion of the body; belief in the body’s ability to
heal itself; and the importance of the relationship between
illness and mental, emotional, and spiritual health. Although
she had had chaotic and difficult early years, Ms. P.
had found solace and peacefulness in her beliefs, which
influenced every part of her life. She not only practiced
Reiki herself but taught the techniques to others and had
even established a newsletter that was circulated among
other Reiki practitioners and their clients. At one point,
Ms. P. stated firmly and calmly, “Even if you told me I
would die next month without surgery, I still wouldn’t
have it.” The physician began to feel that not only would
any further attempt to convince Ms. P. to undergo surgery
be disrespectful of these deeply held core values, but that
her internal sense of integrity could be irretrievably harmed
were she to undergo a procedure that was not congruent
with her beliefs.
As a general principle, the personal beliefs and choices
of other persons should be respected if they pose no threat
to other parties.  The dilemma for the physician in this
case became whether to continue providing care for Ms. P.
when it was clear that her values were deeply held and
durable: There was really no expectation that she would
change her mind and agree to the physician’s recommendation.
The resolution in this case was relatively straightforward.
Because Ms. P. had seen the gynecologist only in
consultation, she was free to reject the recommendation for
surgery and return to her naturopath for continued medical
care, with an invitation to return to the physician if she
desired further discussion. Informed refusal of care was
carefully documented, and the physician contacted the naturopath,
who felt able to continue to provide care for Ms.
P. despite her refusal of surgery.
The question of withdrawal from care becomes more
difficult when the physician in question is the patient’s
primary care provider; in this situation, the ongoing relationship
with the patient must be considered. The ethical
obligation of nonabandonment emphasizes the longitudinal
nature of a caring commitment between physician and
patient. The promise to face the future together is a central
obligation of the physician–patient relationship, which carries
with it a commitment by the physician to jointly seek
solutions with patients throughout their illnesses. [6, 14, 15]The limits of this obligation will vary, depending on
such things as the length of time of the relationship, the
patient’s other medical conditions, and the patient’s desire
for continued care from this physician. Ms. P.’s refusal of
surgery may be seen by some physicians as an understandable—
albeit incorrect—decision, given the totality of her
circumstances. These physicians would feel able to continue
as her primary care provider. In the case of Ms. P.,
the physician should explain his viewpoint—that he considers
her decision a serious mistake but is willing to continue
providing long-term care for her, even if she develops
cancer. Others may find her decision unsupportable, feeling
that continuing to provide care would involve giving
tacit approval to an irresponsible decision. Such physicians
would opt to withdraw from her care and make a referral
to a more like-minded physician. But the obligation of
nonabandonment challenges us to meet and engage with
our patients in conflicts, not shy away from them by recourse
to unbending rules or blindly following some imaginary,
clearcut boundary about what is acceptable or unacceptable.  Patients do not always expect medical solutions to their problems, and withdrawal of care should
happen only when the physician’s own values compel such
a decision. Physicians who feel that remaining in the relationship
provides false reassurance to the patient should
review the patient’s decision-making process (and remind
themselves that the patient came to this choice with full
awareness of the possible consequences) and should avoid
overemphasizing their own role as the patient carries out
his or her decision.
Ms. L., a 48–year-old woman with recurrent metastatic
ovarian adenocarcinoma, seeks an oncologist who will provide
her with conventional treatment but is open to the use
of CAM treatments as adjunctive therapy. She wants her
provider to assist her in evaluating available CAM treatments
and give guidance about their appropriateness for
her. She raises the issue with her oncologist, who dismisses
the subject, explaining that she cannot recommend CAM
treatments because they have not been subjected to scientific
scrutiny and that the patient is risking her health in
seeking these types of care. The physician further comments
that if there were any evidence that CAM therapies
could help, she would offer them herself. Ms. L. cannot
switch providers because her insurance plan requires she be
seen by this oncologist, and she feels frustrated in her attempts
to engage with her physician in this aspect of her care.
Ms. L. has requested information regarding the effectiveness
of CAM treatments as an adjunct to, not as a
substitute for, the planned conventional treatment of her
metastatic cancer. Her oncologist states that there is no
evidence that any CAM therapies are useful in this regard.
Actually, however, data support the use of various types of
CAM treatments for patients with cancer. For example, the
positive effects of relaxation training for improving anxiety
and decreasing pain ; of acupuncture for diminishing
nausea associated with chemotherapy ; and of psychotherapy,
group therapy, relaxation, and imagery for improving
the quality of life in patients with breast cancer  have all been suggested through controlled trials. In
contrast, many herbal and metabolic regimens that have
been proposed for the primary treatment of cancer are currently
not supported by scientific evidence.  The other
relevant factors in this case are that Ms. L.’s condition is
chronic and severe, with a low expectation of cure with
traditional methods, and that she is highly motivated to
use CAM therapies, which may improve her quality of life.
In undertaking CAM treatments, she may also regain some
sense of control over her physical, psychological, and spiritual
health. Also of importance is that she is actively seeking
the guidance and advice of her conventional caregiver.
In such a situation, the patient’s choice of CAM therapies
should be respected. The increasing body of research
regarding the effectiveness of some types of CAM therapies [19–21] helps clinicians to judge the advisability of these
therapies according to the following criteria. As suggested
by other authors, in making such judgments, consideration
should be given to whether available evidence 1) supports
both safety and efficacy; 2) supports safety but is inconclusive
about efficacy; 3) supports efficacy but is inconclusive
about safety; or 4) indicates either serious risk or inefficacy. 
If evidence supports both safety and efficacy, the physician
should recommend the therapy but continue to
monitor the patient conventionally. If evidence supports
safety but is inconclusive about efficacy, the treatment
should be cautiously tolerated and monitored for effectiveness.
If evidence supports efficacy but is inconclusive about
safety, the therapy still could be tolerated and monitored
closely for safety. Finally, therapies for which evidence indicates either serious risk or inefficacy obviously should be
avoided and patients actively discouraged from pursuing
such a course of treatment. This risk framework for CAM
therapies allows the physician to make a thoughtful decision
that will be evidence based, ethically appropriate, and
legally reasonable.  In the case of Ms. L., the oncologist
could therefore reasonably respond to the patient’s request
for guidance about CAM therapies by considering acupuncture
for the patient’s chemotherapy-associated nausea
and mind–body techniques, such as relaxation and imagery,
for anxiety and pain.
It is not difficult for a physician to counsel a patient
about CAM therapies for which evidence exists. No one
would disagree that physicians should be aware of pertinent
evidence and be willing to consider any intervention,
CAM or allopathic, that has an acceptable risk–benefit
balance; apprise the patient of acceptable options; and
make a recommendation. In fact, the obligation of informed
consent requires that physicians raise and discuss
CAM therapy options that have been shown to be efficacious.  An example of one such option is lifestyle programs
that include yoga, meditation, exercise, and dietary
changes for the treatment of heart disease.  Likewise, a
physician should take the initiative to proactively steer patients
away from treatments that are known to be dangerous
or have been associated with clinically significant adverse
interactions with other supplements or medications. [23–26] CAM therapies have become so common that
inquiries about their use should become a routine part of
history taking. 
There are currently many situations for which no reliable
evidence about CAM therapies exists but patients
nonetheless request these treatments; in such cases, physicians
must counsel and advise in the absence of evidence.
When there is no evidence either for or against a particular
therapy, physicians can choose to tolerate and monitor or
actively discourage use of CAM treatments. The risk–benefit
framework presented in the Table is helpful in counseling
because the unknown CAM therapy can be compared
with what is known about the competing
conventional treatment and the relative risk of choosing
the CAM therapy can be assessed. If, for example, the
conventional treatment is effective and the risk for not
treating is great, a patient would be ill-advised to pursue an
unproven CAM therapy, and the physician should actively
discourage such a decision. If, however, the standard conventional
therapy is ineffective, even an unproven CAM
therapy could be tolerated, because the patient has few, if
any, good alternatives. An unproven therapy may have unknown
toxicities, and patients should be clearly informed
about the lack of information on the safety of untested
CAM therapies, particularly if the condition being treated
Such discussions between physician and patient are the
heart of informed consent, and a physician’s recommendations
are often the beginning rather than the ending of an
exchange that will ultimately determine the course the patient
chooses. It can be helpful to begin the discussion by
focusing on general goals of treatment (for example, care
vs. cure) rather than moving immediately to a consideration
of specific interventions, which may lead the patient
to prematurely choose therapy that may not serve his or
her ultimate goals well.  The final choice of treatment
belongs to the patient and should reflect his or her beliefs
and values; physicians should seriously consider and try to
respond to the patient’s needs to the fullest possible extent
without violating their own values. For example, if Ms. L.
refused chemotherapy in favor of some unproven herbal
preparation, the approach should be similar to that of the
physician caring for Ms. P. She and her physician should
first attempt to reach a common understanding of the nature
of her condition, her prognosis, and the goals of treatment.
If she persisted in rejecting chemotherapy, her physician
would have an appreciation of the reasons
supporting this choice and would hopefully feel able to
provide Ms. L. with a continued open-ended commitment
to long-term joint problem-solving despite her refusal of
conventional care. 
Physicians routinely refer patients to other practitioners
and can be expected to be able to judge the skill of
their physician colleagues; however, physicians with no
training in a particular complementary therapy, such as
acupuncture, will find it difficult to be a reasonable judge
of CAM providers. Therefore, it is of paramount importance
that physicians have some understanding of currently
existing methods of credentialing various types of CAM
providers  as well as the potential liability associated
with these referral relationships. [22, 30] The reader is
referred to the considerations of credentialing CAM providers
and malpractice liability associated with CAM discussed
elsewhere in this series. [22, 29]
Increasing use of CAM therapies in the United States
provides an opportunity to reexamine the ethical foundations
of western medical practice with renewed attention to
the commitment physicians make when entering into a
caring relationship with a patient. Physicians routinely balance
risks and benefits in decision making, but the advent
of CAM therapies challenges physicians to deal responsibly
with paradigms of healing that fall outside the boundaries
of conventional medical practice and to make decisions in
these unfamiliar realms, often in the absence of evidence.
Specific details of each case should be factored into a risk–
benefit assessment so that a plan of treatment that is clinically
reasonable and ethically appropriate can be developed.
The commitment to joint problem solving over time
that is a central obligation of the physician–patient relationship
becomes even more important when considering
the use of CAM therapies. Elucidating patients’ experiences
of illness, their hopes and values, and what they see
as their best interests is vital if physicians and patients are
to find common ground when making decisions in areas of
uncertainty. As the body of evidence regarding CAM therapies
grows, we hope that the model of decision making we
have presented will provide an ethical structure for medical
practice in which physicians routinely combine the powerful
tools of conventional medicine with those CAM therapies
shown to be worthy of clinical integration.
The authors thank the two patients who allowed us
to share their stories, Debbie Mosley for invaluable technical assistance,
and Martin Donohoe for critical comments.
By unrestricted educational grants from the American
Specialty Health Plans, San Diego, California; the Medtronic Foundation,
Minneapolis, Minnesota; and the Friends of Beth Israel Deaconess
Medical Center, Boston, Massachusetts.
Eisenberg DM, Davis RB, Ettner SL, Appel S, Wilkey S, Van Rompay M, Kessler RC.
Trends in Alternative Medicine Use in the United States, 1990 to 1997:
Results of a Follow-up National Survey
JAMA 1998 (Nov 11); 280 (18): 1569–1575
The Invisible Mainstream.
Harvard Medical Alumni Bulletin. 1996;70:20-5.
Angell M, Kassirer JP.
Alternative Medicine: The Risks of Untested and Unregulated Remedies: A Medical Opinion
New England Journal of Medicine 1998 (Sep 17); 339 (12): 839-841
Classical medicine v alternative medical practices.
J Med Ethics. 1992;18:18-22. [PMID: 1573644]
Why I would not recommend complementary or alternative therapies: a physician’s perspective.
Rheum Dis Clin North Am. 1999;25:833-43, vii. [PMID: 10573760]
Quill TE, Cassel CK.
Nonabandonment: a central obligation for physicians.
Ann Intern Med. 1995;122:368-74. [PMID: 7847649]
Partnerships in patient care: a contractual approach.
Ann Intern Med. 1983;98:228-34. [PMID: 6824257]
Emanuel EJ, Emanuel LL.
Four models of the physician-patient relationship.
JAMA. 1992;267:2221-6. [PMID: 1556799]
Casuistry and clinical ethics.
In: Jecker NS, Jonsen AR, Pearlman RA, eds.
Bioethics: An Introduction to the History, Methods and Practice.
New York: Jones and Bartlett; 1997:158-61.
Casuistry: an alternative or complement to principles?
Kennedy Inst Ethics J. 1995;5:237-51. [PMID: 11645308]
Ornish D, Scherwitz LW, Doody RS, Kesten D, McLanahan SM, Brown
SE, et al.
Effects of stress management training and dietary changes in treating ischemic heart disease.
JAMA. 1983;249:54-9. [PMID: 6336794]
Long L, Huntley A, Ernst E.
Which complementary and alternative therapies benefit which conditions? A survey of the opinions of 223 professional organizations.
Complement Ther Med. 2001;9:178-85. [PMID: 11926432].
Jonsen AR, Siegler M, Winslade WJ.
Clinical Ethics. 3rd ed.
New York: McGraw-Hill; 1992.
The Physician’s Covenant: Images of the Healer in Medical Ethics.
Philadelphia: Westminster Pr; 1983.
The Patient as Person: Explorations in Medical Ethics.
New Haven: Yale Univ Pr; 1970.
Integration of Behavioral and Relaxation Approaches into the Treatment of
Chronic Pain and Insomnia.
NIH Technology Assessment Statement, NIH Publication No. PB96113964.
Bethesda, MD: National Institutes of Health; 1995: 1-34.
NIH Consensus Conference.
JAMA. 1998;280:1518-24. [PMID: 9809733]
Spiegel D, Bloom JR, Kraemer HC, Gottheil E.
Effect of psychosocial treatment on survival of patients with metastatic breast cancer.
Lancet. 1989;2: 888-91. [PMID: 2571815]
Alternative Medicine: What Works.
Baltimore: Lippincott Williams & Wilkins; 1997.
Jonas W, Levin J, eds.
Essentials of Complementary and Alternative Medicine.
Philadelphia: Lippincott Williams & Wilkins; 1999.
Spencer J, Jacobs J, eds.
Complementary/Alternative Medicine: An Evidence-Based Approach.
New York: Mosby; 1999.
Cohen MH, Eisenberg DM.
Potential physician malpractice liability associated with complementary and integrative medical therapies.
Ann Intern Med. 2002;136:596-603. [PMID: 11955028]
Beyond Complementary Medicine: Legal and Ethical Perspectives on Health Care and Human Evolution.
Ann Arbor, MI: Univ of Michigan
Second thoughts about safety of St John’s wort.
Lancet. 1999;354: 2014-6. [PMID: 10636361]
Lancet. 2000;355:134-8. [PMID: 10675182]
Piscitelli SC, Burstein AH, Chaitt D, Alfaro RM, Falloon J.
Indinavir concentrations and St John’s wort [Letter].
Lancet. 2000;355:547-8. [PMID: 10683007]
Advising Patients Who Seek Alternative Medical Therapies
Annals of Internal Medicine 1997 (Jul 1); 127: 61-69
Quill TE, Brody H.
Physician recommendations and patient autonomy: finding a balance between physician power and patient choice.
Ann Intern Med. 1996;125:763-9. [PMID: 8929011]
Eisenberg DM, Cohen MH, Hrbek A,Grayzel J, Cooper RA.
Credentialing complementary and alternative providers.
Ann Intern Med. [In press].
Studdert DM, Eisenberg DM, Miller FH, Curto DA, Kaptchuk TJ, Brennan
Medical malpractice implications of alternative medicine.
JAMA. 1998; 280:1610-5. [PMID: 9820265]
Return to EISENBERG's CAM ARTICLES